forked from airwindows/airwindows
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathAirwindopedia.txt
4826 lines (2462 loc) · 828 KB
/
Airwindopedia.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
Welcome to Airwindopedia! This is a complilation of the plugin-related bits from all my plugin posts. As such, this is as close to a full documentation of the Airwindows plugins as you'll find. In some cases I'll have added explanation, or I might be able to expand on things as this develops. Bear with me as I must do this while also doing everything else, and it will always be on me: as an open source dev, my experience is that people don't really step up to do the tedious stuff across entire libraries, so Airwindopedia is there because I went through all the posts to consolidate them in one place for you :)
We will start with some basic categorization: you can look these up below for further instruction. We'll use the categorization that was used in Surge Synthesizer when organizing my plugins for their purposes, but I will continue and mention plugins not used in Surge Synthesizer if I feel they fit the category. They won't be in alphabetical order, but in order of how well they fit the category and how good they are, and there will be plugins showing up in multiple categories and ones that don't appear at all.
Actually it turns out that other devs can use this file to do amazing things like populate menus and fill in entries for help files, so what you see below will end up including ALL the plugins eventually. Again, they're roughly in order of goodness, so try the ones at the beginning of the list first: those are also likely to be the most recent ones :)
# Categories
Ambience: ClearCoat, TapeDelay2, Doublelay, PitchDelay, SampleDelay, BrightAmbience3, TripleSpread, Melt, MV2, MV, ADT, kChamberAR, StereoChorus, PurestEcho, TapeDelay, StarChild2, StarChild, Hombre, BrightAmbience2, BrightAmbience, StereoEnsemble, StereoDoubler, Chorus, ChorusEnsemble, Ensemble
Amp Sims: GrindAmp, FireAmp, LeadAmp, CrickBass, LilAmp, MidAmp, BigAmp, Cabs, BassDrive, BassAmp
Bass: OrbitKick, Hermepass, BassKit, DubCenter, DubSub, Floor, Infrasonic, FathomFive
Biquads: BiquadPlus, Biquad, BiquadDouble, BiquadOneHalf, BiquadTriple, Biquad2
Brightness: Air3, PlatinumSlew, DeBess, GoldenSlew, Sinew, SlewSonic, Acceleration2, DeEss, Smooth, EverySlew, Slew3, Slew2, Slew, Air2, Air, PurestAir, Acceleration, DeHiss, Hypersonic, HypersonX, Ultrasonic, UltrasonicLite, UltrasonicMed, UltrasonX
Clipping: ClipOnly2, ADClip8, ClipSoftly, OneCornerClip, ADClip7, AQuickVoiceClip, ClipOnly, FinalClip
Consoles: ConsoleLABuss, ConsoleLAChannel, ConsoleMCBuss, ConsoleMCChannel, ConsoleMDBuss, ConsoleMDChannel, Console8LiteChannel, Console8LiteBuss, Console8BussHype, Console8BussIn, Console8BussOut, Console8ChannelHype, Console8ChannelIn, Console8ChannelOut, Console8SubHype, Console8SubIn, Console8SubOut, Console0Channel, Console0Buss, PurestConsole3Buss, PurestConsole3Channel, PurestConsole2Buss, PurestConsole2Channel, Console7Buss, Console7Channel, Console7Cascade, Console7Crunch, PurestConsoleBuss, PurestConsoleChannel,Console6Buss, Console6Channel, PDBuss, PDChannel, Console5Buss, Console5Channel, Console5DarkCh, C5RawBuss, C5RawChannel, AtmosphereBuss, AtmosphereChannel, Console4Buss, Console4Channel, EveryConsole
Distortion: Distortion, Edge, Dirt, Mackity, ZOutputStage, Density2, Density, Drive, Loud, Hard Vacuum, HighImpact, MultiBandDistortion
Dithers: Dark, PaulWide, PaulDither, TPDFWide, TPDFDither, NotJustAnotherDither, NotJustAnotherCD, Beam, TapeDither, SpatializeDither, VinylDither, DoublePaul, Ditherbox, BuildATPDF, NodeDither, StudioTan, DitherMeTimbers, DitherMeDiskers, RawTimbers, RawGlitters, NaturalizeDither, HighGlossDither, DitherFloat
Dynamics: Pop2, Pressure5, Dynamics, Pop, DigitalBlack, Logical4, VariMu, ButterComp2, curve, Overheads, Recurve, Pyewacket, BlockParty, SoftGate, Thunder, Compresaturator, DrumSlam, ButterComp, BrassRider, Point, Gatelope, PodcastDeluxe, Podcast, Swell, PurestSquish, Pressure4, Surge, SurgeTide
Effects: GalacticVibe, CloudCoat, Fracture2, Dubly, Pafnuty2, PitchNasty, Trianglizer, ShortBuss, GuitarConditioner, Aura, TremoSquare, Tremolo, GlitchShifter, Gringer, Exciter, Energy2, Energy, Facet, Fracture, PowerSag2, PowerSag, Preponderant, Nikola
Filter: Baxandall2, Capacitor2, Pear, ResEQ2, SubTight, CStrip2, Weight, Isolator2, Holt2, Holt, ToneSlant, AverMatrix, Average, MackEQ, Hull2, Baxandall, Hull, EQ, Capacitor, Isolator, TapeFat, ResEQ, Lowpass2, Highpass2, Distance2, Distance, Lowpass, Highpass
Lo-Fi: Pockey2, Flutter, CrunchyGrooveWear, GrooveWear, Pockey, DeRez2, BitGlitter, DeRez, ChromeOxide, Cojones, Vibrato, Bite, Deckwrecka, DustBunny
Noise: Noise, Texturize, TexturizeMS, VoiceOfTheStarship, DarkNoise, ElectroHat, Silhouette, TapeDust
Reverb: kCathedral2, kPlateD, kPlateB, kPlateA, kPlateC, Verbity2, Galactic, Galactic2, sVerbity, Chamber2, Chamber, Infinity2, NonlinearSpace, kCathedral, Infinity, MatrixVerb, PocketVerbs, Reverb
Saturation: Creature, Huge, NCSeventeen, Tube2, Tube, Spiral2, PurestDrive, Focus, Mojo, Dyno, Spiral, UnBox, Desk4, Righteous4
Stereo: Srsly, Srsly2, Wider, StereoFX, ToVinyl4, AutoPan, LRFlipTimer, MSFlipTimer, Sidepass, SideDull
Subtlety: Hype, Shape, Inflamer, Sweeten, PurestWarm2, PurestWarm, Coils2, Interstage, PhaseNudge, Remap, SingleEndedTriode, Coils, Desk, TransDesk, TubeDesk
Tape: ToTape6, FromTape, Tape, IronOxideClassic2, IronOxide5, ToTape5, IronOxideClassic
Tone Color: BussColors4, Channel9, Apicolypse, Neverland, Elation, Calibre, Cider, Crystal, Precious, Luxor, Channel8, Channel7, Channel6, Channel5, Channel4
Utility: Monitoring3, Monitoring2, Monitoring, BitShiftPan, BitShiftGain, PurestGain, PurestFade, EveryTrim, HermeTrim, SlewOnly, SubsOnly, PeaksOnly, Golem, DCVoltage, EdIsDim, MidSide, uLawEncode, uLawDecode, RightoMono, LeftoMono, Balanced, Flipity, MoNoam, VoiceTrick
XYZ Filters: ZBandpass2, ZHighpass2, ZLowpass2, ZNotch2, ZRegion2, ZBandpass, ZHighpass, ZLowpass, ZNotch, ZRegion, YBandpass, YNotBandpass, YHighpass, YNotHighpass, YLowpass, YNotLowpass, YNotch, XBandpass. XHighpass, XLowpass, XNotch, XRegion
############ Acceleration is an acceleration limiter that tames edge, leaves brightness.
As requested, I’ve made a dedicated acceleration limiter, like the one in ToVinyl4. You don’t have to mess with Groove Wear, or highpass anything: as a result, this ends up being the ‘high end’ acceleration limiter, mastering grade and useful in situations where ToVinyl might be overkill (ToVinyl was made in an era where I still had to make ‘amazing marketable plugins with lots of features so people would pay $50 quickly before it got all pirated etc’)
Ah, I remember those days ;P
But now, it’s different because Acceleration is free! AU, Mac and PC VST (stay tuned for a surprise on that front) and if you can’t pay, go ahead and use it anyway. If you can, please do because I think the usefulness of me doing this work will become increasingly obvious, and the usefulness of me opensourcing plugin code has also become increasingly obvious. And it’s brought you Acceleration, here and now.
The reason to be excited about that is, Acceleration’s not a simple filter. It pretty much targets exactly the digital behavior that creates Gibb effect overshoots (the reason you often have to pad heavily limited material down to stop it being crackly and glare-y) and would probably work very well as literally an acceleration limiter for record cutting heads. But, in a largely digital world, it might be even more useful because you can feed it any old ITB mix, any weird edgy source material, and it’ll clean up the nasty digital brightness while not sounding like it’s doing anything. Check the video (it should be relatively obvious if you’ve got half decent tweeters) and try it out. The key is to not try and get the apparent volume of the highs to be less: instead, you’re going after the character, or the presentation. If highs seem to be coming forward with nasty spikey extra energy, Acceleration will fix that tonal problem without altering the basic mix balance.
############ Acceleration2 is Acceleration, better! Updated for high sample rate and more extreme effect.
Acceleration2 is an updated version of my high-frequency limiter, Acceleration. It’s like a brightness control that doesn’t really take away apparent brightness, just glues it really hard. Acceleration limiting is what you’d use mastering to vinyl, to avoid burning out the cutting head: it will get you a nice retro tone without obviously coloring things. It is not an ’emulation’ of any specific gear: real mastering engineers are not looking for ’emulation of Neumann sound’ or anything like that, they’re looking to get the functionality of this in the most colorless way.
It’s been updated to work exactly the same at all sample rates (the previous one had issues adapting to them) and now has more intense depth of effect: if you crank it up you can make things real soft and dull. Don’t do that :) the purpose is not to do what you could do with an EQ, the purpose is to glue things and take the edge off the super-highs without harming the sparkle and air of the recording. Acceleration2 can do that real well: you might even find it useful if you’re doing very digital mixes and avoiding my other stuff like Console7. If you put this on the mix buss of an otherwise super-digital mix, you can make the highs prettier without making them any darker or duller. Do that by applying only small amounts of the Limit control.
############ ADClip7 is the ultimate Airwindows loudness maximizer/ peak control.
Finally! The newest version of Airwindows’ famous ADClip is out, and it’s Mac and PC VST for the first time, and it’s free! If you have dense, busy mixes and you want to push their loudness to the max, this is the one you want. And, oddly, if you’re mastering for streaming services or iTunes and don’t want to do a thing besides convey your mix at a set LUFS and intersample peak level, this is still the one you want :)
Here’s how it works.
You’ve got a boost control, a soften control, an enhance control, and a popup or multi-function control with three options: normal, gain matched, and Clip Only. These are all interactive, so I’ll explain them in the context of that ‘mode selector’ popup.
In normal mode, boost makes things louder. This is a clipper, and no more polite method can ever make things louder than a clipper: you need to either give it a busy and complicated mix to mask the clip artifacts, or use it to snip off non-tonal percussive peaks, at which it excels. A clipper does NOT produce ‘limiter-like unvarying block of sound’, and you shouldn’t try to achieve that. A clipper gives you punchy but LOUD, and tries to retain all the dynamics you’re feeding it, rather than smooth them out for a ‘clean’ sound.
The soften control manages the way that high frequencies enter and exit the clips. It algorithmically reshapes the edges of your clip, stopping it from getting digital glare and fizz. This is the heart of ADClip (also present in my simpler clippers, not counting One Corner Clip, which is still upcoming).
The soften control also balances the outputs of two separate energy-fill algorithms, one for bass and one for highs. This was the response to a certain other loudness maximizer that launched proclaiming clippers were dead, and which is still promising its version 2 (and some bugfixes) while ADClip has gone far beyond it. Turned out the secret of that one was an elaborate way of massaging clipped-off loudness back into the signal, in a way that was supposed to be transparent but ‘cracked’ into artifacts when pushed too hard. The algorithms were presumably very sophisticated, which tends to just make the breaking point more obvious when you hit it.
The Airwindows version is a completely different, cruder and more direct version of doing the same thing, so when you slam the heck out of ADClip7, you get a deep bass slam that ‘overhangs’ a bit to add weight, and still fits into the clipped output. And you get the softened, analog-style clips to add ‘heat’ and overtones, but you also get a high-mids reinforcement that normally just highlights bright transients that would otherwise be lost to the clipping. And the Enhance control lets you go between purely ‘analog clip’ energy, and these added reinforcements.
The gain-matched mode has two uses. You could use it to ‘set your slam level’ in a way guaranteed not to trick you into thinking louder is better: it turns the output down, so instead of hearing everything get bigger and louder, you just listen for the point that the clipping’s adversely affecting the sound. Then you can flip back to normal mode, if that’s what you wanted (maximum loudness without blatant grunge). Or, you can ignore the slam and use gain-matched mode as I demonstrate in the video: ADClip7 already suppresses intersample peaks when they’re part of clipping, already reinforces energy lost to clipping, so you can use it in conjunction with a tool like Youlean’s loudness meter to dial in a specific intersample peak level for iTunes or other such picky streaming services. It’ll work like padding the output. and if you’re already in the ballpark LUFS-wise, ADClip is a far more sophisticated tool than just limiting and then padding the output to get to your ‘true peak’ target.
The last mode is Clip Only, and rather than selecting the various algorithm outputs individually, this version of ADClip gives you them at their respective loudnesses, combined. That means you can engage this mode to hear ONLY the clips, and check that you’re not hearing any recognizable ‘scrunch’ of continuing clippage. But since you’re also hearing the enhance outputs, you can adjust softness and enhance level to balance the stuff being introduced to the sound. My recommendation is to set the controls so no one type of artifact predominates: it’s not necessarily great to throw in a bunch of ‘enhance’ bass just because you can. If you’re hearing that much of it that you’re tempted to use it as an effect, you’re definitely also over-slamming your music.
So my recommendation in 2017 heading to 2018, is to use ADClip7 in gain-matched mode, to keep that ‘true peak’ measurement within the Mastered For iTunes requirements. I’m sure not everybody will stick to that, but I’m happy to say it is actually quite good for doing that, and if people want to smash stuff with it and enjoy the bass thud, that’s their affair. Remember a clipper makes stuff dynamic and punchy, not ‘flattened out’: use a limiter if you need dynamically flat, or perhaps both. In this modern era of replay gain, I’m going to suggest that dynamic and punchy is where it’s at. Learn the lessons of the LUFS meter! They’re available to us all, now.
############ ADClip8 is the ultimate Airwindows loudenator/biggenator.
The last time we saw an Airwindows ADClip, the full version rather than a neat and tidy stripped-down version, was October 8… of 2017. It's been more than six years since the king of loudenators saw an update, in part because ClipOnly got followed by ClipOnly2, which was such a good high-sample-rate clipper that it was all I needed for years. ClipOnly2 is still the cleanest, purest way to take small amounts of peak clipping to your signal: it's great as a safety clip.
But, while working on a new version tailored to the needs of Final Cut Pro (which is also coming), something new arose.
What would happen if the FCP version, which has a built-in slew clip and makes heavy use of Golden Ratios to arrive at its results, was adapted to a new ADClip? What if it was scaled to regular clipping levels, given a ceiling control to dial back its maximum to taste, what would happen if it was stacked up with a whole array of FinalClips in series like some of the other Airwindows plugins like to do? We're no longer dealing with ClipOnly2. Much like the original ADClip7, we're adding extra stuff and running a more complicated algorithm. But we've still got some of the special features of ADClip7, like the gain-matched mode and the Clips Only setting.
Enter ADClip8. It has more settings. They go Normal, Gain Matched, Clip Only… then, Afterburner, Explode, Nuke, Apocalypse, and lastly Apotheosis. No, really! That's the new settings.
What they represent is added stages, so Normal is a single stage (as is Gain Matched), Clip Only shows you where clips begin, and then the new modes are named for the sounds they produce, at two, three, four, five and six stages of special slew-clipped ClipOnly2 style sample-rate aware clipping. In practice, think 'bigger, bassier, thicker' as you continue to add stages. You may not always want to run the full six stages, you might want to voice it to your needs.
The interesting part is how all this affects intersample peaking. Because the process brings in slew clipping and repeated use of the ClipOnly2 algorithm, which resists harsh entry into and exit from a clipped state, dialing up ADClip8 and cranking the gain produces an intensely saturated sound that resists intersample peaking not through filtration, but mechanically. It will literally put in transition samples to resist the intersample peaks from going beyond regular clipping. The more stages, the more thoroughly it resists 'true peak'.
I would say don't use this for evil but who am I kidding. If you beat everyone with 'competitive loudness' this way, please never expect me to listen to what you did, I'll be hiding under my bed, regretting everything. :)
For what it's worth, 'Apotheosis' mode should also be a very nice way to do extremely moderate transparent clipping of challenging material: ClipOnly2 will still be more transparent but only sounds like a slightly brighter version of ADClip8 Normal mode. Apotheosis mode simply does more and I wouldn't blame anybody for preferring it. Hope you like ADClip8!
############ ADT is a double short delay tap with saturation.
ADT means ‘artificial double-tracking’. You could also call it fixed flanging. It’s a single short delay, as heard on lots of Beatles tracks. It’s now my job to explain what’s different about the Airwindows ADT, what you can do with it beyond the obvious ‘stick it on like a preset and pretend you are a Beatle’, and why you’d bother.
With the Airwindows ADT, you get two delay taps (making it A3T?) and an important feature: the mix sliders used to apply the delayed taps are ‘attenuverters’. That’s a word from Modular Synthesizer Land, which means you get both output level control and the ability to invert the output. Here’s why that matters.
If you apply a fixed delay tap, you get an effect called ‘comb filtering’ where you’re emphasizing and cancelling frequencies based on how long the delay is. You’d think that would sound really strange, but it’s the same way you hear a direct sound and also the sound bouncing off a floor or wall: we naturally hear through comb filtering rather well, which is why room design is important in studio control rooms. (you could have a bass frequency getting cancelled, be unable to hear it at all at your listening position, and yet things will still sound perfectly normal.) Applying a quick delay like this can make your sound richer and more textural, and a little more ambient. If it’s a very short delay it may not be heard as an echo at all.
But, if you’re using an INVERTED delay tap, something else happens: the shorter the delay and the closer to the volume of the dry signal, the more it’ll cancel out the bass. You’ll still get all the comb filtering effects, but you’ll also be removing lows, either the deep lows or even low midrange if the delay’s really short.
If you blend two taps that are both inverted, you can cut bass while averaging out the comb-filter effects. If the taps are in phase (not inverted) what you’re doing is reinforcing the bass, because the cancellation effects will run out below a certain frequency and just add together. All this is using very quick delays, though ADT lets you lengthen them to where they’re slapbacks. Don’t be too distracted by that, part of what makes ADT its own effect is the ability to shape the tone with delays too quick to hear as echoes.
Finally, now that you know you can cut bass using these very quick delays, or reinforce it, or any combination you like… there’s a headroom control. ADT will distort like crazy using low headroom. That can be used as a distortion effect… but it’s not just ordinary distortion, it’s a combination of Spiral into Console5’s buss (PurestConsole, for clarity and well-behavedness). These don’t perfectly cancel out. Instead, it produces a slightly leaned-out, skinnier tone to complement the way ADT fattens things up. If you’re using it to thin bass, it’ll be even more effective. If you’re reinforcing the body of the sound, it’ll color things in a subtle but interesting way. And of course if you love it, you can set up mixes that way (swap out Console5Channel for original Spiral, either on mix elements you’d like a little thinner and more energetic, or the whole thing). Sometimes there are new types of coloration that owe nothing to EQ or traditional processing: this is one of those times.
############ Air is a different bright EQ from any other. Requires 44.1K.
See Air2 for more on this. Air is the same, but simpler, and restricted to one sample rate so that if you're at a higher sample rate, the effects will shift up in frequency, possibly making them inaudible.
############ Air2 is a different bright EQ with silk tone and high sample rate support.
Here’s what you get: Air2 has three treble-boost bands, Hiss, Glitter and Air. The first two are very like what’s in Energy2, but Air2 is NOT the same: it’s a different algorithm, not done the samw way Energy2’s ‘Rat’ band is done, and it has a broader, softer air band that’s less ‘raw harsh energy injection’ and more ‘bright and pretty’.
And then, they all run through a ‘Silk’ control… inspired by but NOT the same as the Neve Portico Silk circuit. That is a real hardware transformer biased into second harmonic creation by a DC current. Air2’s Silk control is the same thing as Single Ended Triode (which you can download and use already) but only applied to the highs out of Air2, so it’s the same general concept but is not a clone of the Portico preamp. Plus if they freak out I will rename it to ‘Sow’s Ear’: it’s part of a treble brightener and the same basic functionality and as far as I’m concerned, nothing is stopping me from asymmetrically distorting highs, and Silk is the best general term for what that does. Except with mine you can push it too far for effect, because of course you can :)
And finally, unlike Energy2, the Dry/Wet control for Air2 strikes a new balance between the Energy style of increasing the effect, and the traditional Dry/Wet control. New with Air2, you can now turn it to full-wet and get ONLY the output of the brightener bands. They’re not exactly filters but they act like it, and by adjusting them against each other you can produce insanely treble-boosted sounds and wipe out everything else using Dry/Wet. The way it works is, up to 75% (0.75) you still have full volume Dry. It won’t re-balance your track, just add whatever Air2 highs you’re looking for. Then, between 0.75 and 1 (full Wet) the dry goes away, so if you crank it up all the way you get the fullest extreme, but for most normal use it’s like Energy2, taking an unvarying dry signal (which is NOT undersampled) and adding however much of the effect you like.
############ Air3 creates a new form of air-band EQ based on Kalman filtering.
This should be fun :)
Welcome to the experiment zone, once again, with Air3: an air band EQ based on an extension of Kalman filtering, which isn't normally used for audio purposes. As usual, Airwindows is bringing in algorithms that act weirdly because they come out of finance or science. In this case, it's science.
What's Kalman filtering? You'll be hearing more about it, but it's a method of trying to bring accuracy out of noisy data by projecting what the real data might be, and then incorporating sensor readings based on how closely they correspond to the projection. The key idea there is 'based on', because Kalman is all about willingness to throw out bad data and try and zero in on what's really happening.
Of course, in audio, it's ALL real. There's nothing to throw out, it's all legit. Give Air3 some audio, and it begins to do its analysis and projections based on something that doesn't exist: a hypothetical underlying sound, upon which your real-world cymbal sparkle or voice breathiness are just 'noise' to be removed. It's not even based on frequencies, it's trying to chart out a 'real' sound based on what's already a real sound.
What happens? For one, you get a rolled-off sound that is very, very convincing. It's easy to believe the sound was just that way to start with. You don't get a sense of brightness reduction, it just sounds natural. Air3 is shunting off the whole air band to a separate control, and what it leaves behind is very plausible and doesn't sound filtered. Great for reduction of tizz, glare and detail.
And then you take that redirected air band, bring it back in, and crank it up five times as loud, and Air3 comes alive.
I've not heard a better EQ for this super-high stuff. It seems really good at lighting up any audio you send through it. Bear in mind that much like both Air and Air2 before it, this is not an EQ in the normal sense at all. Air and Air2 used the algorithms I also put in Energy, and what Air3 does has more in common with DeBess or Acceleration than it does an EQ. It's going way into concepts of measuring the rate of change of the rate of change and trying to project that onto expected future samples… it's got a large amount of boost on tap, but can be a bit unpredictable. If you simply cut the Gnd control (for Ground) you get just whatever air there was, if you cut the Air control you remove it, but if you boost the Air control lots and cut Gnd, you get another sound from the sheer intensity of boosting being done.
I think this is going to be hard to beat for this purpose, especially used subtly, and expect to be including it into upcoming stuff that doesn't have to act like 'an EQ'. Still a lot of 'lettered' Console versions that need to be made, to emulate analog gear. This one's for imagining where we can go after that, for a new sort of mixing in the box that doesn't follow the constrictions of analog gear, letting you dial in tone characteristics as much as frequencies.
For now, enjoy Air3, and if you've got opinions on what needs to follow it next, drop by a livestream: you have Air3 today because a livestream viewer asked for it to come out ASAP, and here we are. There's lots more. I'll keep working on all this stuff, it's great to see it all come together :)
############ Apicolypse is a re-release of my old API-style color adder, exacly as it was.
Apicolypse works the same way Neverland did. It’s a drop-in replacement for old mixes, and was a precursor to BussColors. It’s got a simpler method of generating its dynamic impulses, making them a sort of continuous spectrum between the low-level and high-level sounds. Like Neverland, it’s 44.1K though it will still function at any rate you like (sort of pitched up): like Neverland, it’s got a hardness control that at 0 is the ‘Density’ algorithm, at 1.0 is straight digital clipping at the extreme, and at any setting between is a sort of hybrid that turned out to not be the greatest: a switch from perfectly clean, to soft-clip at any desired transition point. Technically if you had it so it only kicked in on hot peaks, it’d be hard to find fault with it: I don’t recommend setting it (on this or any Character plug or ‘Crystal’) to a position where soft-clip kicks in at very low level. Either do Hardness at zero, or high enough that most of the audio stays ‘un-distorted’: you can do what you like, though, I’m not the boss of you :)
############ AQuickVoiceClip softens headset mic recordings that have been super hard clipped on capture.
I hope this is handy for somebody: I made it for myself. The idea is, if you’re a youtuber or something and have the ability to post-process your mic feed, you might have it set up so normal levels give a normal sound, but then if some monster jumps out and KILLS you in your video game, you might scream very loudly into your mic and blow the recording into ridiculously loud clipping.
This does several things. Firstly, it clips and softens those moments further and tries to suppress some of the highs you got from the super-hard clipping. Secondly, it’s got a highpass which can subdue thumps and pops, and can also be used in conjunction with something like a pitch shifter to give cartoony voices (I’ve tried this with DiracFxAU, now known as zynaptiq: only used their demo AU, didn’t use their library in anything code-wise and won’t be doing so as they are unaffordable). I found highpassing before the pitch processing was very helpful for toon voices of that type.
This isn’t the seriousest of plugins, but I’ve seen people ask me for it, so now it’s free and Mac/PC VST too. I think it’s best confined to its intended use, processing single voice tracks that have wildly distorted moments. If you try to use it on the 2-buss I will be very cross with you ;)
############ AtmosphereBuss is Console5 processing with powerful new acoustic distance effects.
############ AtmosphereChannel is Console5 processing with powerful new acoustic distance effects.
I’m pleased with this one: it’s in the Console5 family (so its parts can be interchanged with any other Console5 plugins) but it represents a few different breakthroughs, and if they work out, I’ll be keeping them.
First, this is a multi-stage slew clipper. The effect’s a bit like PDConsole, a gluing and taming of bright digital highs: but, where PDConsole uses the algorithm of PurestDrive to get a more analog mixing-desk kind of sound, Atmosphere’s many stages of slew clipping (across fourteen samples, each with a different maximum slew) enforces the behavior of free air and acoustic distance. This has been a goal for quite a while, but I’ve never got results like these before (might be worth fitting a more exaggerated version into a ‘Distance 2’, if people like). It’s not calibrated to overwhelm, but the scale of your mix should be huge, and since it’s an extended-window slew clipping effect your quieter, subtler sounds don’t get muddied.
This type of processing steps on the annoying digital harshness with a heavy foot, but doesn’t directly EQ: the results you get will depend on what it would be like to have your sound blasting away through a perfect playback system at a distance of ten to thirty feet. If you need up-close, bright and loud sounds, you’ll need to use another system (such as PurestConsole), this one is for space and scale. It’ll probably work well for some electronic mixes if they’re meant to sound loud, it’ll give a real hugeness to rock or metal mixes (so long as the desired effect is more ‘live gig’ in scale) and it will really come into its own summing orchestral stuff from virtual instruments (or anything with similar scale/power needs).
############ Aura is a new kind of resonant lowpass EQ.
So this turned out to be a bear to bring up to date, because the original code was insane. I’m not quite sure how I got there, and I was still bugfixing long after the demo video was made. You’ll find the actual version of Aura has a slightly wider range, better adjustability in the low range, and the Dry/Wet control gets a touch of added functionality: as you go full wet, the resonant quality gets enhanced, so be sure and explore the half-wet or barely-wet settings. If it’s too scary-resonant, just pull it back a teeny bit and it should cooperate.
And this one is a bit scary as it seems to be channeling analog filters. I agree that it would be great to have this principle work as a full-range, synth-style filter that goes all the way into the bass. I can’t do it, though: it freaks out when I try, and it took endless hacking just to expand it a bit from what you see in the video. This is the algorithm derived from GrooveWear, which averages the rate of change OF the rate of change of the waveform. It’s not even slightly normal. You get what you get.
But what you get, is a lowpass with a striking resonant edge that’s implemented in a totally new way, and which has no pre-ring at all… and the way it gets its sound gives it an extraordinary sonority. Pretty much anything in audio that you’d want to project loudly as if from an acoustic space, can be given a sheen and glisten and sonority with Aura. I’ve got it extending down fairly low into the midrange if you’re at 44.1K or so: that should help if you need to use it at high sample rates, because the technique for doing it is not exactly cooperative and I found no way to simply tune it down: everything’s so geared to slew rate between samples that it’s best used for treble effects. I think it’s got a useful tonality for its treble manipulations, and I’ve spent a lot of time coming up with interesting ways to cut or enhance treble. This one’s good at what it does. You can really do stuff with the texture of your mix by aggressively using Aura on suitable elements.
############ AutoPan is for getting some organic stereo movement into your mix.
This is pretty simple: I got asked for an auto-pan. So I Airwindowsified it a bit, and here you go.
It’s set up to get some nice organic movement into your mix in various ways. Rate is how fast it goes (from nearly static, to a rapid flutter).
Phase means each channel relative to each other: on either side, you get full L-R and back again. At the center, it’s a mono tremolo (though, for stereo channels). Off-center, it’ll offset the sines in such a way that the sound sets up a swirling, circular stereo motion: swooping forward from the side, going across, and then back to near-silence again. Basically, centered is ‘not side motion’ and to the sides is ‘more side motion’.
Wide just cuts the mid channel. What this does for the auto-panning is, it’ll make stuff stretch out a bit beyond the edges of the speakers. If you’ve got something full to the side then you’ll get a bit of the opposite on the other speaker. Full wide is ‘only side channel’ and you’ll hear it in both because that’s what you get when you have only side channel: centered ‘mono’ sound, just out of phase completely. You’ll probably want to use smaller amounts of this, unless…
Dry/Wet is the final useful thing here. You can set up extreme stuff, like full Side or weird motions with Phase, and then just dial it back with Dry/Wet and it’ll give you a nice subtle activity of whatever you set up: don’t sleep on Dry/Wet here, you can get nice effects through using aggressive settings elsewhere and just sneaking in a little Dry/Wet. Especially if you’re doing extreme things with Wide, Dry/Wet is how you can integrate that into a sound so it’s just a nice little subtle motion that’s interesting. Or of course you can just crank it out, I’m not your mom :)
############ Average is a distinctive sort of lowpass filter.
This plugin is an exploration of a fascinating filter type that, I think, is really underrated. It’s a straight-up extremely simple averaging filter. If you set it to integer values, it exactly averages that number of taps on the input audio (adjacent samples). If you pick in-between values it interpolates, causing the ‘frequency’ of the filter to smoothly blend between the values.
There’s a reason you don’t see this filter used for lowpasses: it’s not technically correct. In fact it whacks a great big cancellation node into the high frequencies, and the tone (while pleasing) is very obviously affected by this. It’s a bit of a flangey quality.
BUT, we know better than to be limited by technical concepts, right? After all, the classic Scream Tracker resonant lowpass is known to be broken, and that has a real character to it. And even the Roland Supersaw is known to have some quirks that make it what it is…
Average has a fantastic tonality once you accept the response quirks. You can dial in the notches to suppress unwanted content, and (in a post-video revision) it’s got a dry-wet control so you can moderate the effect—and, typically for Airwindows, if this dry/wet is set to full wet, the calculations for doing that drop out of the plugin entirely so you’re not doing an unnecessary multiply.
############ AverMatrix turns averaging into a full-featured EQ.
My Average plugin is pretty handy. It gives you from one to ten taps of simple averaging, as a continuously variable thing. At higher settings, there are comb-filtery artifacts that come in as a result of how averaging works, but you can tune them with the filter setting and they sound pretty natural… and averaging has very good time response because it doesn’t have an IIR ‘tail’, so it’s very clean-sounding in ways normal filters aren’t.
AverMatrix takes that, and gives you one to ten POLES of filtering like in the first Average. That’s continuous too: you can have three and a half poles, no trouble (it just generates the third pole and then fades halfway into it) so the subtlety of adjustment is great.
AND, also by request, AverMatrix uses the Inv/Dry/Wet trick some of my other plugins use, and it’s a great idea to include. Go to full ‘inverse’ and AverMatrix is made to keep the dry and subtract the inverted wet from it… which means, now it’s a highpass. An amazing-sounding highpass, with great clarity and airy-ness and just as clean transient behavior, and just like applying dry/wet on the regular averaging, you can apply dry/inverse to give only a subtle treble lift. I think this is going to work real well for people. And one more thing: since native averaging at high numbers of taps gives a funny notch that you can tune, and since this is inverse, you can use this behavior to give a highs boost with a funny, tuneable peak. Set it wrongly and maybe it’ll make your mic sound nasty (don’t reinforce existing spikes in the mic’s response) but place it just right and you can get treble lift that neatly avoids the resonances of your mic… and enjoy the response of a much fancier microphone. (your settings will be sample-rate dependent so tune it again if you change sample rate)
############ Balanced lets you run an XLR input into a stereo input, sum it to mono, and use it.
Balanced is inspired by the adorably tiny Teenage Engineering TX-6 mixer. You might know that the TX-6 is notorious for being wildly overpriced as a small line mixer, since you can get a Bastl Dude for one-twelveth the price and have almost as many inputs (five jacks to plug in, rather than six). If you multiply that by twelve you've got sixty jacks, ten times as many inputs per dollar, and pretty much as tiny.
But the TX-6 is a digital mixer, not analog. And its inputs are stereo. And it also functions as a ADC, a converter for getting stuff into your DAW (at 48k, granted, but 32 bit audio). And that means it can serve other functions, like replacing other multichannel converters… and it's normal for multichannel converters to be sort of costly.
But how much can it replace? Enter Airwindows Balanced (and some patchcords you will have to make).
Firstly, with breakout cables (also very expensive, from Teenage Engineering) you can get twelve inputs. You will have to process them in pairs or leave them untouched: the TX-6 will do stuff ITB on input before sending the result to your DAW, but I don't think it can operate on sides of its stereo channels, so the processing has to match.
With Airwindows Balanced, you can plug a balanced, professional XLR output into your TX-6 (or ANY consumer grade stereo input of any kind) and turn it into a mono channel with all the benefits of balanced operation. In fact if your poor-quality consumer input hums or buzzes the same on both channels, Airwindows Balanced will cancel that out as if it was interference in the analog domain. It flips one of the sides and then sums to mono, with a handy gain trim based on BitShiftGain to let you gain-stage stuff that might come in very quiet.
The reason that's an interesting option is this: with Airwindows Balanced and Teenage Engineering TX-6, you can record certain things at professional quality in multichannel WITHOUT professional mic pres. And that makes TX-6 the budget option. Completely aside from its tininess and portability… or the fact that everything in this portable rig is internally battery powered and can exist, in any combination of professional mic input and stereo line input, miles away from mains power.
The key here is that you can get a battery-powered 48v phantom power box (I'm using Xvive P1s, which charge off a laptop or charger) and power a high-headroom and high-output condenser mic (I'm using Roswell Mini K47s), make a XLR to TRS 1/8" cable, and put the mic in front of LOUD things. The key is that the TX-6 (and some other bad quality line inputs on laptops and such) is not professional line level. It's consumer line level. There's no headroom, it maxes out way short of professional grade.
And that means when you put your mics in front of loud things (and I mean brutally loud drums, stage amps, etc) the peaks won't be limited by the phantom power box. They won't clip a mic pre inside the TX-6 as there isn't one. And it's a bit like trying to run your mic into a line input and getting no signal… BUT, with the right signals, you will actually get peaks within 6dB or so of clipping. Hell, if you clip the TX-6 it doesn't sound particularly bad, it handles clipping gracefully compared to some boutique ADCs: and it would take impossibly high loudness to do that off a mic.
The budget gear you can record with can't do that, not even close. In the 'loud into condenser mic into DAW' scenario, this setup with the Xvive P1 and TX-6 compares most closely with APIs and Lavrys, just without 96k capability, and everything on the budget side comes off as congested and flat by comparison.
My hope is that Airwindows Balanced also helps other recording situations. You should be able to take a hot XLR condenser mic, a 48V phantom power box, make the cable (these do not exist as far as I know) and a cheap old laptop with a 1/8" stereo line input, and get a respectable and good sound as a mono capture… also entirely on batteries, and anywhere. My hope is that if you're not buying tiny $1000 digital mixers on a whim you can still get value out of this.
But if you're able to get the TX-6 and scorned it, check out what I was able to make it do. I could have put up four other mics, all on things that would crush budget pres and converters, and carried the whole rig in a backpack including the laptop to anywhere I wanted to record, and also would never have to set a level because by definition nothing the mics can do can ever overload the inputs in a negative way. That makes setup quicker. Just saying.
I can give you Balanced. You gotta make the cables yourself, as I did :)
############ BassAmp is an old plugin with new tweaks, that gives some bass guitar tones.
This is BassAmp. (I’ve been asked to do amp sim type plugins, so it’s a start). Back in the day I experimented with some of these things, though I’ve always been more interested in wiring up physical hardware, especially for guitar distortions and things. I’ve brought this one up to date in some ways, debugging a problem with the ‘dub’ channel and using my modern techniques for dithering to the floating point bus, and so on.
It’s got basically a bright channel, midrange (which is just dry, as a slider), a bass channel with a characteristic woolly tone that has a kind of noisy blur on the small amount of treble it does have, and a ‘sub’ channel that really doesn’t work, so use it as a crazy special effect. BassKit might get you a cleaner sub-octave, but this will give you a trashy sub-octave that’s mostly garbage, even on a single-note line. So you’ll be leaving it off, I think.
############ BassDrive is an old secret weapon, like a kind of bass amp.
There was this plugin called BassDrive… and it was a secret weapon. It was unlike any Airwindows plugin because it ran on painfully hard-coded biquad filter code, looked up from calculators online: which also meant it was locked to 44.1K (at least, to get the expected behavior). But it was also unlike anyone else’s plugin of that nature, because it used lots of Airwindows overdrive inside it. The controls seemed to say normal EQ things like ‘bass, mid, treble’, but they didn’t give you those things, they gave you funny bands voiced in a weird way.
And this is because it was the eccentric Airwindows way of trying to mimic the SansAmp Bass DI. Not the pedal form… the rackmount RBI version.
The weird part is I ended up not liking that one (I came to prefer the pedal-form VT Bass) but before I got rid of it, I’d gone and shaped biquad filters to its voicings. But then, I implemented them strangely (note the intense notch at 10K, not present in the real one!) and made ’em overdrive a lot more. So it’s a lot LIKE a SansAmp RBI, but then again it’s totally not because it turned into its own thing.
And then, even though this could be used for bass guitars, I started to hear about how it was working as a secret weapon on heavy guitar sounds…
So this is exactly how the ‘secret weapon’ plugin was. It’s the blast from the past, now also available for Mac, Windows and Linux VST. And I know how eccentric it is, but sometimes that’s the point. I’m working on a bunch of stuff that’s meant to be normal (or at least flexible). This ain’t. It’s just weird. If you hate it, it’s not getting any better, abandon it with a clear conscience. On the other hand if its dark magic speaks to you… then hey, enjoy the new weapon. And rest assured that people who need good behavior out of their plugins will not be finding your secret, because BassDrive will scare them away first.
############ BassKit is centered bass reinforcement with subs fill. Clean and controllable.
This is much like DubSub or DubCenter, except it’s strictly mono bass and is designed to be super controllable. The controls are simplified and kind of optimized so only good-sounding results come out. You need to have good subwoofers (or use SubsOnly to test) to use the Sub output, otherwise you may not be able to hear what you’re doing as it’s much deeper and more filtered than you get with most DubSub patches (the filters are somewhat rearranged).
The Bass reinforcement works like if you were using the head bump in ToTape, except it’s mono-only so it will only reinforce usefully. Because BassKit is meant for mastering and 2-buss duties, and not the full range of madness available in DubSub/DubCenter, it uses the bass and sub augmentation ONLY as additions to dry: never ‘wet only’. You can exaggerate it, but the intention is to make it easy to add bass and subs in a sensible, controlled way. I hope it proves useful, and I know it will be more well-behaved than DubSub in case that one was too unmanageable for normal use :)
############ Baxandall is a killer general-purpose EQ.
Baxandall starts with a two-band filter that, zeroed out, subtracts an inverse lowpass from a lowpass and gives you bit-identical, perfectly transparent sound. That’s if you’re being subtle. If you boost or cut, lows or highs, it gives you the gentle broad boosts you expect, centered on the vital midrange. As you get more intense with the boosting, it gets more extreme, to where if you’re doing double boosts to get an intense exaggerated sound, a mids notch will naturally develop to accentuate the boosting. The whole voicing shifts to accomodate what you want to do with it, and you can play bass against treble or vice versa to get really wild voicings, such as for extreme EQ treatments (in terms of lows or highs)… but using the same natural Airwindows Baxandall tonality, so it won’t sound ‘filtery’, it’ll sound like it was meant to be that way.
Airwindows Baxandall uses my interleaved biquad filters (original, purest form, not meant for rapid automation) that run inside a Console5 instance to deepen the sound of the filter.
############ Baxandall2 is Baxandall, extended and made more powerful.
Sometimes the only way you can make something bigger is to strip it down. I'd been exploring EQs for the purposes of putting one in a future Console version, and revisited Baxandall… well, by the time I was done, a lot was different and a lot was still the same. It still uses my interleaved biquad filters (an unusual choice that helps the shallowness of the filter slopes in this very 'broad strokes' EQ). It still uses the technique of sweeping those filter center points to make increasingly extreme effects as you get crazier with it: it gets almost synth-like, starting with general 'tilt EQ' effects of great subtlety but bringing in wild boosts as you crank it. The purpose of that is to do the sort of anti-Soothe thing I favor: if you have a track and it's carrying highs, in no way do you need to also have full bass on it. Instead, you build a mix out of composite parts each of which bring a different voicing to the mix. And Baxandall2 is designed to get you to whatever broad EQ place you need, from a place of naturalness.
But there's nothing natural about the cranked-up 24dB boosts and cuts you can now do. This is newly tuned with the filter sweeping to take you straight past the original Baxandall's settings into wild new places. Also, the original Baxandall was made more forgiving by running it inside an internal Console processing, analog-ifying it. But it turns out that I can leave that to actual Console versions and strip Baxandall2 down to the guts of it, which actually gave it a lot more power.
This is Baxandall in the form that I will need it, for doing things with. But since it's Airwindows, you can use it too. And you can still use the original Baxandall, that's not going anywhere and has its own merits. Baxandall2 is just… more, at any cost, no holds barred. Hope you like it.
############ Beam is a wordlength reducer that tries to heighten sonority.
Beam’s a wordlength reducer like Dark, but instead of seeking to always give you the least departure from whatever trajectory the audio’s on (like Dark: and that suppresses highs, including noise in the highs), Beam seeks to make all waveforms converge on the same angle (either ascending, or descending). Imagine a world of triangle-waves, the deepest frequencies taking the most energy, and progressively quieter until high frequencies are in balance with the lows. That’s Beam. Beam has a ‘focus’ control that sort of optimizes the frequency range it’s providing a window into (as in, lower settings SEEM to encourage focus on deeper sounds, higher settings SEEMS to highlight the treble, and the 0.5 is set to zero in on where our hearing’s most sensitive.) And it uses those frequencies, across a broad range of possible sounds, to emphasize the energy and depth of the audio to shocking effect.
Except, it’s not. It doesn’t know what a frequency is. It’s only ‘dithering’ in such a way as to try and accentuate certain waveform slope angles, and has no real power to do even that beyond statistical averaging of a bunch of samples.
On top of that, its noise floor is WEIRD. Dark gates into silence, sort of well-behavedly. Normal dithers become noise seamlessly. Beam goes nuts and screams, and the only thing I can say for it is, raw truncation is worse. It’s super weird… and yet, even while it’s doing that, the background audio retains a startling depth and personality. Or at least so it seems… even more than Dark, with this one you turn up DeRez and nothing happens to the sound at all. You don’t even hear the noise until it’s silly loud. (there will be a series of dither reissues with DeRez, both for auditioning and for lo-fi duties.) It might even be a ‘sonic maximizer’, putting tone qualities into the audio that weren’t there to begin with.
I admit I don’t know what to make of this at all. I know how I got it, and I know it’s doing what I made it to do. But nobody knows what you get if you sculpt audio, not by frequencies or loudness, but by reinforcing certain waveform slope angles, because to my knowledge nobody has ever wanted to do that or had a way to make it happen, even in subtle ways like this. And now that I’ve done it, I’m not certain it’s better than Dark. I understand what Dark’s doing, and why that’s useful. This is a wilder beast. Looking forward to getting people’s reactions… because one of these may replace NJAD in my flagship Airwindows plugin, Monitoring. (followup: turns out it is Dark in Monitoring2)
############ BigAmp is a very flexible amplike thing with a taste for the bizarre.
Let’s go a little farther! BigAmp steps even farther away from what you might call ‘realistic’ amp sim behavior. It’s more like the Z series filters, or perhaps the X filters: you can make it act sort of ‘normal’, but it also lets you twist the knobs way beyond what’s realistic. So, you can use BigAmp for layering, or to supplement a sound, or just do strange things. There’s one more amp sim before I get into other, different plugins (and indeed different whole things like the sample instruments I’m developing, as heard in ‘Skronk’ and these demos) but BigAmp is perhaps the widest-ranging of the lot.
############ Biquad is the Airwindows implementation of a biquad filter.
This little puppy is really flexible!
That’s true of any biquad filter, but I find most of them are really flexible and slightly plastic. Traditionally biquads require special handling, because you can’t push them too far before they break, and also they sound ‘DAW-y’ and synthetic. Not in the same way as linear-phase EQs, mind you, but still with something about them that feels digital.
So I fixed that :) with a couple caveats. Firstly, part of it is brute force. I figured that if biquads run into trouble with 32-bit coefficients on some filter curves, and it’s widely known that you should do things like run ’em in series to get better behavior, we could always run long doubles and be totally sure they work as intended. Overkill? Meh. Sure, but it’s known that these break due to losing control of themselves due to not long enough wordlengths. Why wouldn’t they work and sound better with long enough wordlengths? And so they do. Also, you get to run simpler biquads if you don’t have to run lots of them, so it averages out. I think it’s just a matter of audio DSP coders stubbornly insisting on not using long data words on the grounds it doesn’t matter to the sound. This, when the filters go obviously wrong under some conditions for just that reason? Anyway, here’s long double biquad, because nobody else was doing it far as I know.
The second part is trickier, because if you use it you remove some functionality. Biquads make great DC blockers. But, biquads also work through the summing of many delay taps, all very close together. Some are on input, others on output. They’re a mixing of multiple copies of the signal.
That means Console applies to the biquad filter.
We already knew that, of course: I’ve long observed that you can put Console (or BussColors4/ConsoleBuss4) around a bog-standard DAW EQ and get an enhanced sound out of it. Mind you, I wasn’t controlling that EQ, but it still worked. But now, I have Console5 (not 6, but I COULD do 6 in the right context) built right into Biquad. So, it does expect to have a signal between -1 and 1 (or lower than 0dBFS in the DAW), and that means Biquad itself can’t remove huge DC offsets ten times the size of the audio content. (It can still remove smaller offsets just fine). And it runs into basically PurestConsoleChannel, does the biquad, then uses PurestConsoleBuss and goes out to an inverse/dry/wet control (which itself has multiple uses).
What’s a biquad, you ask? The Airwindows biquad uses four of the six common biquad options (I prefer to do shelves with subtractive/additive use of filtered audio). The top control has settings one through four. They go lowpass, highpass, bandpass and notch. That’s what I consider the platonic set of biquads, and it lets the filter design be simple and clear. You can set a frequency (in ‘amount of the audio range’) from ‘zero’ to ‘one’. There’ll be EQs with more specific frequencies, that’s easy, but this is proof of concept and to be used by ear, plus it will always cover DC to Nyquist at any sample rate. Zero is not really DC, because that kills the biquad: also, One is not really the Nyquist frequency (half the sample rate) for the same reason. But, they act a lot like they range that far, because the high resolution lets Biquad calculate things accurately even that far out of the normal range.
They don’t update/recalculate every sample, but the way I’ve defined the data structures means they could if you wanted them to (at a cost in higher CPU-eating). The code’s MIT licensed open source, so GPL people can just take it if they credit me, and all the projects that are using the Airwindows library are advised to get up to date and include this one. Set right, it is THE ultimate sample-instrument tone shaper to sit ‘under the hood’ and voice somebody’s musical product after the sample-playback stuff is taken care of. You’ll be seeing a lot of stuff come out that uses this code, as there are many plugins that require this type of filter to work, often ones that will benefit from the sonic improvements that are part of Biquad.
More will be revealed. Suffice to say this is a very useful building block that’ll allow for some very special plugins.
############ Biquad2 is the Airwindows biquad filter that's more sweepable and synthy.
This time, it’s time for the impossible!
As in, there was a reason nobody was doing sweepable, ‘synthy’ biquad filters…
Turns out it simply doesn’t work. One uses a different type, like state variable filters, for the synthy stuff. The reason is that, while biquads can sound pretty great (especially implemented like I do ’em), they fundamentally can’t cope with changing the filter coefficients mid-calculation. They flip out: you’ll hear some of that, especially at the frequency extremes. Low Q makes low frequency motions flip out, and high Q (and boy do I have a high Q for you this time) makes ultra-high frequency motions flip out in a really wild. glitchy way.
So obviously I gave up.
:D
nope! Instead, I just kind of forced the filter into zones where it mostly is controllable. This is partly through REALLY smoothing the filter cutoff, especially at low Q and low frequencies. If you try and update biquad coefficients every sample (and I already changed the form from the more CPU-efficient to the more stable form: didn’t help much at all) the algorithm gets super twitchy, so part of what I’ve done is just stabilizing things. I tried for ages to come up with some bizarre hack to force the filter back into stability: no dice. So, the range has been limited a little, the Q doesn’t adjust below a Q of 1, and it reacts really slowly, because that was the only way I could get it to behave at all. (the original Biquad can sometimes be better behaved, because it’s only recalculating coefficients every new buffer. So, zipper noise. When you get rid of that your troubles get infinitely worse, with biquads)
I’ve also got the resonance (on low and highpass) working differently. I’m scaling back loudness, but I’m also applying a distortion, then averaging after that, to try and get an ‘overdriving filter effect’ not present in the purer original Biquad. You can go quieter into Biquad2 and avoid this, or not use as much resonance: I feel it’s kind of like the Roland Alpha Juno filter resonance distortion, not an everyday thing but when it pops up it has its own distinct quality that’s interesting.
Anyway here’s Biquad2 ;)
############ BiquadDouble is a handy Airwindows cascaded biquad filter: steeper roll-off before resonance.
So it’s a boring old biquad.
Except, if you’re into the details of how stuff works, it’s not THAT boring. It’s two biquad filters stacked, which means you can get twice as much cutoff steepness before resonance. Since it’s Airwindows, you can have any possible resonance setting anyway, from ‘impossibly high’ to so wide that it’s basically full bandwidth. That interacts interestingly with the stacked filters: some of the weird effects with super-wide filters will act different here. On top of that, it’s two stacked filters inside an internal Console mixing system: so the tone of the filter itself is ‘expanded’ in a way not common to boring old biquad filters. And it’s the most efficient (but least cooperative) biquad, meaning that it runs and sounds great but doesn’t always cooperate with rapid automation, so that might get interesting if you’re not careful.
Why do this at all?
Because I sometimes like mocking up effects and plugins out of component parts. And if I was to use a stacked biquad filter as part of something else, it’s just more convenient to dial in the correct settings and get the tone exactly right: I could use two ‘Biquad’ instances, but remember that BiquadDouble stacks its filters INSIDE the Console processing, so it might not be as good to fake it, plus I’d have to set both the Biquad instances the same…
The real answer is ‘because that’s how I roll’. Nobody asked for this. But maybe you reach for a biquad filter for simple tone shaping, and you keep trying to find a butter zone between too shallow, and too resonant. This might become your go-to basic filter. I can’t predict what will take off: for all I know, this is THE basic digital filter everyone’s been waiting for, the one that just sounds right for every purpose. Or not. But you can’t know until you try :)
############ BiquadOneHalf is an interleaved biquad filter like Biquad.
BiquadOneHalf is tech that’s going into my future plugins. You have to set the frequency twice as high as you would using Biquad, and it can only go up to 1/4 Nyquist, but the way it interacts with the sampling frequency is special and works very well for some types of filtering. You’ll be getting more developed plugins of that nature soon, and all through January, but for now I got this: working version of the new Biquad filter variation, that you can use on stuff.
It’s the first version, not the ‘sweepable’ version, because this is going to find its way into fixed-frequency EQs, ‘head bump’ etc… compare it to Biquad remembering to use twice the frequency you otherwise would. It’s happiest on bass or midrange: much like regular Biquad, trying to force it to get real close to its high frequency limit can get messy.
############ BiquadPlus is Biquad plus zipper noise suppression! For twiddling the controls.
By request, a kind of bugfix!
Understand, a simpler implementation of a biquad filter isn’t a ‘bug’, exactly. I may have not had everything figured out, but if you’re designing fixed filters or looking to tune in EQs on something in your mix, the biquad filters I’ve been making are actually better. Without the extra smoothing code they run more efficiently and eat less CPU, and they’re still useful, plus there’s less to them so they’re more approachable.
But, a lot of musicians seem to enjoy cranking the filters around… and now that I’m using Bespoke all the time, I can put an LFO literally on any damn thing by rightclicking it. And it wasn’t all that difficult to do… I rolled it in to some documentation upgrades I did on the advice of Paul from Surge Synthesizer. One thing about hanging out with the open source music people is, stuff starts happening faster than you could possibly imagine. I had a crash bug in Bespoke on some strange things I needed to build into my main music making procedures, and Ryan had a fix the next MORNING. It’s daunting and wonderful to hang out with these earnest, motivated people.
You’ll see more on that, sooner than you think.
But right now… enjoy a cleaner, smoother Biquad, plus zipper noise suppression! I needed to come up with my Airwindows way to accomplish this, and got it done. More to come.
############ BiquadTriple is a handy Airwindows cascaded biquad filter: steeper roll-off before resonance.
Here’s another utility plugin! This is the same as BiquadDouble, except it’s three stacked biquads. Otherwise the same: you get to set the frequency, Q etc. for all the filters in parallel. This saves time over doing three instances of Biquad, plus they’re run in series inside Console inside the plugin, so there will be a deeper tone out of this one.
It’s for if you wanted to get steeper filter slopes without the cutoff going resonant. This, like BiquadDouble, is very much ‘tools to make tools’, for if you’re mocking up some kind of effect and know you’re going to be using filters. You can test out what you’ll use in your effect, this way. Note: seems like the inv/dry/wet is really just wet/dry/wet, and I have too many new and more worthwhile plugins to put out to fuss over it so I’d advise just taking BiquadTriple as it is. I’d be using it as full-wet anyway though you can bleed in dry for more of a shelf effect if you’re doing a steep rolloff of some kind.
I expect to be doing some more variations on this as there are known ways to stack filters where the Q factors are very specific and staggered: however, those will be more aimed at the mixer/end user, and probably will be called isolators (that being a common term for this type of steep crossover-like filter). And again, I’ve got other things to put out that are more interesting, so more will be revealed as I do those experiments. I think an Airwindows isolator-style effect would be good. Bear in mind the biquads can be fussy about extreme settings and moving the sliders abruptly.
############ Bite is an unusual edge-maker.
Bite puts on a sort of midrange edge. At high sample rates it’ll be more of a trebly edge. It’s no specific frequency, so much as it’s just a harshening factor: you can also use it inversely, to take out midrange edge. It runs a couple samples of latency: on VST that works as a couple samples of delay. Again: an experiment, a science project.
############ BitGlitter is an old-sampler style digital lo-fi plugin.
DeRez is the simpler, purer bit and frequency crusher, and is still the best ‘analog setting’ bitcrusher (because it lets you use floating-point or fractional frequency and bit crushes). It’s a very pure example of those things and you can make it gate with a touch of DC offset from DC Voltage, and it’s even got a touch of grit softening when it frequency crushes to improve its tone.
BitGlitter, however, isn’t DeRez. BitGlitter’s something a lot more sophisticated: a kind of sampler emulator. At every stage it’s designed not for bitcrush alone, but to get the particular tonalities you can get out of primitive old samplers. An earlier attempt intentionally went after the old Akai sound, but currently BitGlitter has no specific model. It’s just there to dial in a kind of punchy grit that will make beats sit well against other elements: the video demonstates this.
Now, I know there are people who get mad when I make plugins like these. They say, ‘stop making the sound worse!’ and I understand what they mean, but sorry, I won’t stop because I know there are elements to certain ‘bad’ sounds that aren’t just ‘bad’ but usefully different. BitGlitter maximizes this as much as I can, and might be the go-to textural element for this sort of thing if generic bitcrushing etc. just never works for you. And then for some people I think it’ll immediately be their best friend, but I don’t need to explain to that crew what this is. For those who aren’t used to ‘crappy old sampler’ magic…
First, BitGlitter’s got gain trim going into a stage of Spiral analog-style saturation. You can overdrive the input effectively. Then, it does a hint of bitcrushing and splits into two separate frequency crushers, each set slightly different. This isn’t ‘accurate’ to any real retro sampler, but it helps broaden the sound. The output of these are blended and given an output gain and a dry/wet in case you need to sneak a little clarity back in there, and a slight averaging blur is added to the blend to further emulate analog circuitry.
The result is a coarser, more opaque sound which still lacks modern digital ‘edge’: you can plainly see on a metering plugin like Voxengo SPAN how the highs are softened. It’s not a digital bright-maker, it’s a texture-changer and impact-maker. Especially if you go for darker regions of the Bit Glitter control, you can use this to add ridiculous amounts of midrange punch in that ‘retro hip-hop’ kind of way. There’s a visceralness and aliveness to the grunge because it’s made by an algorithm to act like analog gear might: you won’t get the same result out of just a pile of typical DAW bitcrush and EQ. BitGlitter will do the extreme damage you might be looking for, but it’ll do it with a personality that contributes instead of detracts.
############ BitShiftGain is the ‘One Weird Trick’ perfect boost/pad, but in 6db increments ONLY.
Here is one final trick for clean gain aficionados.
Turns out the only way to get cleaner gain trim than PurestGain, with its high mathematical precision and noise shaping… is not to do any of that. No fancy math, no noise shaping or dither. Just a very narrowly defined boost or cut, in the form of a ‘bit shift’.
Doing this means your waveform is scaled up or down by increments of 6 dB exactly. No 3 db, no 9, no 7 or even 6.001! Only 6 or 12 or 18 and so on, up or down. Select the number of bits you want to shift, and BitShiftGain applies the exact number, not even calculating it in floating-point through repeated operations: from a look-up table to make sure it’s absolutely exact and precise.
And when it does, all the bits shift neatly to the side inside your audio, and whether you lose the smallest and subtlest or gain up and fill it in with a zero… every single sample in your audio is in exactly, EXACTLY the same relative position to the others. Apart from the gain or loss of the smallest bit, there is literally no change to the audio at all: if there was a noise shaping, it would have nothing to work with.
Perfection, at exclusively increments of 6 dB. That’s the catch. You probably can’t mix with gain changes that coarse (though it’s tempting to try!) but here’s what you can do: you can take 24-bit dithers, gain down 8 bits in front and 8 bits up after, and have a perfect 16 bit dither. Or a 17 bit, if that pleases you… or shift 16 bits down so you can hear what your dither’s noise floor acts like (we’ll be doing lots of that when I start bringing out the dithers). +-16 bits of gain trim is a very big boost or cut. The overall range of BitShiftGain is huge. But the real magic of BitShiftGain is the sheer simplicity of the concept. Provided your math is truly, rigorously accurate and your implementation’s perfect, gain trim with bit shift is the only way in digital (fixed OR floating point) where you can apply a change, and the word length of your audio doesn’t have to expand, AND every sample which remains in your audio continues to be in exactly the same relation to all the others.
Digital audio is like some crystalline structure: it’s fragile, brittle, and suffers tiny fractures at the tiniest alterations. There’s almost nothing you can do in digital audio that’s not going to cause some damage. But as long as you stick to 6 dB steps and rigidly control the implementation (BitShiftGain doesn’t even store the audio in a temporary variable!), you can chip away at that least significant bit, and the whole minutes-or-hours-long crystalline structure of digital bits can remain perfectly intact above it.
############ BitShiftPan gives you a gain and a pan control that are ONLY done using bit shifts.
This is a request from YouTube comments, but I had no idea how well it'd work out! BitShiftGain is a long-standing secret weapon of mine. On almost every video, I'm losslessly dropping 6dB using BitShiftGain. But what would you get if you applied this to pan?
You'd get a pan where center was quite a bit louder than sides (there's no 3 dB pan law from bit shifts), but the first steps to left and right are QUITE a lot to the side. It's not at all LCR panning, but if something's center you know it, and if something's to the side it's WAY to the side. Then, you have a succession of further-to-the-side positions that are progressively quieter, all the way to hard L and R. If that was all it was, this would seem really pointless and arbitrary.
BUT.
If you can construct a mix this way, you can construct a mix where every single gain setting, every pan position, every location in the mix, is Bit Shift Gain: utterly and completely lossless. No requantization, just like with BitShiftGain itself, but in full stereo (within these constraints). You're picking locations, but they're not LCR locations, they're a range of potential locations.
There's more. Mixing with BitShiftGain in mono is impossibly crude. 6dB increments are seemingly impossible to mix with, absurd, insulting to even consider. But if you tick a track one step over to the side… that's now 3dB down, not 6. You've losslessly cut one side 6dB while leaving the other one unaltered.
If you had two tracks, a doubletrack, and did this with just one of them, your 'track' (that's really two tracks) can shift 1.5 dB down, to the ear. Starting to sound more interesting? If you had four tracks and shifted just one of them, that's an apparent shift of 0.75 dB.
Mixing isn't just about taking a single track and making it perfect. Mixing is how tracks sit relative to other tracks. If you have a full mix, and a track 40 dB down steps closer to the center bringing it up 3 dB in total, that makes all the other tracks seem just a tiny bit quieter in contrast. At every step, your ability to fiddle with 0.1dB adjustments is completely hobbled… but the framing of the TOTAL mix can take a whole new form.
And at every point, across every track, at every position in the stereo field, the mantissa of every audio sample is EXACTLY as captured by the converter. Once it's mixed, you'll get a composite, but everything being fed to the mix buss at every level in every position is exactly the raw sample… scaled to fit.
If you liked BitShiftGain for its utterly uncompromising transparency, beyond anything else even possible… now you have it, but with panning.
If this approach, so ruthless in the desire to hang on to raw unprocessed intensity from the original digital captures, seems interesting… next week is Console Zero, built from the ground up to work using almost entirely bit shifts even inside the saturation/antisaturation calculations and anti-alias filtering. BitShiftPan is an ultimately clean gain trim, and apart from the 'steppy' positioning and lack of pan law, it's very normal and approachable. Console Zero… is radical.
Thanks for the suggestion to try applying this to panning. Who knew so much would happen as a result?
############ BlockParty is like a moderately saturated analog limiter.
What do you get when you cross an Airwindows compressor, with OneCornerClip?
Pretty much this. Okay, so it wasn’t a very difficult riddle, was it?
BlockParty acts like a somewhat distorty limiter. It’s not at all about lookahead (in fact it doesn’t have any) or preserving tones pristinely. Instead, it takes the onset of sounds that would compress, and manipulates the attack in the way that OneCornerClip does. The threshold gets kicked way down, and gradually expands to full scale, and since the threshold’s determining compression, that means BlockParty doesn’t have a stable compression threshold. It’s interactive with the audio you’re giving it.
Because it’s on the OneCornerClip model, that means it’s a mostly-compressor with OneCornerClip-like behaviors. That means bass which blooms and has fullness even under heavy load, and highs that don’t poke out or distract. The result is a thing that sounds real analog-y but not super clean. You can use very small amounts of it (there’s a lot of gain on tap) to do peak limiting for loudness maximizing, or you can slam things into it for effect. It’s called BlockParty because heavily limited stuff sounds like blocks of loudness: it’ll get you some of those sounds, but not as cleanly as your classic ‘loudness war’ limiters. It’ll also smash drums and things in its own distinctive way, which might be its strongest suit. On the end of it is a clipping stage to make sure nothing you do will ever produce overs. The clipping stage is AFTER the dry/wet, so to get a true dry you’ve got to turn it off: this is because raw digital clipping is another style of loudenating, so if you were going super-hot into BlockParty and wanted to dial in some pure digital clipping you could use the dry/wet to do it (or, if including some dry would have given you overs because your direct buss signal includes overs).
BlockParty is a fierce loudenator with a voice and style all its own, using techniques that are distinctly Airwindows. It might be just what you needed, or it might be a little too grungy for you… but either way, there’s nothing quite like it, so check it out.
############ BrassRider is One Weird Trick for drum overheads in metal!
So this is pretty simple, and pretty distinctive. I made it for my friend Tim (one day I’ll repost his amazing rant, on the brink of his possible death, to musicians about ‘stop buying DAWs! Stop buying mic pres and plugins and becoming sound engineers! Learn to play your instrument!’) and it’s possible it has only one user, Tim. If so, it’s okay. This is a weird trick, a very specialized little toy. I do think it’s good at its strange purpose, though.
BrassRider is a drum overhead rider for metal mixers.
Here’s the situation. If you are doing a metal mix of extreme brutality, you are probably making very intense, sculpted spot-mic sounds. You probably have a mic on every tom, every drum, perhaps also on key cymbals. And you probably have overheads, because hey, overheads. Drums use overheads, right?
And your overheads probably aren’t helping you AT ALL. You’re getting ‘jazz drummer’, ‘classic rock’ snare and tom sounds poking through from the overheads, and you hate them. Yet, there are some cymbal crashes that should get into the mix. What to do?
BrassRider does one thing (one weird trick!) to help you. It watches for crashes (noise like white noise) and it cranks the volume WAY up when that goes beyond a threshold. You turn up the threshold control to start engaging this behavior. There’s a dry wet control that you might not even use if your mixes are truly brutal and heavy (who needs reality?)
So then, BrassRider is cranking your overheads when the crashes are hit. So what you do is you bury the overhead mics in the mix, completely. And you use BrassRider to make them peek out only when there has to be a decent hint of crash cymbal in there. And most of the time, BrassRider totally kills the overheads so your drum sounds have maximum sculpted brutality and work the way you want them.
I don’t sit around making metal mixes so I can’t demo this properly, and I made it so I’m not that concerned with doing the ultimate demo for people who don’t know what they’re doing and have to be sold on the tool. Tim would be disgusted with me if I went around trying to popularize this one and teach noobs how to use it. So, if all that confused you, this is not the plugin for you. Also, I’m not going to teach you how to use it in a metal mix (except maybe on Monday livestreams. ;) ) And if you know exactly what I’m talking about and you’re already dialing it in, a wild mad light in your eyes, well… you’re welcome :)
############ BrightAmbience is a plugin for very bright gated reverbs.
BrightAmbience is one of the old secret weapons of classic Airwindows. It’s nasty enough to be pretty unique, and nice enough to be useable. What it does, is it sets up a BIG pile of delay taps spaced by prime numbers. There’s a decay control and a sustain control: the sustain will help with its CPU hungriness, the decay won’t. Dry/wet is at the top (what can I say, this was an early one) and wet tends to run HOT so you might like to use it as mostly dry and add its coloration to things as needed.
It’s CPU hungry because it’s using a really naive method of doing all that, but it also has a really distinct sound that differs from what you can get out of convolution reverbs. And since the distribution of delay taps is prime-numbered but also kind of naive, instead of getting a convincing room you get a sort of steel-chamber effect with a nasty resonance up near the Nyquist frequency. But that can work as a special effect. Back in the day I meant this to be a great reverb, but it’s not. However, it IS a great ‘crazy bright ambience effect’, though it’s pretty CPU hungry like I said. You can apply it to anything that you’d like a halo of shiny around, and its tone is nicer at 96K as I demo it. (the sustain is tied to maximum sample delay, so higher sample rates will give you shorter sustains and denser shininess)
############ BrightAmbience2 is more BrightAmbience with better tone and more slapbacky effects.
This is just like BrightAmbience, except different in pretty much every detail. Techwise (skip to next paragraph if you like) it is using a totally different prime number series, ‘super-primes’ for its delay spacing, and then it’s offsetting alternately left and right delay taps to the next prime number in line, meaning that it’s a dual mono ‘ambience’ but center signals WILL get stereo spread. Also, there’s a kind of regeneration that was real tricky to do, but it means you can get a clean digital slapback, or any degree of ‘fuzzed out’ bright ambient slapback all the way to the wash of sparkly atmosphere, either subtly regenerated or cleanly gated.
Okay, but what does it do, and what does it sound like?
Super ultra bright ambience space that doesn’t get in the way. You can put it on things like drums for 80s gated verb, you can do dub-like things through using it to make a blurry slapback, but this thing is bonkers at putting Star Quality Vocal Glitter on voices. You don’t have to have it loud, and you don’t have to stretch it out so long that it feels like a reverb. That’s not the point, this is about doing that classic Lexicon thing (without, I might add, using ANY actual Lexicon sounds or algorithms) where you can fill in a bright, glossy atmosphere around the voice that makes it sound like star quality. You probably don’t want to treat it like a reverb, on a send or whatever (maybe on a vocal bus? It’ll handle split harmony vocals very elegantly since it’s dual-mono). Instead you want to use it like your lead vocals alone go into a special chamber. Might also be an inspiring thing to monitor while tracking: I sure had fun playing with it in my headphones, and if it’s on your mix while tracking you might not have to print it on the actual vocal track. Everything about it is evolved from BrightAmbience, even the algorithm that makes it.
Oh, also if you’re a coder and want to get your hands on the delay taps, it took hours of looking up and typing in specific prime numbers, as there is no such thing as a ‘list of super-primes except every other one uses the next real prime after the super-prime, making the list pan every little echo to alternate sides using inter-aural delays’. And maybe you never even thought of such a thing. But if you think you can make use of such a thing, in the .h file (for the AU, anyway) is a definition of ‘primeL[]’ and ‘primeR[]’ that you can simply copy and paste. It’s 489 total entries which will get you a half-second or more out from the dry signal, even at 96k, and you just use ‘primeL[]’ and ‘primeR[]’ to specify the delay taps you want, typically in a range (like, entries 40 to 60 will give you a little ambient blur starting at whatever ‘primeL[40]’ is, which is 1031 samples)
It’s MIT license so you only have to shout me out and you can do anything you want with it… so don’t say I never gave you nothin’ :D
############ BrightAmbience3 adds undersampling for high sample rates, and better feedback.
BrightAmbience is a very old plugin. The original was all about taking sounds coming in, and transforming them into lengths of extruded and very bright reverb. BrightAmbience2 transformed that, in turn, into a more adaptable creation that used inter-aural delays to create a subtle stereo effect like an aura around mono content.
BrightAmbience3 adds undersampling. Now high sample rate mixes retain a consistent tone and reverb length to what the CD-rate plugin would do… and it’s more CPU-efficient running at the elevated rates… and the subtle darkening in tone makes it worth a re-listen.
But now, on top of all that, we’ve got a new way to apply feedback at the ‘wider’ reverb settings, which allows you to feed THOSE back too. And that means, BrightAmbience3 has just taken on a new life for a variety of vibey, distinctly analog-feeling blurred delay effects. Even the really wide reverb settings will still feed back at full crank (though they just give you a sort of droney resonant quality) and the medium settings produce a variety of unusual sounds that are a little bit like when you have a crummy old antique echo effect, and it has no clarity, but when you turn up the feedback strong retro flavors begin to take over… you can’t get clean infinite regeneration that way, but tune it to taste and dial back the feedback control until you have enough echo for your purposes.
Or, ignore the feedback and just use it as BrightAmbience, but with a greater range of effect at higher sample rates, and a richer tone thanks to the undersampling.
############ BuildATPDF is a dither-making toolkit.
I've explained how DoublePaul was made: taking a TPDF highpass dither, and including additional nearby samples to cause the tone of the noise to be more high-pitched.
Wouldn’t be fair if I didn’t give you a chance to play with it yourself :)
To use BuildATPDF the way I used it, set the middle sliders to -1 and +1. Then adjust the other sliders until you’ve tailored the sound of the noise in a way you like. With the other sliders at 0, you have PaulDither. If the surrounding sliders ‘oscillate’ (going minus, plus, minus, plus) you have more of what DoublePaul is. Tailor the highpass in ways gentle or obvious, or even darken the highpass while leaving its dither functionality intact! It’s the combination of a -1 and +1 tap that gives you highpassed TPDF. (This won’t work with +1 and +1, because it’s the same random noise passing through the plugin: it needs to play a positive node against a negative node)
Or, you can set up the -1 and +1 taps, and then just play with the sliders to see what it sounds like!
############ BussColors4 is the Airwindows console emulations, now working up to 192K correctly.
This plugin’s a pretty huge deal: it’s been around since before console emulations were a thing, and I’ve been steadily working on it year after year. There’s not been much to do: my original algorithms, once I moved past the earliest incarnation as the ‘Character’ plugins, were so solid that it left little to do. BussColors has sold $22,445 of plugins over the years, as Audio Units alone, in spite of the fact that it has always been a 44.1K plug (due to being based on a convolution model).
The story there is, a guy at NoiseVault always wanted someone to make a plugin which used his sets of hardware convolution kernels, which he’d done at all different levels. Sort of the Nebula thing (Giancarlo has been a great help when I was getting the VST thing going, check out Nebula)
I’m not that. I’m Chris from Airwindows… so I set out to make my own version of this sort of thing, with hardcoded algorithms generated on-the-fly, as raw and close to the metal as I could get it. And I got some pretty respectable CPU efficiency with all of the tone quality I wanted, limited to 44.1K.
UNTIL NOW :)
Try it. BussColors4 is the newest version of BussColors, and it works at any sample rate and gets you the same sound. Better, in fact: I like what it does a lot better at 96K and 192K. It also doesn’t greatly expand processing and CPU hit at higher sample rates. I think this one’s a real highlight of the Airwindows arsenal. It would definitely be worth your $50, like its predecessors (well, I always did give existing BussColors users free updates for life, I mean for new buyers). Feel free to shoot it out against the most expensive and/or CPU-hungry options and I think it will clobber all comers.
The sources of the models in BussColors are as follows:
Dark (originally Cider) is a Focusrite and resembles an MCI console a bit
Rock (also seen in the Logical compressor) is from SSL impulses, obviously it's SSL in nature
Lush (originally Neverland) is from Neve impulses
Vibe (originally in the Elation compressor) is from LA-2A impulses
Holo (originally in Precious) is from Precision8 impulses
Punch (originally in APIcolypse) is from API impulses
Steel (originally in Calibre) is something else, I don't remember what
Tube (originally Luxor) is from Manley Massive Passive impulses
All of these run a form of dynamic convolution which means they are never EXACTLY what the source impulses would have them be: it's more complicated than that. So, treat them as original things, not as attempts to clone whatever original gear sourced the impulses. In the end, they're BussColors(4).
############ ButterComp is my softest, smoothest compressor.
Sometimes a plugin can be a sort of cult favorite. That’s the story of ButterComp, a compressor of great subtlety that’s no use for quite a few normal compressor tasks… but still has a fervent following. I’ve been begged for the new version of this one (not merely VST, but revamped with all the current Airwindows sophistication and purity) and I’m delighted to bring it, though I think it might puzzle some people. If the stock Logic comp would do as well, this isn’t the plugin for you. But if you’re looking for a certain thing and thought it unattainable in software, this might be your lucky day.
Buttercomp, under the hood, is absolutely unique (or unique, until I start working on variations and until other people try to copy it. They’re free to, but it’ll be possible to test that quite easily with special audio files).
It’s a bipolar, interleaved compressor (with rather slow attack and release: a version with access to much faster dynamics is available in CStrip). This one’s the original, the cult classic. What is a bipolar, or an interleaved compressor? For the purposes of Buttercomp, it’s four totally independent compressors per channel. Two are sensitive to positive swing, and flip back and forth every sample. Two are sensitive to negative swing, also flipping back and forth every sample. The compression factor’s reconstructed through combining these poles, through the screen of the interleaved compressors switching back and forth at the Nyquist frequency… that mysterious digital frequency that is on the one hand the literal highest frequency that can be encoded, and the lowest frequency that ought to be totally rejected and filtered out.
What happens is this: the tone of things gets some added second harmonic, wherever the compression is more strong on one side than the other. High frequencies take on a particular airy openness, since they too get second harmonic, plus individual sample spikes can only affect one out of four compressors: ButterComp deals with all samples only as samples relative to other samples, and doesn’t get thrown off by isolated samples that don’t represent the actual waveform. It’s got a sound, but the sound is hard to define because of its extreme fluidity and purity.
If this sounds like your idea of fun, enjoy ButterComp. If you’re looking for the ‘all buttons in’ mode, I’ll get back to the drawing board and probably do something totally different for you. This compressor is not for everybody. Also, if you start slamming it really hard (demonstrated in the video), it’ll volume invert: you can squish it down to become more quiet than the quiet passages, but still it will retain its tone quality. It’s perhaps best used as a particular kind of ‘glue’ compressor, at which it is exceptional… or maybe I should say, it’s unique. Start working with it and you’ll soon work out whether you’re part of the ButterComp cult. Not everybody will be, and it’s only one type of sound… but nothing else can get that sound, and this is why I’ve kept this purest form of the algorithm available.
############ ButterComp2 is improved ButterComp with an output control and sound upgrades.
So here’s what happened: in working on the new ButterComp, I found a mistake. Because of a thing C programming lets you do (assign, in an if statement) it turned out the original ButterComp didn’t actually use the interleaved compressors after all. The one in CStrip does, but actual ButterComp (which has its own distinct fans!) doesn’t. It’s strictly a bi-polar compressor: it does each half of the wave different, and blends them.
Because of this, I’ve made the source code (also being released) represent what the plugin actually does in practice. It’s a little simplified, and it’s worth paying attention to, for people who like the simplest most minimal form of ButterComp.
But, because of this, I get to release ButterComp2 as very much its own thing! I even came up with a subtle tweak: it modifies its release just a touch, slowing it when the signal’s hot. That’s on a sample-by-sample basis… and it’s on the OUTPUT of the compressor. So, this further smoothing effect is subject to the output level control. And the dry/wet. In fact if you had it all dry, the release modification is therefore as if you had it on the input… making it blend not only between positive and negative wave compression, but also between feedback and feed-forward release time modifications :)
But really what you need to do is listen to it.
With the interleaved compressors fully working AND the bi-polar compression on each, there is indeed the four distinct compressors working in parallel. The whole thing is very gentle (hence the name) but you’ll get a glue and tonal reshaping out of it as it will even out the bulk of the waveform, making it balanced between positive and negative. It’ll also soak up treble detail in a characteristic way, and you’ll really hear the quality of ButterComp2 on ambiences and reverb tails. It’ll float things in space in this holographic way… I thought it made for a significant tonal improvement over the simpler ButterComp.
############ C5RawBuss is the original Console5 algorithm, with optional very gentle DC suppression.
############ C5RawConsole is the original Console5 algorithm, with optional very gentle DC suppression.
Back in December of 2017, I launched Console5. It had some innovations I’d come up with that got it to a height of sonic deliciousness I’d not reached before. I put it out Xmas morning… and immediately panicked and scrambled, because I’d been testing it with a sort of ‘choir’ and hadn’t thrown enough softsynths, etc. at it. And when you hit it with a intensely asymmetrical slew (how fast the waveform moves) it flipped out and began producing powerful DC voltages.
Ever since, I’ve been doing variations on Console5 (which all interoperate with each other) that avoid the problem. That same day, I rushed out a bugfix. I tried an alternate method for getting that tone, which wasn’t as good but was more resilient to DC offset. I came up with a good workaround for Console5, which watches for silences as a good place to pull back on DC, and set up the final Console5 that way. I released PurestConsole to offer a cleaner, clearer version that totally avoids the problem in the first place. I released PDConsole, because if you can’t have the original Console5 tone algorithm, why not have PurestConsole crossed with PurestDrive (which doesn’t have the DC issue)?
All the time, there were SOME sorts of audio that were fine… or mostly fine… with the original, special-toned Console5. Technically, nothing’s perfect: if you went for long enough under the original Console5, the DC would inevitably drift, unless the tones were really perfectly symmetrical. But it was a lost opportunity: the DC bug was so severe that the original version was forever lost, banished for endangering DC-coupled equipment and never to be heard again.
UNTIL NOW :D
Don’t use this Console5 version if you’re not willing to work with it and keep an eye on it. Much like the plugin DC Voltage, which simply turns the output into a fixed DC offset for signal-processing or testing purposes, C5RawConsole is capable of giving you signals that are bad for your gear. The final Console5 is very nice and automatically handles DC problems, PurestConsole can’t possibly give you DC issues, neither can PDConsole, and you can mix and match these if you like. There’s plenty of Console5 stuff out there if you just want to get some sounds and go.
But, that original algorithm captured people’s hearts. The way it handled the fabric of sound was something a little special… so now, that original Console5 lives again. The Xmas morning one, the ‘wow factor’, dangerous one. With one tiny detail added, that can make it tame if you know how to handle it.
The control ‘Centering Force’ defaults to 0.0, which is the ‘Xmas Morning’ setting. If you turn it up to maybe 0.7 or 0.8, you have the first bugfix, where the correcting force took away some bass just to make sure the problem was fixed. Cranked up, it’s a little bit like a highpass and you can use it to tighten non-bass tracks, seeing as it’s there… it’s like a free highpass, but very gentle slope. But if you keep it at 0.5 or below, Centering Force reacts so very slowly that it won’t touch your bass. I don’t think it’s really perceptible at 0.5, but you can always be even more cautious and go to 0.25 or so. At settings close to 0.0, it will take minutes to erode away any DC offset that sneaks in. And at 0.0 it’s completely defeated and there’s no centering force at all.
This is for the people who wanted that very first version back. You can couple the plugins with other, outboard DC correction filters if you like, or you can use the Centering Force (especially if you see voltages lingering anywhere in your mix, that tells you that you need to apply some). You can treat it like there’s a built-in highpass as part of the circuitry of the Console, and tune the dominance of your subs very conveniently without having to run an extra highpass (much less a multi-pole, steep-cutoff highpass). I guarantee C5RawConsole can be used for this, because it dials between the extremes nicely, from noticeable subs reduction at 1.0 to absolutely no reduction at 0.0, with a really really extreme exaggeratedly logarithmic curve making it so the middle setting is still pretty much no bass rolloff.
I think this should give people a chance to play. I could’ve stuck with the final Console5, in practical terms I think that’s the one most people should use. But Airwindows people like pushing the limits, so I needed to make that possible. Now, if you use C5RawConsole and leave the centering force off, you have that untame, primitive version that came out first and was gone in mere hours in a frenzy of bugfixing. It’s what I made when testing obsessively against the ‘choir’ sounds, trying to get something special while still mourning the loss of my Mom, who sang in a choir. That’s where that first Console5 came from, and it pleases me to (with caution) bring back that original tone, and let people have access to it.
Just be careful how you apply it, and if you see voltages hanging around (on tracks or the buss), put some of the Centering Force in! We could all use some of that :)
############ Cabs is an Airwindows guitar speaker cabinet simulator.
I’ve been asked for ITB guitar plugins before. Never been all that interested, because I like real amps and real tubes etc: have never found a lot of enthusiasm for going to great lengths to digitally fake all the colorations of some amp/speaker or other, because when the soul’s not there the color is irrelevant.
This is different. I took on the challenge of updating my old Cabs plugin: originally dynamically convolved cab impulses, except they hadn’t been MADE with that in mind so it became a series of wild experiments. I wasn’t happy… then.
Cabs completely overhauled everything I’d attempted. It’s undersampled, so it’ll run on whatever sample rate you like, efficiently. It’s got the same tricks of allowing you to heighten the particular speaker coloration it uses, or dial it right back to nothing.
But around that core is a completely different approach to getting ITB amp tones. I’m using the cab size and ‘room loud’ controls to let you dial in the exact presentation you want, from clear and up front power-soaked clearness to complete wall-of-stacks meltdown, and beyond (in classic Airwindows fashion). This is done through an alternate approach to loudness emulation based on the destruction of audio signals through ultimate volume, such as a Space Shuttle takeoff, and what happens to the waveforms there.
I think you should listen to it. Careful that you don’t jump for odd settings right away and get confused: for instance, settings of Room Loud beyond say 0.6 are not really real-world things. Settings of the cab tone intensity beyond 0.6 or 0.7 might be a bit weird. The way the cabs take on character and intensity is related to cab size: everything is basically related to everything else. You have to know what sound you’re trying to get, to clone something real-world. This is not a preset box, at all, at all. You have to be a real guitar tone maven to dial this in to convincing amp clone territory.
OR… or… or!
You can use the range of adjustment here, with the controls that give you a basic idea of ‘what you’re doing and how you want it’, combined with for instance Edge or the upcoming Dirt plugin (softer saturation) to get tones that DO NOT EXIST but act like real amps in significant ways. You can get tones that you can’t get in the real world, and dial them in to match what’s in your head: in some cases maybe to a point that the real world amps couldn’t reach.
And track through them because the whole rig (when using entirely Airwindows plugins) runs NO LATENCY, so it’ll feel as close as it can to the real thing, and respond with electric immediacy.
And then unlike any real amp situation… once you’ve got the final mix done, you can wiggle any of these parameters to where they need to be for THE SONG and tune the tones to best suit whatever else is there, even if it didn’t exist when you were tracking it. This is a freedom of the DAW experience (don’t go too overboard, make decisions!) and it might bring you the results you need.
And all of this can be lightly applied to any other tone: real genuine amp-and-cab that was tracked too quietly or didn’t pick up enough mojo, synths, drums, you name it!
############ Calibre is a re-release of another old Character plugin.
Calibre is another Character plugin, re-released in AU and VST form. This one’s got kind of a steely quality and if you hype the Character control there’s a point where bass cancels out: weird variation on a tone control! Also, that means if you duplicated the track (or used Blue Cat Patchwork or something) and flipped phase, you’d get a really unusual lowpass and would have only the bass and a lot of strange color.
############ Capacitor is a lowpass/highpass filter of a new type.
Sometimes half the fun is just inventing :)
People say all filter plugins are just biquads with different GUIs. It’s not true… or at least it’s not true at Airwindows. (actually I have to see what I can get out of biquads: I generally don’t even use ’em at all)
This is a new filter. When I say new, I mean that you don’t have it. I can demonstrate that, because due to the way it works it has a little characteristic response quirk at 1/3 the sampling rate: nothing that hurts the sound, just an odd little notch under some circumstances.
Because it’s a typically Airwindowsy, not-normal EQ algorithm, nobody has tried it, and because of the little quirk, if anybody did try it they knew they had to reject it, because it wasn’t perfect. It had a quirk! And even if they experimented, they probably weren’t that invested in concepts like the sequence of operations on stuff that, in theory, could be done in any order. Why care so much about the implementation details of an idea that had a quirk anyhow and was therefore useless?
More fools they :)
Because musically, this filter has tone for DAYS.
There’s almost an analog-synth-like effect to it: though it doesn’t accentuate the rolloff point, it behaves almost like a DJ ‘isolator’ in the obviousness of its filtering. It has a highpass, a lowpass, and dry/wet, and the high and lowpass are melded into each other as part of the algorithm to get the sound more fluid. It’s also got new experiments in rejecting zipper noise and clicks, because I’m figuring people will want to automate this one. But, even though it’ll work for crazy effects and ‘isolator’ madness, the fullness of the tone will also let you use it in mastering: if you need to tighten bass or roll off just a hint of brightness, you’ll pay no penalty in the body of the music. And the dry/wet is there to let you treat it as a shelf… or to create a presence boost by isolating an area of interest, and then leaving it mostly dry with a hint of the clarified zone.
I’m thrilled with this filter, and nobody else has it: nobody else would be allowed to do this one, because it has a quirk at 1/3 the sampling rate, and typical commercial developers are not allowed DSP quirks. But, since I’m supported by Patreon, I am.
############ Capacitor2 is Capacitor with extra analog modeling and mojo.
Didn’t see this coming!
My researches led me to a webpage by the electronics company Murata, and an observation: for a particular line of capacitors they make, namely high dielectric ceramic caps made of barium titanate, there’s a concern. The capacitance drops sharply if you put the capacitor under voltage pressure. How much? As much as 50% for a little over six volts. It’s pretty linear. Thing is, the signal is ALSO a voltage. What if it tended to modulate the cutoff? As part of analog modeling?
I have plugins, the old Lowpass and Highpass, which frequency-modulate the cutoff based on the input signal. But they did it symmetrically… what about doing it the way the real-world capacitor would do it? What would you get, in the event that other capacitors had some of this behavior? It seemed like you might get a lot of even harmonics, and people tend to like that. Why not give it a try?
And that’s how something interesting got discovered.
Capacitor2 is Capacitor, already a popular plugin, but with this analog modeling built in. There’s a ‘NonLin’ control that lets you crank up the distortedness from very minimal, to quite extreme. It’s sensitive to input level (naturally) so that’s another reason to have it on a control.
And what you get is INTENSE analog coloration, and something unexpected: it emphasizes transients and brings out the articulateness of sounds in a really distinctive way. You may not have heard anything quite like this… or if you’ve been using analog gear, maybe you’re used to hearing it. I really didn’t plan for the result I got: if real-world caps have any of this behavior, it explains a lot. Literally all that’s happening is modulating the cutoff frequency of the rather Airwindows-y Capacitor algorithm. There’s no dynamics processing in there at all, but the result is incredibly dynamic. (you can even use it to boost narrow bandpasses for effect!)
############ Chamber is a feedforward reverb based on the golden ratio.
I actually meant to do something completely different. Honest.
It was time to dig into the feedforward reverbs again, maybe rearrange them in some interesting way. Do something with the geometry of them, have them go two-into-three-into-four or something along those lines, and we were hanging out in the Monday coding livestream, starting to experiment… and one of my crew tossed out a sequence of numbers. Fibonacci numbers. Could we hear what those sounded like as delay constants?
Wasn’t bad. Chat got buried in Fibonacci numbers for a while: we are always enthusiastic at finding weird applications of things that shouldn’t be any use. After all, to get reverb tails to become seamless, the delay constants have to be set up properly. Prime numbers are key. Fibonacci sequences have nothing to do with this. And then, someone in chat observed… as the Fibonacci numbers get bigger they start approximating to the golden ratio.
Yoicks, scooby! We’d better try it! AND THEN…
Chamber is a feedforward reverb, using three banks of four delays each in a Householder feedback matrix, except it’s feedforward. Only the very end feeds back into the beginning again, just like Verbity, just like Galactic (it is dual-mono like Verbity, as its peculiar merits fit well with a dual-mono arrangement). And the delays go to a longest delay (maximum delay size) and each one in turn, back to the first, is exactly the golden ratio smaller than the previous. It’s like a big spiral of delay times, perfect to lots of decimal places. If you listen to just one instance of each delay (by turning Longness to zero, and Chamber lets you HAVE literally zero feedback), that’s a weird stuttery slapback. By itself, an arbitrary little chirp, a complicated slapback that doesn’t sound particularly interesting.
And then if you turn the feedback up, with Longness, it stretches out into a continuous, seamless, perfect reverb tail, just as if all the delay times had been worked out to be perfect little prime ratios.
This was an astounding discovery. It means you can dial in any degree of feedback or none, use any delay time (everything’s calculated out on the fly), do anything with it and it’ll adapt. It’ll always sound like a chamber, hence the name, but it’s maybe more malleable than any reverb I’ve ever made. And to make it even more malleable, Chamber’s Darkness control is tweaked so that the fall-off over time is always accurate to the sound of audio decaying in air in a theoretically ultimate room (studied from recordings of giant underground concrete cisterns) but the tone-shaping is darkened using very warm, basic IIR filters. And on top of that, a new control for the feedforward reverbs: since Chamber is such a studio tool, I gave it a highpass. So you can plunk it on any sort of buss or channel, run it mostly dry, bring in the reverb (Chamber and Verbity are designed so as you add verb, the dry remains unaltered until you get to 0.5 on the Wetness control, at which point the verb is at full volume and you start fading the dry signal) and then begin dialing back the bassiness of the reverb without touching the dry. Very useful for a chamber or plate send, and built right in!
Hope you like it. Oh, and one more thing: in working on this, I found a bug that was in the Undersampling code I use. The bug was making a bit of unwanted edge, only in high sample rate stuff, only above 20kHz. I’d had someone discover this in Verbity, but I didn’t know what to do other than filter it at the time, and hadn’t done anything yet.
So now (as of right now: redownload what you need) Verbity, Galactic, IronOxideClassic2, and Chamber are FIXED. Go back and redownload them, or get them out of the collections for various different platforms. They have all had the ultrasonic noise cleaned up: there is still a touch of audio there as part of the algorithm, but it turns out it’s much less than I thought and that’s reflected in cleaner sound. It should not change saved mixes significantly as it is entirely supersonic, but if it did it would only help as the traces of noise weren’t useful for any purpose, they were a bug, samples being slightly out of order at 96 and 192k. No change at 44.1 or 48k.
Chamber actually goes a step farther, in that I added more code that subtly averages the supersonic samples… that can apply to new plugins going forward, but Verbity, Galactic, and IronOxideClassic2 don’t have that as it could work like a tone change. So, compare the new Verbity to Chamber if you’d like to check out the tiny amount of added depth we’ll have going forward.
############ Chamber2 is a feedforward reverb, a blur delay, and a glitch topping!
Sometimes you just want to watch the glitch BURN…
So here's what happened. I wanted to try a modification to Chamber. It's a reverb where every delay time inside the feedforward network was exactly the golden ratio of the next. Why? Why not, I thought. What happened with that was, I got a sort of oddly-colored echo, but one that turned into very seamless reverb as long as you had some regeneration in there. Interesting! And so I coded a reverb where some of the delay taps were quite tiny, and that's Chamber.
But what would happen if it wasn't always the golden ratio? What if you tried other ratios?
Well, nothing for it but to try it, right? And I had to take ALL the delays inside, and make them potentially full length echoes, meaning the amount of memory it takes is WAY larger than what original Chamber wants. You can get the original tones out of it, but in doing that you're wasting huge amounts of delay buffer. The plugin just doesn't see them at all, and they sit there doing nothing. So, don't use Chamber2 where Chamber will do.
What happens when you have the 'thick' control at 0? You have the most expensive, wasteful, CPU and memory hogging delay ever. You've got over 4000 delays, all precisely the same. So don't do that either (note: if it were only that simple)
But what if you put 'thick' slightly off 0? You now have a blur echo. You've got a delay which is also a Chamber reverb in which all the echoes are ALMOST the same. And you can dial in the blurriness of this echo. Not only that, regeneration will further blur the echo. So you can take the no-blur setting, and sweep the 'thick' control higher while regenerating. And it'll (somewhat glitchily: you are buffer smashing) blur its way from direct echo into a chamber reverb, which will also make the echo happen faster (your internal delays are getting shorter, all in synchronization).
And then you can let the regenerations fade away. And then… what if you snap the 'thick' control back to 0 again?
Suddenly you have a full-on glitch buffer effect, from audio you had in the sample buffers when you went to the chamber reverb effect. Boom, there it is, at whatever delay rate you have set on the 'size' control at top.
Obviously this is extremely nasty. But it's also a shocking, bold effect with a tinge of the accidental. And when the effect turns up in the VCV Rack dailies, or is used in Bespoke or wherever… it's an open invitation to throw crazy LFOs and sample-and-hold on the 'thick' control, and just use Chamber2 as a glitchy noise generator. It'll grab buffer snippets from its delay mode, it'll blur them into reverb, it'll throw other echo bits on top of that: a proper mad scientist laboratory for sonic mayhem, from your friendly neighborhood Chris.
You can dial in nice verb/echo hybrid sounds and use those too, I won't stop you. I'm just making sure everyone understands the possibilities of this one. Chamber2 glitches in very special ways. Hope you like it :)
############ Channel4 is a tone coloring plugin that gives more of an analog feel.
Channel4 uses a very good-sounding saturation function, a calibrated slew clipper and a highpass to approximate the tone coloration of a recording console. (the plugin Console4 approximates the space and depth on the mix buss, which is different.) Because the bandwidth-limiters are set up based on measurements of real gear (through impulse responses), Channel4’s console type is labeled ‘Neve/API/SSL’. It doesn’t actually contain any of those things or any convolution kernels, but the presentation of the sound ought to suggest those voicings.
There’s also an argument for doing it that way, because in many cases you can’t hear those limitations unless they’re being way overloaded. Channel’s a very clean, pure ‘coloration’ plugin, especially with the drive kept at 50% or under. It’ll voice extreme sounds the way it’s labeled, but you would struggle to consistently set plugins like Slew to those targets because mostly you can’t hear what it’s doing, especially on normal/musical sounds.
############ Channel5 is Channel for 2018, with new subtleties.
Channel has long been a sort of test-bed of mine, and so when I started working with new methods for some of my stuff it was an obvious choice. I’ve been asked for an extra Output level control, which it has (for those of you who are really into level matching, can’t fault you for that though sometimes I think it’s not relevant) but if you set Output to 1.0 it bypasses itself and becomes exactly, exactly the same as Channel4.
Except for two things, which first appeared in Atmosphere: this lets you A/B them with the original versions, using Channel 4 and 5.
One, I introduced the concept of live air denormalizing: before this (and on plugins I think are suited to it), Airwindows plugins watched for total digital black silence, and if it was present, produced an impossibly faint hiss sound, at around 300 dB down. If there was any audio at all that wasn’t digital black, this got switched off. With Atmosphere, it’s different: the denormalizing signal is always there, and if the result WITH the signal is digital black, then the signal’s subtracted again. Same result, but there’s a permanent noise hundreds of times quieter than typical ‘noise adding plugins’. The idea is that there can be no ‘dead’ silence, only ‘alive’ silence: you can’t save the file in normal formats and retain it (24 bit audio saving will simply turn this noise to digital black again), you most likely can’t hear it ever no matter what you do, but I wanted the deepest darkest silence to be alive rather than dead black nothingness.
Two, it noise shapes to the floating point buss differently, and on the one hand I bet you can’t consistently hear it but on the other, I think it matters. Before, I used an interleaving technique much like I often use in IIR filters and plugs such as Capacitor. It related to the sample rate by forcing everything to be considered as the interaction between two or more samples, and the energy being brought back into the output audio provided corrections at high frequencies. What began with Atmosphere, and what you’ll (maybe) hear with Channel5, is a completely different and more radical approach. These plugins noise shape ALL the error energy back into the audio. It accumulates, sticks around, and so it can contribute to bass energy in tiny amounts because it persists over thousands and thousands of samples. There’s a ‘pull-back’ in the form of multiplying the amount by something like 0.999999, which will bleed built-up energy away, but this multiply doesn’t happen on a per-sample basis. It happens per BUFFER. So, in the world of noise shaping floating point error, this is the one that stores up every tiniest bit of error and applies it, not interleaved but very smoothly, to the audio.
What should happen as a result of this is, there’s a sense of ease from switching from interleaved to the smooth way. It might seem a little less lively, and I’ll have some plugins that retain the other way. But due to the greatly increased time during which the error is held, plugins that use the new noise shaping will be crazy good at presenting analog-like, seamless bass and mids. So, more soothing, possibly less exciting, but a huge jump in depth and naturalness.
…in RELATIVE terms to what was there before. That’s the thing. I feel it’s possible that when you get used to it the change can be dimly sensed as a good thing. I don’t think you, or I, or anyone would be able to pick this out double blind. I can think of a mastering guy or two that MIGHT, but they’d struggle: these are people who take pains to dither to 24 bit and say that they sense something wrong if they don’t. It’s on the order of that, or even more subtle.
But I do feel that one eventually senses the full range of a sound one works with, and more importantly if I can make something better I will do so, without hesitating.
Have fun playing with Channel5. And if anyone needs to sound the alarm and say HEY! You need to stop using the new stuff! That noise is driving me crazy, and your noise shaping sounds awful now!
…well, that’d be pretty funny :)
############ Channel6 uses the Spiral algorithm.
I use the plugin Channel to test out my latest code regarding things like how to prevent denormalized numbers, what’s the best noise shaping to return to the floating point buss, and so on. That’s because after a fairly brief debug period (at first, the model-choosing popup wasn’t actually changing anything and you always got SSL behavior) the plugin was established as a totally complete plugin: it always used the same ‘Density’ algorithm for the saturation, it always blended that with dry signal using a drive control that was really a dry/wet control (which gave it its openness at partial settings), and it always used the same slew clipping and simple IIR highpass code each time, which was so simple that it couldn’t be improved.
And it’s always been ‘one’ of my most popular and successful plugins, through all those variations. Once there was a big blind shootout on the most popular plugin forum (back when people did that) and someone put up my stuff against several contenders including the most hyped plugins out there. At the time I was selling plugins for $50 a pop, and I’d hoped someone would pick BussColors and do a comparison against my flagship tone-emulator, and have it clobber its rivals in blind testing.
Instead they picked the latest Channel, and used its settings of ‘API’ and ‘Neve’ and ‘SSL’ (in other words ‘slew clipping and a simple highpass and Density’, no real ’emulation’ just shaping set to calibrated amounts), and that free plugin was the one that clobbered the pricey rivals. And soon blind testing went out of fashion, because such shenanigans get embarrassing I guess :) And I didn’t get any sales out of it as Channel was free, but it was fun to watch.
So will this be…
Channel is back in version 6, and there’s one key change. The distortion algorithm which I thought was unbeatable has now been beaten by Spiral… so now Channel uses Spiral, not Density, for its distortion model! For once this should be a plainly audible change: Spiral’s less fat-sounding than the Density algorithm. (I fixed the bug with the dry/wet control in the video: now dry doesn’t clip)
What won’t change is this: you’d be surprised how little ‘processing’ can give you a great, analog-style sound. Literally all Channel does is apply a very simple IIR highpass, add a distortion that’s very clear and pure-sounding, and do a slew clipping set so high that it will almost never kick in. The highpass isn’t very steep, because it’s just one stage of interleaved IIR filter (a trick I use in a more complicated way in Capacitor). These are subtle, gentle differences… but they’re done with absolutely minimal processing, not tons of processing, and the data integrity is kept pristine through long double math and noise shaping to the floating point buss. In simpler terms: this doesn’t suck your tone like heavy digital processing does, this applies ‘hardware-like’ changes calibrated to measurements of real hardware, and this does it at impressively low CPU compared to other plugins of this type.
############ Channel7 improves slew clipping and lets you blend Channel 6 and 5.
Short and sweet. You know how Interstage got an update because the slew clipping was too intense on high sample rates? This applied to Channel, too. So, Channel7 now handles high sample rates better than any previous Channel.
There’s more: Channel5 was the last one with the ‘Density’ algorithm, that’s extra fat-sounding. Channel6 got the Spiral algorithm, much purer and clearer… but without that fatness effect Density gets. My attempts to make a special algorithm that does both gave me Mojo and Dyno… neither of which are Channel or a replacement for either Spiral or Density.
So, Channel7 simply goes to 200%, not 100%. Up to 100% it’s the same as Channel6, but with the slew clipping fix. From 100% to 200% it does a crossfade into Channel5’s Density algorithm—so it ends up being best of all possible worlds. You can seamlessly go from perfectly clean, to the purest saturation there is, and beyond that to add any degree of fatness you like to the sound (and it fades between Spiral and Density, so there’s extra subtlety to the way it transitions).
Channel7: yeah, you’re pretty much going to want this upgrade no matter which Channel you liked. It is all of the best, more effective than ever.
############ Channel8 is Channel7 with updates from Capacitor2 and Slew3. More and better of what Channel is.
Channel8 is a subtle ‘analogifier’ that applies saturation boosts, subtle highpass filtering, and slew clipping. What that means is, there’s a ‘drive’ knob you can turn up to boost things, and you can pick three ‘flavors’ you might recognize. They’re not really the consoles associated with those things, but they are calibrated to take effect about the same way, and when you use Channel8 (on tracks, on the mix buss, wherever) you add a little ‘analog flavor’ by restricting what the audio can do, in ways that are characteristic of real analog gear’s limitations. This restricting can help make things sound bigger and more solid.
This is basically the newest evolution of Channel, which has always been one of my most popular plugins. It uses the improvements in Capacitor2 and Slew3, so basically it’s still relatively subtle, still very much Channel, but it’s taken on a new level in quality and natural tone. I’m real happy with it, and I’ll be returning to trying to fix up my video setup etc (and finishing up the dither re-releases, etc etc) knowing that I’ve dropped a nice audio bomb on the world. If you ever liked Channel, this should make you real happy.
############ Channel9 is Channel8 with ultrasonic filtering, and new Teac and Mackie 'low end' settings.
Channel9 retains everything it had—the two-stage drive control, the newly improved highpass, the slew clipping—the same calibrated Airwindows algorithms to not clone, but give you the general sense of various fancy name brand consoles in a cleaner, less ’emulated’ way which lets the music through…
But now, Channel9 has the same ultrasonic aliasing-filtering that’s come to Console and other recent Airwindows plugins. In a new way! Because Channel9 isn’t just sticking to ’20k’ as its definition of supersonic. Instead, Channel9 steepens its filter with a teeny resonant peak at this cutoff point… and then selects it in keeping with what the real console would be doing, if you had it! The Neve is the most extended, well beyond 20k. The SSL cuts off tighter, gives more audible sparkle (due to the additional gain stages in a really huge SSL desk, the cutoff will be steeper over the entire desk). API is between the two. The lift at the peak comes before the saturation stage, for better smoothness when the console is being driven.
And then I went and added two MUCH more affordable ‘models’, with their own usefulness.
The new settings are ‘Teac’ and ‘Mackie’. That’s right, old school house/techno classic basement mixers! The Teac, I was able to reference recordings of a real unit. I didn’t get it perfect as the real board had a noticeable high-mids peak that doesn’t belong in Channel, but it’s the correct kind of dark and vibey. And the Mackie’s my take on what you get out of the classic vintage 1202: A hair leaner than the Teac, but brighter. They both grind a bit harder than any of the big expensive desks, they both have slightly more exaggerated reshaping of the deep bass (using the Capacitor2 algorithm, like the others), they both control the brightest highs much like you’d get in a classic old low end mixer. It’s two new settings that follow entirely different rules than the big guns, for folks who know how to use an actual mixer much like metal guitarists use a Tube Screamer.
You can still have your fancy desk models (only better: the ultrasonic filtering brings just the right additional distinction to the models) but now you can go cheap and get the tone and vibe of some house music warhorses. The sonic reshaping these low end models do is just the ticket for sculpting relentlessly synthetic sounds into an appealing result. And since it’s Channel… it won’t have the noise of the real ones, it won’t have weird extra colorations, it’ll honor more of your real mix as it reshapes it: sort of best of both worlds! You can always add funky colorations or noise to your digital mix, but you can’t remove that stuff from the real mixers. Channel9 will shape your sound in ‘classic’ ways but along Airwindows lines: getting out of the way so your sound is interfered with as little as possible.
Channel9 can be used anywhere you like. If you’re doing a Console mix, I’d put it after ConsoleBuss. If you’re not doing a Console mix, you can literally do anything you want with it: it’s a subtle distortion/fattener combined with a set of careful tone shaping algorithms. Hope you like it!
############ Chorus is a mono chorus, also works as a vibrato.
Here’s the start of some modulation plugins: Chorus will give you a nice basic mono chorusing effect, sweeping one moving delay tap against the dry signal. What makes it unusual is a pile of odd Airwindows things to adjust it this way and that.
The sweep is done with a peculiar Airwindows interpolation which both applies a bit of an averaging function, and also a subtle lift to highs to compensate for that. The idea is to have the moving part be totally fluid, analog-like, even though it’s digital.
To drive that, there’s a treble boost going into this smoothing delay tap, and it’s based on the Airwindows Energy algorithms: that means this plugin interacts with the raw sample rate in two different ways. It’s designed to make it so, whatever the sample rate, the sounds project through to the most extreme highs without edge or dullness. That means Chorus won’t sound exactly the same at 44.1K, 96K and so on: it will try to deliver the most finely tuned treble for each sample rate, whatever that is, and the tonality might be different. Also, since it’s running a delay buffer, it’s giving you full use of the buffer at all times: the broader settings might be more useful at high sample rates.
In general, this should be pretty approachable. Hope you like the tone: there’s more where that came from.
############ ChorusEnsemble is a more complex, multi-tap mono chorus.
Here we can fill out the Airwindows palette of modulation plugins a bit… like Chorus, this is using my special slightly dark interpolation with a little pre-sparkle to get an adaptable, rich chorusing effect. But ChorusEnsemble uses a bank of chorus taps to get a more complex, textured sound that’s farther from the original. You can set it wrongly, so don’t assume all the settings are appropriate: that said, a little care should give you nice lush chorusing that’ll work great on pads and backgrounds. The reason I allow for the ‘ugly’ settings is, who’s to say you might not have a use for them, and if you find that use you’ll have a tonal element that other people don’t have on tap (generally, it’s so hard to sell plugins that can sound wrong and broken that people will tend to shun that and limit you to ‘nice’).
############ ChromeOxide is an alternate path to vibey old tape sonics.
Chrome Oxide was an experiment, one that didn’t go further than this. It is a dual-band tape emulation, where the lows are a bit saturated but the highs are delayed by a random noise warble that can also be biased to delay them a bit further. My pursuits of tape emulation have always gone toward the ‘BETTER than digital’ direction, where I tried to capture the magic without diving into the audio degradation.
But revisiting Chrome Oxide (and re-releasing it, with modern wordlength handling etc and dithering to the floating point output buss) showed me a plugin that excels at some tonalities I didn’t even know about when I made it. For instance, your Boards of Canada type stuff, mulch-core audio where it sounds like it’s coming off an old Walkman or Wollensack? This will not do crazy pitch wobbles or dropouts… but you can instantly, effortlessly get the tone of it. The intensity controls a noise effect that is FM, frequency modulating the highs against the lows. Bias further delays the highs, and this sculpts the phase aberrations of the output and the flavor of roll-off… so, without ever getting aggressive or obvious, you can just dial-a-mulch and go as fuzzy and old-sounding as you like, but musically. It is subtle enough to use on anything and aggressive enough to completely change the mood of a track.
And now you can have it. Mulch away! You don’t have to obliterate a track to get into the vibe you crave. (and of course some people hate this sort of thing: if you doubly hate this one, I’ll know I’ve done it right :D )
############ Cider is a re-release of another old Character plugin.
Cider is another Character plugin. Works like the last few, sounds kinda like an MCI console? This plugin was posted on a week full of troubles, so there's very little instruction about it. It's a tone coloring plugin, and you can find a more current version of this sound in BussColors4 on the Dark setting.
############ ClearCoat is an array of bright ambience reverbs.
So, after months of work on refining Householder reverb matrices and realism and depth, and getting started on the huge tasks of mastering the tones of Bricasti reverbs and million-dollar analog recording consoles… I make a dual-mono bright ambience.
What's that for? First of all, it's for me. I'm using it immediately on album mixes. But beyond that, it's an example of something very Airwindows-like: once I get access to something, I share it. I wasn't able to do sounds like these when I did the kPlate reverbs. I tried. My tools for making the Householder matrix stuff hadn't evolved enough yet: this is why kChamberAR sounds the way it does. I had bright ambiences, but they were very artificial (though, nifty in their own right)
And this is what I'm sharing now: I got it to work. Next stage unlocked, with not one but effectively seventeen bright ambiences in one. ClearCoat is the one-stop shop for sparkle and atmosphere. It's not about lingering verb tails: the space it produces is dual-mono and remains tied to the sound that's creating it, and while there is feedback in the algorithm, it's calculated to be exactly halfway between the purity of infinite sustain, and the purity of BitShiftGain in the feedback path cutting the feedback by 6.08 dB. The idea there is to produce a consistent tone in the feedback path while allowing it to bloom a little. Basically more Chris from Airwindows experimentation: see if you like the results in the sense of space the plugin creates.
On top of that, I'm exploring a refinement of SubTight, the same plugin that backfired on me in ConsoleMC. I think if I scale it to sample rate a certain way, that might tame it, and the kind of energy it brought is perfect for this plugin. The brightening and subs-cutting is happening after the reverb matrix, to enliven the space.
It's simple to use: top slider picks a setting, bottom slider is your dry/wet. Add just a bit for in-line air on an instrument or submix, go all-wet for use on an aux. I'm using it as my Verb 1, with the upcoming kCathedral for Verb 2.
Why go for specifically this effect? The theory here is simple: you get better mixes when you're serving a purpose rather than throwing tone colors around just to make them happen. ClearCoat is there to be an extremely refined and intense form of 'nearfield reflections', like singing in a bathroom, or drumming in a live drum room. Yes, you can turn it way up and it'll sound like a really bright reverb, like a small (to medium) room with tile or stone walls. But what you can also do is turn it right down once you have the tone color you want… and then HINT at that tone color, with it.
And if you do that, ClearCoat ought to give you the gloss you're looking for, while minimally affecting the rest of the mix. No clutter, no excess energy, no trace of a tone character beyond what you need, and a cleaner mix as a result.
Let me know if ClearCoat works for you, and especially whether it's acting OK at diverse sample rates with the changes I made in SubTight. If so, I can and will update the actual SubTight plugin AND ConsoleMC to correspond (if you're working at 1x sample rates there will be no change for you, but if you're working 2x or 4x or oversampling 2x or 4x, you might not have to compensate for the brightness as much to get your sound)
############ ClipOnly is a clipper plugin that suppresses the brightness of digital clipping without affecting unclipped samples.
Here’s what I’ve done. I started with Clip2 and the original Clip freebie, and I used the new plugin as a way to experiment with ideas about the recurrence of mathematical constants. For instance, a common value used in reverb allpass filters is 0.618… which can be continued into a mathematical constant, the Golden Ratio (0.618033988749894848204586). It’s common for these things to turn up in disparate situations, so I look for them. And in the code where OneCornerClip rapidly converges onto full clipping with the broadest resulting variation of clip onset tonality, I found it zeroing in on cos(x) == x (0.739085133215160641655312) and selecting that optimal value for the guts of the thing I ended up with ClipOnly.
If you watch the video (one of the least necessary to watch for instructions: it’s a clipper at -0.2dB) you’ll see a lot about the other factors that went into ClipOnly. I show ADClip and its many controls, but then make an impassioned case for designs like ClipOnly, a known quantity with a predictable sound and no controls or adjustments whatsoever. Under normal circumstances there should be nothing you’d want or need to adjust, since the clip level combined with the anti-glare prevents any ‘digital reconstruction overs’ in normal use. It doesn’t do anything tricky to ‘enhance loudness’ beyond what it already is, unlike recent ADClip versions. It does have a very sophisticated enter/exit clip behavior, but to put it bluntly you don’t need to be thinking about that when mixing into a safety clipper (or tracking into such a mix).
In fact, you should not be thinking of that OR me while tracking your music and mixing it, you should be thinking about the sounds you’re making, and any distraction into the lands of ‘oh gee Chris is so clever, ClipOnly was well worth the money, look at all these knobs’ is hurting you even as it rewards me, the plugin developer, with attention.
And THIS is why I program things like ClipOnly, where possible. It might be the most forgiving safety clipper yet, as far as avoiding glare and harshness (I’m measuring extreme highs as much as 12 db down from the max under heavy load, and usually clippers by their nature exaggerate ‘crunch’ even when perfectly implemented). But ClipOnly is also part of a concept that I can run with now that I use Patreon: I get to make stuff that’s not showing off, either with controls or concepts or fancy faceplates, and just delivers the sonic goods.
I hope you like ClipOnly, but I also hope you quickly stop noticing anything about it, so you can focus on your mixing. You should be able to work into it, even hit it with heavy load and severe clipping for effect, without fussing or being distracted. The best silver bullet plugins are not the trickiest :)
############ ClipOnly2 suppresses the brightness of digital clipping without affecting unclipped samples, at any sample rate.
ClipOnly2 is the heart of my mastering-grade clipping algorithm. Instead of trying to define the cleanest possible nasty sharp edge, or doing a soft-clip thing, ClipOnly passes through ALL nonclipped samples totally untouched… but when you get a clipped sample, what ClipOnly does is it takes the sample entering clipping, and the sample exiting clipping, and it interpolates between the last unclipped sample and the clipped stuff. So, it is synthesizing a soft entry and exit from what is otherwise total hard clipping, and if only the one sample clipped? That very bright clip simply goes away, turned right down.
This produces a hard-clip suitable for safety clipper purposes, which is purely ‘bypass’ (plus a one sample delay to allow for the processing), with softer highs than you’d get from any pure hard-clip, no matter how oversampled. It’s an alternate technique, and is also pretty CPU-efficient.
ClipOnly2 takes this principle and changes the ‘one sample’ to ‘the space of one sample at 44.1k’. Same tone, same ear-friendly approach to clipping extreme highs, except that now it’s effective at high sample rates. I’m demonstrating it and its predecessor at 96k, but ClipOnly2 is designed to work up to 700k or so, in case people get giddy with their newfound power :)
############ ClipSoftly is ClipOnly2 but as a softclipper: a very handy building block.
What more needs be said? ClipSoftly is ultimate softclipping, using the techniques used on ClipOnly2 for softening the edge of the hard clipping, but on the softest possible clipping. Where ClipOnly2 manipulates the onsets of the brightest transients to give you pure hard-clipping without digital grit, ClipSoftly is doing that without even hardclipping.
So what we have here, basically, is a control-less (like ClipOnly and ClipOnly2) safety clipper, but it's a saturation effect. It's also a building block I (or others) can use inside larger plugins or new versions of Console, much like the recent Pop2 uses ClipOnly2 on the output, to give you the ability for really aggressive attacks.
You'd use something like ClipOnly2 in mastering specifically because it won't touch the values of any unclipped samples. ClipSoftly is different: it will touch the values of ALL samples, reshaping the whole sound to make it bigger, fatter, tubier.
I hope you like it, and the way it'll affect my plugins going forward :)
############ CloudCoat is an array of blur effects with a taste for evil.
Welcome back to the dark side of Airwindows!
CloudCoat starts as an experiment. What if there was ClearCoat (a bank of small reverbs) except all the delays were allpasses?
Okay, so a bit of explanation: allpasses are like reverb parts, but they make stuff sound smeary and blurred, like bloom reverbs. I've got a bloom reverb, MV, which is just a stack of allpasses. Most of my recent work is about avoiding allpasses completely by using reverb matrixes that give me LOTS of echo returns, impossible numbers, so I don't have to cheat with allpasses. This kind of works and kind of doesn't (work in progress, see my recent livestreams). But what would happen if you took what was clearly a reverb, and just replaced all the delays with allpasses?
That's ClearCoat and CloudCoat. The idea was I could give a completely different texture, but using literally all the same reverb constants, and then I could hear what it was like. I expected it to be more a cloudy, diffuse texture, hence 'cloudcoat'.
You might notice one difference right away: ClearCoat sounds a lot roomier. That's because it's designed with a little bit of feedback to fill it out. Sustain, if you will. It's also way more spacious, and way more metallic and ringy. This is in line with how it only uses delays, and is all part of the research.
CloudCoat with sustain all the way off, is quite different. Depth and spaciousness is almost gone, but there's no metallicness either. It's like essence of fake artificial reverb blur. Remember, this too is a 4x4 Householder matrix: it's a complicated pile of allpasses, not something primitive like MV. I think it might find uses on pads or ambient sounds, or could be used to feed into ClearCoat at the same 'select' setting to create a more powerful room sound. With sustain all the way off, you can do many polite and nice things with CloudCoat.
Then, throw it on some drums and turn the sustain up and ALL HELL BREAKS LOOSE :)
CloudCoat is applying a kind of nonlinear reverb, like your classic 'nonlin' gated verb settings, but Airwindowsized. I'm allowing feedback, but of the four 'channels' of Householder feedback, each is also modulated by the output of the one next to it (a totally different sound)… WITHOUT smoothing. So the feedback is broken up by four banks of allpass output and cranked up to the point of meltdown, and that's CloudCoat. It disrupts the signal wildly and fiercely.
Why?
Because I tried it, on a livestream, and the drums absolutely exploded like nothing I've ever heard. There's rasp and rattle and an effect much like extreme compression while in a stone drum room, but there's no compression and no stone room. It just makes that sound. Instead of making it out of compressed, flat-topped compression smash, it's making the madness happen INSIDE the sound, meaning you can make it brutal and unbearable with ClearCoat and THEN loudenate it, unlike any other sort of compression or distortion. CloudCoat adds a whole new type of trash that is dynamic with your sound (again, nonlin) and automatically dials itself back if the source energy level does. It's a huge, nasty, energy-laden meltdown that can be escalated to pretty much any degree… and then dialed back down again, to pretty much any level of controllability, so long as you're using it on noisy percussion and as long as you're okay with its signature trash-sound, which is not like anything else and which is independent of added compression and distortion.
And then you can do blur/texture effects with the sustain on 0. Just don't bother trying to make it work with tiny amounts of sustain: it basically can't. 0 or trash are your only options, and then there is a universe of trash and meltdown if trash you choose.
But seriously, why?
Because it makes me laugh with delight when snares go off like something out of the imagination of Trent Reznor. I don't know what to tell you. Enjoy CloudCoat, if it's the kind of thing you enjoy.
############ Coils models the types of distortion you'll find in transformers.
Coils is like the inverse of Focus. Instead of boosting/distorting UP the mids, it saturates DOWN the extreme highs and lows leaving the mids untouched. Because it’s a fixed mid shape mimicking the Neve transformer picture I was shown, the biquad doesn’t have to run inside Console as it’s not a high-Q filter at all (you’ll notice way more of an effect on steep filters) so it trades off that effect for less processing. The saturation is the Density algorithm, but without clipping, so if you trash it you get wrap-around on the transfer function curve, which works out to be kinda transformery.
The top control increasingly distorts (still kinda subtle), the middle one DC biases the core (second harmonics!) which does more when you’re less saturated, and then there’s a dry/wet. Enjoy :)
############ Coils2 is a transformer overdrive emulator.
This is to Coils what Srsly2 was to Srsly. In other words, Coils was always ultra subtle. Coils2 lets you dial up the cheapness until the tone is wrecked (if you so choose).
There’s two controls, Saturation and Cheapness. Saturation determines what happens outside the ‘resonant band’, and Cheapness narrows the band in which the transformer’s putting out clean, pure sound. This is not a ‘stompbox distortion’ type of distortion. It’s shaping the way the transformer produces sonority. If you narrow the ‘sonority’ band with more Cheapness, you increasingly distort and lose the lows and highs, and also get a hysteresis effect of magnetizing the transformer core. More Cheapness lets you hear the transformer crap out better when you overdrive it.
Saturation gives you the maximum overload you can get to, NOT ‘more total gain’. It’s clamping down, not boosting into. That means if you turn it up all the way you get a sort of bandpass: it’s no longer really a model, you’re hearing only what’s left over after the transformer dies. To hear the grind, you have to set it to less than full crank, and halfway should already be quite a lot of overload (except if Cheapness is really low, it might be hard to overload the transformer, so it’s a matter of taste.
That’s a lot of talk to say: play with the knobs. They should do what they’re labeled to do, and as long as you’re not thinking ‘turn everything up all the way’ you’ll be fine. It’s meant to pass through a great deal of sonority even for tiny cheap transformers, just in such a way that you can really hear it this time. Coils2 is still in the spirit of Coils, in that you’ll get the most accurate ‘modeling’ by not treating it as its own stompbox. It’s the output stage, for shaping and sculpting things that already exist. You might combine it with tube distortion effects to get an ‘amp-like’ character, or use it subtly on a 2-buss to tighten lows and give you more impact and vibe. Remember to not use too much saturation and cheapness for full mixes :)
############ Cojones is a new kind of distorty.
Cojones is one of the promised releases from back when I started all this. What’s interesting is, Cojones is also the seeds of Dither Me Timbers and StudioTan. That’s because it does a similar thing: it tracks the trajectory of the waveform (over five instead of three samples) and either heightens or minimizes any disparities it finds. It’s called Cojones, because I thought it highlighted that sort of quality in voices and guitars, though it’s easy to just make it be distorty and strange.
You’ll find that boosting Cojones can give a peculiar sort of midrangey sonority. I’m not going to say it’s GOOD sounding, but it is at least distinctive. There’s also a ‘breathy’ which is more three-sample stuff like Dither Me Timbers and StudioTan, and a ‘body’ control that can beef up or cut bass and low mids.
Pretty much play with it and if you hate it, throw it away and curse its name and mine. It’s all the rage! :D seriously, if you’re the sort to like this, you know who you are. If you’ve been putting Dither Me Timbers or StudioTan in places that aren’t the output dither, you need to try this instead as you’ll get a lot more out of it. And if its seasoning seems way too spicy and always produces trebly grit, try very slight amounts of its mojo, as this is one that’s set up so you can apply too much.
After all, what good is an ugly new distorty if you can’t overuse it and make unpleasant yet unrecognizable noises? :)
############ Compresaturator fades between compressing and soft clipping.
One of the final plugins from the Kagi era, this stopped being available for a while when Kagi went out of business. It’s like a soft-saturator in which you can dial in even more cleanness, but couple it with a weird springy quality that’s very dynamic. It occupies more of the role of a stem or buss compressor than a peak limiter, but makes for a very powerful loudenator. The idea is that you can get a loud punchy sound but with more attitude than usual.
Since it’s one of those strange Airwindows algorithms, be aware you can push it too far and get it to ‘flutter’ or oscillate like a tremolo or do other odd things: in normal use that shouldn’t ever be a concern, I’m just saying that this plugin isn’t normal and doesn’t sound or act quite like other plugins.
In other words, classic Airwindows :) hope you like it!
############ Console0Buss is a radical, minimalist Console system.
############ Console0Channel is a radical, minimalist Console system.
If the weirdness of Airwindows Console systems tends to throw you, it's probably good to skip this one. You'll either hear what it's about… or you won't. This is sort of an extreme position of a line of inquiry that not everybody buys into. It's really quite clear what's going on and why, but people disagree very intensely on whether it matters. Rather than make allowances, I thought I'd push it to the most radical extreme possible to see what happened.
Meet Console Zero.
There will probably never be anything that goes farther in this direction… not in the purely digital domain, anyway. Not as free open source plugins that everyone can have. It's easy to get external hardware to do some of these things, but that's costly. Console Zero can be used by everybody… if, that is, you can deal with the demands it makes.
Console Zero is the Airwindows Console concept, crossed with the BitShiftGain concept, and pushed literally as far as it can possibly go. It has built-in gain and pan… kinda. It has aliasing-suppression filtering… kinda. It does the saturate on channels and then anti-saturate on the buss… kinda.
Everything, everything is sacrificed to the Mantissa Gods. The idea is, with many analog-to-digital converters, with even fairly humble ones you can get quite a lot out of them if you just pass the audio straight through. Analog to digital straight to analog again? Often, it'll capture a really good sound. It all goes to hell when you start trying to work with it in the box.
Even on a system like Console8, with NO other processing, there might be hundreds of mathematical operations on every channel, thousands. Things like sine and arcsine functions do a lot of processing to be accurate. Biquad filters can get you nice accurate filtering, but require lots of math operations to function. The stuff modern mix topologies get up to could be tens of thousands of calculations, PER sample, PER channel on the way to the final mix. We take this for granted and nobody tries to make the opposite approach work, certainly not while including filtering and analog emulation.
Console Zero does maybe eight math operations, per sample, per channel, between input and the final mix output, that touch the mantissa of the input values. INCLUDING the actual mix. Including the Console saturation/antisaturation system. Including the aliasing-suppresion filtering. Eight, if I remember correctly… for the ENTIRE chain.
How? By leaning on bit shifting, to an insane, excessive degree. It's a giant trade-off. All level and pan is done with built-in BitShiftPan. (even on the buss, and you shouldn't touch it… but why not? The point is that it doesn't alter any mantissas). That is mixing in 6 dB increments, 3 if you count moving stuff one notch to the side to make it quieter. The filtering is strictly simple averaging. One operation, and a bit shift to get back the original gain: the bit shift doesn't touch the tone. The Console system is simplified so much that it's just one calculation and another bit shift. Everything is traded off for minimalism. The saturation produces slightly more harmonics than something like PurestConsole, but almost without calculation.
You get a mix together in Console Zero through arrangement, through broad strokes. You CANNOT fuss with it. It's almost LCR panning, except there are obvious left-center and right-center points included, and a range of pannings nearer the edges. All these points sound exactly like unprocessed raw digital audio, all of them pass through the unaltered mantissa from the input sample. The entire mix, with all its levels and pannings, puts through every single track as if it was the untouched raw signal without even a gain change… because there is not a single mantissa change to any track on any channel, going into the Console processing.
This may mean NOTHING to you. If you got here and that's you, thank you for the great patience. And… does it sound good? All of this is in pursuit of a particular KIND of sound, very unlike typical DAW sound. Do you hear that in my example, do you hear it if you attempt a Console Zero mix? If you hear nothing unusual, move on, this is not for you.
Some of you are going to lose your minds over how good this can sound. This is for you :)
Wait, it gets worse! You have to run 96k this time, 192k if you really want extended highs. If you're at 44.1k, the averaging will still work. It will be REAL DARK. Sorry! That means some people could do mixes at 44.1k with some of the desirable qualities we're seeking here, just so long as it's meant to be dark. Plenty of genres that can do that. But you should be running 96k. These plugins are SO lightweight, it's hard to even express how light they are: again, eight operations across the entire mix buss, not thousands or hundreds of thousands, per sample. That's from input to mix output. That will stand up to a lot of tracks… and the other secret is, Console Zero LOVES high track counts. The more stuff you have layered, the more freedom you have to do subtle loudness adjustments: a 3 dB nudge on one layer of one track made of four layers means the total loudness of those doubles, is the same as adjusting all of them by 0.75 dB. It will nudge one of the layers sideways in the stereo picture. And? The task becomes managing aggregates of tracks, thinking only of the big picture, letting some stuff be buried in the mix because you simply can't nudge it up in volume to balance.
And it can all come together and work, better than you'd ever imagine, and quicker… and sound like life, not like a DAW.
Either this will work for you, or it absolutely won't. Either it'll blow away anything you've tried, or completely frustrate you. This one's gonna be polarizing. It goes so far in the directions I try to achieve, that I'm certain it'll blow some of you away, and it's gonna make other people really mad. What price is an untouched mantissa? How much can you hear the inevitable digital erosion of overprocessing… and just how little processing can you do in a mix and still have a mix at all?
Console Zero… if you dare :) I will return to more accessible stuff, analog emulations etc, now. There will be no beating Console Zero for what it does. The rest is doing interesting tone colors, which I promise I'll bring you… but if this is your jam, you're done. There won't be any better from me or anybody else. It will beat many external summing busses. Console Zero is yours now.
############ Console4Buss is the original Airwindows ITB mixing system.
############ Console4Channel is the original Airwindows ITB mixing system.
Console4 is a two-plugin mix buss replacement for DAWs. This was also the release plugin for my switch to Patreon, and the announcement that Kagi had gone out of business owing me hundreds of dollars, so for this Airwindopedia entry from August 3, 2016, I'm including the Patreon blurbs and explanation of what my business was going to do. Turns out everything I promised, I delivered on, and now everything is MIT-licensed open source. Over to the 2016 post!
Console 4 is the most recent version of basically my flagship Airwindows plugin. It works through a channel plugin on every sound source, feeding at unity gain into the buss plugin which overrides the digital summing with Airwindows summing (more internal space, more depth, and with Console 4 it now has glue making the top-end more listenable).
You can also watch the half-hour long introduction video that goes into a lot more detail about how it’s used, how it works, how you make ‘big kicks and snares’—Console doesn’t let you crank any given channel up to full volume, but it’s designed to layer stuff so if you need giant sounds the best way is to make them composite sounds, everything layered from separate channels in the mix. Rather than layer samples and put them on a single loud track, keep the layers accessible in mix. That does mean if you wanted things like sines to be superloud you’d have to get creative, but the principle’s clear.
Speaking of principles…
You may notice no demo link, and also there’s no Kagi shopping cart link. There’s a reason for that. Kagi went bust Sunday. I may still get my last two months of sales out of whoever’s divvying up the assets, but as of this Monday you cannot pay me for my work through Kagi, because there is no Kagi. They served me for close to ten years, always with perfect efficiency and scrupulous honesty, but the commission off my work wasn’t enough to keep ’em going.
I could go find the newest-trendiest shopping cart e-commerce people, but when I started shopping-cart shopping I found them all horrible, tacky hypemongers offering to do things like find people who’d left a cart un-checked-out, and spamming them with reminder emails like ‘Hi, I’m your shopping cart and I’m looonely! Do you miss me? I miss you! Surely you just forgot me?”
I ran away before I threw up all over their e-commerce portals. After all, I have never spammed or bugged people, never advertised, and what’s more I give people free updates for as long as I live, none of which ‘makes sense’ in this happy future of badgering people for every cent, preferably by hyping them into a rental arrangement and DRMing the stuff until it’s nearly ready to explode all by itself, never mind when the rent is due or the authorization servers are having a bad day. (but I digress)
There’s something completely different I can do—something I would never have done, except Kagi went out of business and I have no reason not to be completely rebellious and flip the marketplace table.
Ever heard of a thing called Patreon? It’s not for discovering new artists. Really, it’s more of a… payment processor. For people who are already well established, who are appreciated for what they do, and who are busy devoting their lives to giving the world something for ‘free’ (like comics, or perhaps music, or art). Much like I already do for existing customers: I’ve promised all Console owners that they will have all Console updates for free.
There you have it. I am going all in on Patreon, and that is the full release version of Console 4, with the new Mac and PC VST builds, for free. Please remember this when I have figured out the details and started my Patreon. From now on, I will be relying completely on that to survive. I’ve seen more than one person (for instance, Vechs, or Jim Sterling) who are doing great with Patreon, usually because they too are rebelling against some commercial thing and making a bid for total freedom from obligation.
From now on all Airwindows plugins will be ported to Windows and Mac VST as well as being Audio Unit, and they will all be free from now on, and if I get enough Patreon support I will release all source code under the MIT license and document it as I go so that everybody can use the tools and concepts I’ve built to create their own software. That will be my legacy, and if that ain’t a worthy Patreon goal I don’t know what is.
This includes old versions of plugins for when people preferred a certain version, so the total number of plugins to cover is over 250. That means if I do one a month it will take me more than 20 years. If I work like a maniac (well, more like a maniac) and put out one a week, that’s still around five years just to turn Airwindows into a sort of audio plugin library and DSP school. And it will all be AU/VST with VST covering Mac VST2 and Win VST2 (built on older OS versions so the plugins work on the very widest range of hosts). I feel this isn’t a bad strategy because if I set up the Patreon so I can actually survive on it, I can make the rate of these ports (and free releases of the existing AUs) conditional on whether I was able to eat that month, which seems fair. :)
That’s not counting new research and new plugins… but I have a lot of VST plugins to do, to catch up. Over 250 of them. I’d also ask pirates and haxxors (who don’t have any VST airwindows but what I make, and don’t have most of the AUs) to please leave this ‘making them all free’ process to me through Patreon and my own website. If all goes well, all the plugins will be free in the end (even with source code!) so if you could not mess this up I’d be grateful. ;) leave it to me, please.
Oh, and the VST versions do ‘double replacing’, which means they noise shape to the 64 bit floating point buss. Technically that makes them higher sound quality than Apple’s CoreAudio can offer, though I promise you really won’t notice (and the VST versions also do 32 bit ‘replacing’ so they’ll work on all hosts)
Consider Console 4 an advance on this new concept. If it works, and if I can live for five or twenty years doing it, every possible Airwindows plugin will be part of everyone’s toolbox and the code will be out there making people’s products better.
############ Console5Buss is a richer, warmer Console system.
############ Console5Channel is a richer, warmer Console system.
############ Console5DarkCh is a richer, warmer Console system.
Console5 uses some more ‘expensive’ math operations, where previous Consoles tried to do their thing while keeping the Channel component as low-CPU as they possibly could. This might mean a heavier CPU cost, or it might be not that much of a difference. It’s a change (the math here more closely resembles Density or PurestDrive).
What do you gain? Using this more advanced math means there are functions which can exactly ‘undistort’ what comes in (more on that property later: there’s a variation on Console that perfectly nulls when only one track is active). This brings an added level of bigness and signal purity. Then, Console5 applies a similar behavior to the slew factor, but backwards to what the amplitude factor is getting. Doing that takes Console5 away from perfect transparency (and subtlety) and gives it a big, beefy, large-console sound that still doesn’t radically alter individual tones… but throws in TONS of ‘glue’ and solidness compared to the raw digital mix.
This is not a thing you’d struggle to hear (listen for depth and space, not frequency changes). This is not a thing that’d get washed out in mp3 encoding (in fact, because of the way it restricts slews in Console5Buss, it’ll actually help encoding a teeny bit, because superhigh frequencies waste bandwidth better used on the mids). This is the new Console, and it should be a real revelation to mix through, no matter what style or genre you’re working in.
As seen in the video, if you’ve got a DAW that can enable/disable plugins on selected channels, you can audition it with one mouse click to switch. Console5 works like this: you want Console5Channel on every channel feeding the 2-buss (with all submixing and all post-plugin faders at unity gain), and Console5Buss first on the 2-buss. That’s all, just replace digital summing with this system. If you can do post-fader plugins, you can use the faders (otherwise, best use the trims on the Console5Channels, or any earlier gain trim). The point is to replace your digital summing network with the Console5 system.
If you have that mastered, you can start playing with stuff like putting things ‘inside’ Console: delays, reverbs, EQs. Plain digital EQ in particular benefits from being post-Console5Channel on the track. Gain stage everything so you’re not slamming Console5Buss more than about +3 dB: it should survive hot peaks but there’s no special benefit to clipping it, and Console5Buss will clip there. Ideally, you’ll frame a mix with Console5 in place, and you may find you don’t need to do nearly as much ‘twitchy DAW stuff’ to get things sounding acceptable. Console5 addresses the root of the problem in a way no other ‘console emulation’ does. (if they do, you’ll find they have exactly the same constraints: needing to keep unity gain between Channel and Buss plugins is a dead giveaway they are using the Airwindows design)
This is the Console version that can generate DC offset from sawtooth waveforms due to its interaction with slew rate handling, and the final Console5 you can get attempts to accomodate that. For the launch version (with a different DC-block strategy in place), use C5RawConsole.
############ Console6Buss is a new more intense Console that peaks at lower dB.
############ Console6Channel is a new more intense Console that peaks at lower dB.
A little while ago I got a note from veteran GSposter torridgristle, entitled “Simple Encode/Decode (Inverse Square)”. It was a very simple little note.
“Something I’d been using in GLSL shaders for brightening the image turns out to work pretty well for an encoder / decoder pair when demoed with Maths. It may already have a name but I refer to it as inverse square. 1-(1-x)^2 and 1-(1-x)^0.5 https://www.desmos.com/calculator/3vhxkwjjyi The quiet parts are shaped like a plain multiplication, just a simple scale up or down, and then it curves back toward 1.”
Now, this wasn’t quite code yet (for audio purposes, it has to be able to do negative voltages, perhaps suppress inputs that would give broken outputs) but I asked if I could have it under MIT license anyway, and torridgristle said yes, and off I went to code it into a Console-like thing to see what happened.
Right away, I noticed that it was kicking in WAY harder. You could still put full-scale audio through it as one track, but for a dense mix, it broke up a lot sooner than Console5 or any previous Console had done. The channel plugin made stuff a lot louder and more aggressive, and the buss plugin cut stuff back harder. I had to pad my choir demo back a lot simply to stop it exploding in distorty noise. However, when I did… it was a hell of a big, intense, Console-style sound. This one wasn’t subtle at all.
And then, when I remembered how many times people had wanted a Console that did NOT have to peak near 0dBFS… I knew me and torridgristle, whom I’ve never met and don’t even know the first thing about… were onto something.
That’s the neat thing about open source when people are willing to use it. I mean, yes Patreon, you wouldn’t have this at all if not for that, you wouldn’t have future stuff inspired by this etc etc… but the heart of it is, you have the newest greatest Console because two complete strangers got together and cooperated. torridgristle used this in shaders and has never heard what it sounds like until today, and had no idea what it would do or what it would sonically mean. I know an incredible sound when I hear one, and I know it’s useful for people who are NOT trying to mix super-loud into their DAW buss, but I didn’t think up the very simple and elegant little algorithm. torridgristle didn’t code the Mac AU and VST Mac/Win/Linux ports, I didn’t code the GLSL shaders and have never seen what they do.
But the whole is greater than the sum of its parts, and without either of us being too troubled with great struggling, Console6 happened, and here it is.
So the bottom line here is simply usability: mix into this like it’s an analog mixing console that can be overstressed. You can still put full scale single tracks through there, losslessly, because that’s what any Console does. You can mix into it provided you don’t stress the buss too hard (or don’t mind the distortion you get). Gain stage before hitting the Channel plugins, one Console6Channel plugin on each sound source with unity gain through the mix structure into the Console6Buss plugin. If you need to slam stuff louder afterward you must do it AFTER the Console6Buss plugin, in whatever way you like. Console is always a digital mix buss replacement without an EQ-style sound of its own, so if you want further color you need to use other plugins outside Console. If you’re daring or just curious you can use things between the plugins (‘inside’ Console) such as a DAW EQ, echoes, subtle reverb etc. If you choose wrongly, it’ll distort or be weird, and things like EQs will be much more sensitive than usual to boosts and cuts.
############ Console7Buss adds anti-alias filtering and special saturation curves tied to the slider position.
############ Console7Channel adds anti-alias filtering and special saturation curves tied to the slider position.
If you don’t already know what Console is: it’s the Airwindows digital mix buss. You put the channel plugin on every channel, feeding directly (at unity gain) into the 2-buss where the buss plugin lives. It applies saturation and anti-saturation functions so that, for individual sounds, there is no change, but when there’s two signals interfering with each other, it makes the channels saturate easier if the buss’s ‘input impedance’ is fluctuating based on other signals coming in. You set it up, and then mix with gain trim controls or the controls on the plugins because to change the faders would violate the need for unity gain between the plugins.
That’s been the case for six previous versions of Console, and now it’s Console7. Here’s what’s new, that I didn’t have before.
Every stage of the Console system now runs ultrasonic filtering. Not ‘the Ultrasonic filter’, which is heavier in CPU and steeper: it’s a system designed and built for Console, optimised for use with Console. It’s a gentler, less phase-smeary version equivalent to the Isolator filter across the entire Console system, but set up backwards: the Channel plugins lead off with the steepest stage of filtering, causing highs to hit the saturation in a particular way. Then, on the Buss plugins, the remaining two stages use decreasing resonances, so the end result is as flat as Isolator’s fifth-order Butterworth filtering: but one stage runs before the processing, and one after. Doing this causes aliasing to be repeatedly removed at every step it might occur, rather than trying to whack it completely on input and then expecting the whole chain to be clean. You can still drop Ultrasonic in there, anyplace that you think needs extra attention… but this is actually better. Especially if you’re working at 192k (but it’s designed to be fantastic at 96k).
Every channel and the buss now gets a dedicated seed value for the dithering to the floating point buss. This might seem (and in fact is) a mighty subtle point, but it turned out to be fine to do at no cost to the CPU of the actual mix (it’s just a little extra getting done as each plugin loads). In a (real, though kind of theoretical) sense, that means every single channel produces its own dedicated noise for dithering, even though it’s just to the floating point buss. No previous Airwindows plugin has done this, but it worked so well that it’s now the new standard for how they’re built.
Every channel and the buss now has a dedicated saturation/anti-saturation algorithm that ONLY exists in Console7. It’s based off of a blend of Spiral, and Density, with the first instance of Spiral run as a ConsoleBuss algorithm, ever. They go to the trouble of blending between this new Spiral/antiSpiral sort of Console, and the Density-based one as seen in Console5 and PurestConsole, because doing this allowed a tweak in the way channels hit saturation, where the harmonics are generated in a balanced way, a smoother onset of saturation than I’ve ever had before in a plugin. Console7 channels saturate in an incredibly sweet, non-edgy way, and that’s before the ultrasonic filtering.
All the channel plugins now default to 0.772 on the gain control. That, not 1.0, is ‘unity gain’. Why? Because you can now push Console channels into the red in a special way. For the first time, the gain staging is flexible and tied to the Fader controls on the plugins (the Master control on the buss also does this in its own way, but that’s normally kept at 1.0). Unlike any previous Console, and opposite to what you get if you use the DAW faders, these channels saturate MORE as you push them, and saturate LESS if you pull them back. By the way the Density algorithm works, that means stuff tends to come forward as you nudge the gain up, and drop back into the soundstage if you pull the gain back. It opens up in a very literal way when you pull channels back, like some idealized analog console. What that means is, if you use these controls (they are smoothed for zero zipper noise) stuff will practically mix itself: the mix ought to fall into place more easily and quickly, plus if you’re whacking around the controls in some mad dubby way it ought to romp with you quite delightfully! They are simple 0-1 controls specifically to get you to set them by ear: there is no such thing as ‘dB’ with these, and even if there was, you’re adjusting the saturation curves so it’s completely down to what sounds right. I recommend using control surfaces to ride these Fader controls in the plugins: this is another way to get back to (automatable) analog console days.
############ Console7Cascade is a drop-in replacement for Console7Channel that allows for MUCH higher gain.
I heard you liked slamming consoles. So I put five individually ultrasonic-filtered stages of slamming into your console so you can slam console while you Console7 :D
This is pretty straightforward. It’s a drop-in replacement for Console7Channel, right down to the gain staging that works with the trim control to fit the result into the mix.
The difference is, this uses FIVE gain stages of the same processing in Console7Channel. And of course it’s always better (I’m learning) to filter more gently between individual stages, rather than try to super-filter all at once and then do all the distorting. And so, Console7Cascade is born: turns out to be a very very ‘consoley’ type of crunch.
By that I mean it seems to barely crunch at all. It just intensifies and gets REALLY LOUD. Might work as a guitar amp sim too? It worked so well for me on my drums that I might end up just using it by default for that: just all of the channels, all get Console7Cascade. You could also put it in place on a submix… or all the submixes, if you’re kind of insane. This produces a really intense tubey loud effect with very little scratchyness or grind. I’m pretty sure it’ll be kind of brutal on the CPU as it’s not only an Ultrasonic, but also five Console7Channels, each of which run two sine functions.
You might just find it was all worth it, though. Why compress when you can cascade Console7Channels?
############ Console7Crunch shows you Cascade but with its EQ weighted for more edge and grit.