Conflicting MEI bounding boxes in Salzinnes 037v

[dchiller]

Two bounding boxes in Salzinnes 037v are identical and correspond to two syllables with different texts.

The offending bounding boxes are the ones for the syllables on lines 1341 (has text "u") and 1347 (has text "a").

production_mei_files/cdn-hsmu-m2149l4/cdn-hsmu-m2149l4_037v.mei

Lines 1319 to 1354 in 990508b

	<syllable xml:id="syllable-0000000408461961">
	<syl xml:id="m-7c0fee67-340a-48f4-a9cd-9e9a288b4733" facs="#m-bb7f1507-ff9f-4df4-b8f2-d7d948cbe5ea">e</syl>
	<neume xml:id="m-c717013a-a615-4193-91c9-71f5bfbff2ca">
	<nc xml:id="m-a3d5633f-7edb-4e53-bd51-4834817ed4d9" facs="#m-b767f6fd-8ee3-4422-8905-7fe7c21f9105" oct="4" pname="c"/>
	</neume>
	</syllable>
	<syllable xml:id="syllable-0000000096533162">
	<neume xml:id="m-f8f6c626-5b09-4f11-8e16-1fea20ee541d">
	<nc xml:id="m-f201659f-b8f0-4e81-85f4-da8305d9c638" facs="#m-1d467706-3294-46df-8bc2-48fd3168d809" oct="4" pname="c"/>
	</neume>
	<syl xml:id="syl-0000001522003717" facs="#zone-0000002076080015">u</syl>
	</syllable>
	<syllable xml:id="syllable-0000001277423745">
	<neume xml:id="m-19b4f2d6-b223-4457-b564-4a472cff60ee">
	<nc xml:id="m-5d203544-762b-4456-9ec4-fc9b85786158" facs="#m-99b23432-c7d8-4790-b903-acbb848f189c" oct="3" pname="b"/>
	</neume>
	<syl xml:id="syl-0000000912535654" facs="#zone-0000001424196309">o</syl>
	</syllable>
	<syllable xml:id="syllable-0000001260803421">
	<neume xml:id="m-68ff2bc5-2d05-465b-8dc8-52a412fa8e8a">
	<nc xml:id="m-dba73913-335c-4efe-aac3-65ba310cce0c" facs="#m-0d80982c-621c-4756-b877-8e6b4ab753b1" oct="4" pname="c"/>
	</neume>
	<syl xml:id="syl-0000000629762335" facs="#zone-0000000914023447">u</syl>
	</syllable>
	<syllable xml:id="syllable-0000001900591886">
	<neume xml:id="neume-0000000570505371">
	<nc xml:id="m-451dc1e7-2b23-4913-b448-af2a36bfc922" facs="#m-48d2d194-b151-4ee2-b895-65309ee33102" oct="3" pname="a"/>
	</neume>
	<syl xml:id="syl-0000000268901873" facs="#zone-0000001957259248">a</syl>
	</syllable>
	<syllable xml:id="syllable-0000001428452390">
	<neume xml:id="m-0e5a8419-c1c7-4d22-82d7-746ecaf5d89d">
	<nc xml:id="m-3676d872-9275-4b79-a25c-d872323ea8d5" facs="#m-fd41e892-0993-4542-a8a3-b88d78a82205" oct="3" pname="g"/>
	</neume>
	<syl xml:id="syl-0000000406029329" facs="#zone-0000000116389595">e</syl>
	</syllable>

The problem is the evovae in this screenshot (I assume that the four letters here are an abbreviation of that abbreviation!) -- and that either the automated process or the human correctors wanted to add the missing syllables to the mei. But the added "a" was added in a locatioin identical to the shown "u".

[dchiller]

Tagging @annamorphism @JoyfulGen @fujinaga

[JoyfulGen]

@annamorphism's comment in #6 applies here as well:

The bounding boxes are pretty sloppy here generally (the one for E includes the whole text, the others are sort of floating in the staff...) I don't know how these two bounding boxes ended up so perfectly aligned (seems like it would take effort to do?) but cases like this one are a lot easier to "see" (at least in something like Neon) than when the zones are associated with neumes--if the BBoxes are highlighted it's pretty easy to see they need to be moved, especially if syllables are highlighted or if one scrolls over the text. Basically it's not significantly more wrong, or harder to correct, than the "euouae" lower down on the page.

[annamorphism]

@JoyfulGen beat me to the comment! Again, there is another euouae lower down (also missing the last letter) and also with the bounding boxes floating around--but for whatever reason none of the boxes lined up precisely in that case. If we want to be flagging text box weirdnesses we would probably want to flag both these cases. (I suspect most--all?-- of the text box problems will be in the euouae's, because it took a while to decide whether to deal with them as separate syllables or as one entity, what to do about cases where the whole thing isn't written out, etc.)

[dchiller]

thanks for taking a look at these. I agree -- even if the bounding boxes are a little off, it does seem quite suspicious that you'd have the boxes for two different letters/syl's be the exact same, which is what tipped us off to look at this. The good news is that this kind of thing only happened twice in all the folios that have been processed for Einsiedeln + Salzinnes.

So long as the results seem good enough to you (eg. just part of the difficulty of processing an evovae) and aren't evidence of neon doing something weird, that seems fine.