Skip to content

Update num_frames output for LongCatAvatarWhisperEmbeds #2020

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
drphero wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Update num_frames output for LongCatAvatarWhisperEmbeds #2020

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e30028c
Update num_frames output for LongCatAvatarWhisperEmbeds
drphero May 24, 2026
ddecf6c
Update log message
drphero May 24, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
28 changes: 26 additions & 2 deletions LongCat/nodes.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -295,7 +295,31 @@ def linear_interp(features, output_len):
offset += w.shape[-1]
out_audio = {"waveform": mixed, "sample_rate": sr}

return (multitalk_embeds, out_audio, num_frames)
# Calculate actual frames based on audio duration
actual_num_frames = num_frames
if len(audio_outputs) > 0:
if multi_audio_type == "para":
# For parallel mode, use the longest audio duration
max_audio_duration = max([ao["waveform"].shape[-1] / sr for ao in audio_outputs])
actual_frames_from_audio = int(max_audio_duration * fps)
else: # "add"
# For sequential mode, use the total audio duration
total_audio_duration = sum([ao["waveform"].shape[-1] / sr for ao in audio_outputs])
actual_frames_from_audio = int(total_audio_duration * fps)

# Use the smaller of requested frames or actual audio frames
actual_num_frames = min(num_frames, actual_frames_from_audio)

if actual_frames_from_audio < num_frames:
log.info(f"[LongCat] Audio duration ({actual_frames_from_audio} frames) is shorter than requested ({num_frames} frames). Using {actual_num_frames} frames.")

# Debug: log final mixed audio length and mode
total_samples_raw = sum([ao["waveform"].shape[-1] for ao in audio_outputs])
log.info(f"[LongCat] total raw duration = {total_samples_raw/sr:.3f}s")
log.info(f"[LongCat] multi_audio_type={multi_audio_type} | final waveform shape={out_audio['waveform'].shape} | length={out_audio['waveform'].shape[-1]} samples | seconds={out_audio['waveform'].shape[-1]/sr:.3f}s (expected {'sum' if multi_audio_type=='add' else 'max'} of raw)")


return (multitalk_embeds, out_audio, actual_num_frames)


NODE_CLASS_MAPPINGS = {
Expand All @@ -305,4 +329,4 @@ def linear_interp(features, output_len):
NODE_DISPLAY_NAME_MAPPINGS = {
"WanVideoLongCatAvatarExtendEmbeds": "WanVideo LongCat Avatar Extend Embeds",
"LongCatAvatarWhisperEmbeds": "LongCat Avatar Whisper Embeds (v1.5)",
}
}