Preview override: Add LTXAV audio preview synced to video - #640
Open
drozbay wants to merge 1 commit into
Open
Conversation
Adds an optional audio_vae input to Model Preview Override. On LTXAV (audio+video) models the audio latent is decoded each step and played back in sync with the video preview via the shared playback clock, improving in quality as sampling proceeds. Decode runs directly on first_stage_model, pinned to GPU for the run (the same device-handling the node uses for the TAEHV previewer), bypassing comfy VAE.decode()'s per-step load_models_gpu. Audio is muted unless the cursor is over the preview frame or the step plot. No behavior change unless audio_vae is wired.
drozbay
force-pushed
the
feature/ltxav-audio-preview
branch
from
May 24, 2026 01:48
935c0b6 to
928189a
Compare
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds an optional
audio_vaeinput to Model Preview Override. On LTXAV (audio+video) models the separated audio latent is decoded each sampling step and played back in sync with the video preview, so the audio sharpens alongside the video as denoising proceeds.The
<audio>element rides the same playback clock and double-buffer/seek pattern as the existing MP4 path (showMp4), so pausing, scrubbing, and per-step lock work for audio too. The audio VAE, which is relatively small, runs directly onfirst_stage_model, pinned to GPU for the run (the same device-handling the node already uses for the TAEHV previewer),No behavior change unless
audio_vaeis wired and the audio is muted unless hovering over the relevant UI elements.