From e803f24ea090de7108772d65957fd6388d3f2085 Mon Sep 17 00:00:00 2001 From: comfyanonymous <121283862+comfyanonymous@users.noreply.github.com> Date: Sat, 1 Aug 2026 13:38:42 -0700 Subject: [PATCH] Let the VAEDecodeAudio node decode nested audio. (#15211) --- comfy_extras/nodes_audio.py | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/comfy_extras/nodes_audio.py b/comfy_extras/nodes_audio.py index 4ac5ced53..1c6e91fe4 100644 --- a/comfy_extras/nodes_audio.py +++ b/comfy_extras/nodes_audio.py @@ -96,10 +96,14 @@ class VAEEncodeAudio(IO.ComfyNode): def vae_decode_audio(vae, samples, tile=None, overlap=None): + latent = samples["samples"] + if latent.is_nested: + latent = latent.unbind()[-1] + if tile is not None: - audio = vae.decode_tiled(samples["samples"], tile_x=tile, tile_y=tile, overlap=overlap).movedim(-1, 1) + audio = vae.decode_tiled(latent, tile_x=tile, tile_y=tile, overlap=overlap).movedim(-1, 1) else: - audio = vae.decode(samples["samples"]).movedim(-1, 1) + audio = vae.decode(latent).movedim(-1, 1) std = torch.std(audio, dim=[1, 2], keepdim=True) * 5.0 std[std < 1.0] = 1.0