Audio Mono to Stereo

Audio Mono to Stereo does exactly one thing: takes a mono AUDIO clip and duplicates the single channel to the left and right. That's the whole node, and it earns its place because a lot of AI video models - including several LTX and Wan audio paths - hand you mono sound, and mono can sound thin or wrong when you mux it into a stereo video or play it back where stereo is expected.

It's part of the ComfyUI-FBnodes pack from François Beaudry, the same grab bag that brought you Save Video+. Like most of the pack it's a quiet utility: no settings to mis-tune, no model to download, one socket in, one socket out.

How it works

The node inspects the incoming AUDIO dictionary's waveform tensor. If it already has two or more channels, it passes the audio through untouched - stereo in, stereo out, no harm done. If it's genuinely mono (one channel), it repeats that channel into a second one, producing a (batch, 2, samples) waveform at the same sample rate, and hands it back as a normal AUDIO object.

In other words: it's the "duplicate the channel" conversion, not a fake-stereo widening or a phase trick. Same signal in both ears, just now with both channels populated. That's the honest, lossless fix for the actual problem - downstream muxers and editors that balk at mono tracks - and it won't do anything for audio that's mono because it was supposed to be.

The one input

• audio (AUDIO) - wire whatever's producing the mono audio: an LTX audio latent decoded through its audio VAE, a Wan audio path, a video's extracted track.

Output is an AUDIO that slots back into any node that expects stereo, like Save Video+'s audio muxing or an audio preview.

Installing it

Search "ComfyUI-FBnodes" in ComfyUI Manager, or:

cd ComfyUI/custom_nodes
git clone https://github.com/FranckyB/ComfyUI-FBnodes.git
pip install -r ComfyUI-FBnodes/requirements.txt

Restart ComfyUI. The pack depends on av (PyAV) and color-matcher; no model files, this is pure waveform math.

Real-world placement

Where people actually stick this: right before Save Video+ in an LTX-2 workflow when the decoded audio comes back mono, or between Get Video Components+ and an audio preview. One honest caveat - if your audio is already stereo, this node is a no-op, so don't treat it as a fix for quiet or muffled audio. That's a different problem (usually the model's audio quality, not the channel count). But if you see one channel in your waveform and hear only one side, this is the two-second fix.