Save MusicGen Audio (Legacy)
The old audio saver this pack kept around for compatibility
- audio
- filepath
This is the pack's original audio saver, and the source code says so outright: the docstring is "Simplified audio saver (deprecated - use SaveAudioStandalone instead)". Displayed as "Save MusicGen Audio (Legacy)", MusicGenAudioToFile still works fine - it's just the old shape of a node that got a better version. If you're starting a workflow today, reach for SaveAudioStandalone. If you're loading someone's older workflow or a saved graph that already has this node in it, now you know what it does and why it's there.
What it does
It takes your generated audio and writes it to disk as wav, flac, mp3, or opus. That's the whole job. Underneath it calls the exact same saving function as the newer node - same PyAV encoding, same output/audio/ destination, same timestamp-appended filenames.
The differences from SaveAudioStandalone are all about what got left out:
- filename instead of filename_prefix - one name, no extra prefix concept (a timestamp is still appended).
- No quality control - no bitrate or
V0selection for mp3/opus. You get whatever the encoder defaults to. - format - still the full wav/flac/mp3/opus menu.
Inputs: audio, filename, format. Output: filepath (STRING), the absolute path of the written file. Like the newer node it's an output node, so it ends the queue's run when audio hits disk.
Why it exists
Same pack, two savers, one awkward history. The author shipped the simple version first, then expanded it into SaveAudioStandalone with quality settings and kept the old one around so existing workflows don't break when the pack updates. That's actually decent behavior - plenty of packs would have just deleted it and made your saved graphs throw missing-node errors. The cost is a little confusion when you search "save" and see two nearly identical nodes.
Install and usage
It ships in the same ebrinz/ComfyUI-MusicGen-HF pack as the main generation node, so installing that covers it. Via Manager: search "ComfyUI-MusicGen-HF". Manually:
cd ComfyUI/custom_nodes
git clone https://github.com/ebrinz/ComfyUI-MusicGen-HF
cd ComfyUI-MusicGen-HF
pip install -r requirements.txt
Restart ComfyUI and you'll find it under the audio/musicgen category.
Troubleshooting
There's almost nothing that can go wrong here that isn't shared with the whole pack - missing PyAV means nothing encodes (reinstall requirements.txt), and your files end up in ComfyUI/output/audio/ with the filepath output telling you exactly where. One honest tip: since this node has no quality setting, a 30-second track saved as mp3 is going to be a bit chunkier than the same file from SaveAudioStandalone at 192k. If file size or bitrate matters, that's your cue to switch to the non-legacy node. It's a compatibility leftover, not a trap - just not the one you'd choose first.
Inputs (3)
| Name | Type | Default | Description |
|---|---|---|---|
| audio | AUDIO | — | |
| filename | STRING | musicgen_output | — |
| format | COMBO | wav | 4 options: wav, flac, mp3, opus |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| filepath | STRING | — |