This Space demonstrates the SNAC (Scalable Neural Audio Codec) model (hubertsiuzdak/snac_32khz
).
- Upload an audio file (wav, mp3, flac, etc.).
- The audio will be automatically resampled to 32kHz if needed.
- The 32kHz audio is encoded into discrete codes by SNAC.
- These codes are then decoded back into audio by SNAC.
- You can listen to the original, the 32kHz version (if resampled), and the final reconstructed audio.
Note: Processing happens on the server. Larger files will take longer. If the input is stereo, only the first channel is processed.