UniverSR2026music
High-Frequency Restoration - UniverSR (UniverSR)
Model card for the UniverSR model behind the High-Frequency Restoration - UniverSR preset in Neural Analog.
Up to 9x
Available in Studio. Processing minute usage varies by model option, up to 9x.
UniverSR Description
An upgraded version of AudioSR for music, voice, and sound effects. UniverSR reconstructs missing or poor-sounding high frequencies to add brightness. Try it on dull voices, old recordings (such as cassettes), or AI music from SUNO. It does not change the low frequencies (use an AI cover model instead).
UniverSR Architecture
UniverSR Parameters
| Key | Label | Options | Description |
|---|---|---|---|
| frequency_cutoff | Strength | Subtle (12 kHz cutoff) (12000), Medium (8 kHz cutoff) (8000), Pronounced (6 kHz cutoff) (6000), Strong (4 kHz cutoff) (4000) | No additional help text. |
| reconstruction_method | Reconstruction | Soft Reconstruction (seamless_reconstruction_soft), Continuous Reconstruction (original_signal), Authors' Technique (original) | No additional help text. |
| model_name | Model | Original (universr-audio), Fine-tune v1 (universr-audio-finetune-v1, 9x), Pitbull EDM (pitbull_edm_206, 6x) | No additional help text. |
| stereo_mode | Stereo | Stereo (Mid/Sides) (mid_sides), Sides only (sides_only), Mid only (mid_only), Stereo (L/R) (left_right), Mono (mono) | No additional help text. |