R.
Audio restoration and stem separation
It can bring back clarity without making everything sound too artificial or overprocessed. Stem separation is also surprisingly clean.
Upload audio
Drag & drop or click to browse (multiple files supported)
Extract stems using SAM Audio when you need to describe a custom sound. For vocals, drums, bass, and other common parts, compare specialized presets such as BS Roformer, MelBand Roformer and MDX23. Every mix is different, so listen to the examples before you choose a pipeline.
These generations were shared by Neural Analog users
| Model | SAM Audio net win rate | Separation score |
|---|---|---|
| SAM Audio | Reference | 100.0 |
| LALALAI | 8.4% | 91.6 |
| AUDIOSHAKE | 11.4% | 88.6 |
| DEMUCS | 17.6% | 82.4 |
| MOISESAI | 19.5% | 80.5 |
| FADR | 28.3% | 71.7 |
| SPLEETER | 52.8% | 47.2 |
| CLAPSEP | 92.4% | 7.6 |
| FLOWSEP | 93.2% | 6.8 |
| SOLOAUDIO | 95.4% | 4.6 |
| AUDIOSEP | 98.1% | 1.9 |
R.
Audio restoration and stem separation
It can bring back clarity without making everything sound too artificial or overprocessed. Stem separation is also surprisingly clean.
N.
MP3 and CD audio restoration
Does amazing job in restoring Audio from bad quality MP3s and badly extracted audio from CDs. Always my go to for audio restoration.
Amazing tool for audio. Clean, simple, and effective. I would spend hours in RX to get the same results. Give it a try. I can save you hours of production time.
M.
Restoration and mastering
The flash upscaler mixed with the composite mix system makes it super easy to get closer to a song master that I'm happy with
I've been loving the app. The audio restoration works amazing on my suno songs
IMDK
I love the interface. I love the bulk upload/download features. They're a life saver! Also I just realized that Apollo is magic and I don't even need to use denoise. Apollo somehow removes noise much more naturally. So I'm actually spending way less credits than I expected.
Eve St. Jude
Production workflow
Neural Analog has been a godsend to me and my workflow, such a game changer. I'm more thankful than words can express, I don't know what I would have done without it.
Eve
amazing!!!! love this! THANK YOU! I just discovered batch dl, this feature SAVED ME
Wow thank you so much i upscale videos and take out live recordings from music because of my autism i hate the crowd
It worked! Well done :) Many thanks :))))
Bjark
Easy to use and high quality results.
AI audio enhanced with Neural Analog
Stem separation can expose compression damage and missing high-frequency detail. Compare the stem with a restored version before you export.


Suno offers built-in stem extraction for supported songs. Neural Analog adds prompt-based SAM Audio separation, so you can describe and isolate a specific sound, such as "saxophone," "electric piano," "synth lead," or crowd noise. Start with Suno Stems when you need a downloadable stem or a custom cleanup pipeline.
The public SAM Audio playground is for short mono experiments. Neural Analog puts SAM Audio inside Studio so you can import full songs, preserve stereo pipelines, select source regions, and download the resulting stems.
Start from SAM Audio Online if you want prompt-based extraction, or use the general stem splitter for standard vocals, drums, bass, and custom prompts.
Import the Suno track, choose a custom stem prompt, and describe the target sound, for example "acoustic guitar," "violin," or "synthesizer." SAM Audio uses that prompt to separate the matching sound from the mix.
Start from Suno Stems or use the general SAM Audio Online pipeline.
AI stem separation can produce clean vocals, drums, bass, and instrument stems, but it cannot recreate the original studio tracks perfectly. Sounds that overlap in the same range can leave bleed or artifacts.
For the best result, use AI Stem Splitter presets for common stems and SAM Audio custom prompts for unusual instruments or sound layers. After splitting, you can restore or master individual stems in Studio.
Piano extraction example
Original mix
SAM Audio piano stem
Processing time depends on the operation, selected model, input length, and queue load. Short jobs often finish in about a minute, while long tracks, batch jobs, and heavier models such as SAM Audio Large or UniverSR can take several minutes or more.
You can leave the page and come back later. Finished imports, restorations, masters, and stems stay in Studio, where you can reopen the project and download the result.
Click the Export button, then choose Download all current stems.
This downloads the stems currently visible in your project, using whatever version is selected for each stem. For example, if your vocal stem is set to a restored version and your bass stem is set to the original version, the export will include those current versions.
To change what gets exported, first select the version you want on each stem waveform, then click Export again.
If you need to create stems first, use the Universal Stem Splitter or open the project in Studio.
Paste the link. Choose or describe a stem. Preview both files before you download.