FAQ
Questions and answers
If yours is not here, ask when you request early access and we will answer it directly.
Privacy
Does my audio ever leave my computer?
Not in the current build. ToneTrace is a desktop app: both models run on your own CPU, the library database lives on your disk, and the app makes no network requests.
Local processing is the default and will stay the default. For launch we are evaluating an optional faster processing path for machines that cannot keep up, which would send audio for analysis. If it is offered, it will be opt-in, clearly labelled, and announced together with pricing.
Do I need an account or a licence server?
Not during early access. Install the app and use it. Licensing at launch, and any optional processing options, will be announced together with pricing.
Does the app send usage data or telemetry?
No. The current build collects nothing and sends nothing. This website stores an email address only if you request early access, and analytics on the site, if enabled at all, are cookie-free. See the privacy policy for the full picture.
Files and formats
Which audio formats are supported?
MP3, WAV, FLAC, M4A, AAC, OGG, WMA, AIFF. Decoding is handled by a bundled copy of FFmpeg.
Does it work on long files and stems, or only one-shots?
Both. Analysis is based on the parts of a file that contain sound, so a long stem with a quiet stretch is judged on the music, not the silence.
What happens if I move or rename files after analysing them?
They are recognised by their content. Each file is hashed before analysis, and a file with the same bytes is never analysed twice, whatever it is called or wherever it lives now.
Does ToneTrace detect BPM or musical key?
Not in the current build. ToneTrace tags instrument, sound events, mood, genre and production character.
DAWs and workflow
Which DAWs does it work with?
Any of them, in two ways. Files drag out of ToneTrace as native operating-system file drags, which every DAW, sampler and file browser accepts. And ToneTrace can write its tags into the files themselves as standard metadata (ID3 for MP3, Broadcast Wave for WAV, a JSON sidecar otherwise), so browsers and sample managers that read that metadata will show the tags. No plugin is needed and none is planned to be required.
Does it copy or reorganise my sample folders?
No. Files stay exactly where they are. ToneTrace only reads them, and only writes to them if you explicitly export tags.
Can I make collections or save searches?
Not in the current build. Searching, Find Similar and the folder tree are the ways to move around a library today.
Requirements and pricing
What are the system requirements?
Current expectation, subject to change before launch: Apple Silicon and Intel Macs, Windows 10 or later, and Linux, with about 4 GB of free disk space for the app and its models. Analysis runs on the CPU; no GPU is needed. We will publish tested minimums with the first public build.
How much does it cost?
Early access is free. Pricing will be announced at launch. Early-access testers will hear first, and there is no obligation to buy.
Which models does it use, and are they open?
Two open models: YAMNet from Google, trained on AudioSet, for instruments and sound events, and CLAP from LAION for mood, genre and production character. Both run locally through ONNX Runtime. Licenses and credits are listed on the credits page.
Is ToneTrace a sample marketplace or subscription library?
No. It works on the samples you already own. There is no content store and nothing to download except the app.
Still have a question?
Ask it when you request early access and we will answer directly.