Bring a track you know.
Import a local audio file and choose the codec chain for each side. Both chains start from the same decoded source.
Trust your ears.
Test the difference.
Two codec chains. Your own audio.
A blind listening test that keeps the labels out of the way.
Audivane 0.1.0 is in preparation. Release status
Blind audio comparisons.
The same as A or B.
Can you tell which?
Audio stays on your machine.Your files. Your listening setup. No audio uploads.
Less expectation.
More listening.
Prepare a fair comparison.
Audivane aligns and level-matches the pair, verifies the delivered audio, and stages the first trial paused. You decide when to play.
Listen. Choose. Review.
Switch between A, B and X, then decide whether X is A or B. Complete the planned trials and review the results and preparation evidence.
Understand the method.
Short, exact answers about how Audivane tests — and how to set up, run and read a comparison that means something.
Documentation How blind ABX testing worksWhat an ABX test asks, why the labels stay hidden while you listen, and how Audivane keeps a codec comparison fair.Scientific policyHow Audivane prepares, verifies, blinds and scores a comparison — and precisely what its evidence does and does not establish.Audio formats and limitsAccepted file types, size and duration ceilings, sample-rate and channel policy, and what Audivane refuses rather than shortens.Release statusWhere Audivane 0.1.0 stands — platform readiness, the download formats planned first, and what is not available yet.Licenses and third-party creditsWhich licenses cover Audivane's own code, the FFmpeg and codec libraries, the fonts and the desktop runtime.
Guides Run your first ABX testA step-by-step walkthrough of a full comparison — import audio, configure two chains, prepare, listen blind, and read the evidence.Set up a fair listening testWhat you control around the app — output route, volume, room, material and expectations — so a result means something.Understanding your resultsFixed trial counts, guessing rates, binomial evidence, and the difference between a valid run and an audible difference.
A little more listening
before launch.
Audivane is preparing its first public release. This site is the project's home; documentation, guides, honest release status and — when published — downloads with checksums and provenance all live here.
- Linux desktop
- Tested by the developer. Final release packaging and publication are in preparation.
- Windows desktop
- Pending testing on Windows. No Windows download is available yet.
- Browser demo
- Try the working browser preview. Source and codec materials accompany it; the tagged 0.1.0 release remains in preparation.
Evidence, with its limits.
Audivane records preparation, delivered-audio identity and fixed-trial ABX results. Numerical checks alone do not prove an audible difference or certify a codec. Your hardware, operating-system processing and listening conditions still matter.