A2 Listening Tests: Now Open to the Community
A2 is ready for your ears.
We've spent months building Architecture 2 (A2), the next generation of Neural Amp Modeler (NAM) with Steve Atkinson. It's designed to sound better, run more efficiently, and work across more devices.
Today, we're opening blind listening tests to the community. Your feedback will help us finalize A2 and ensure it delivers the best possible accuracy before launch.
Why a Blind Listening Test?
Objective metrics like error-signal ratio and spectrogram matching are useful, but they don't tell the whole story. The real test is whether you can hear the difference between a captured tone and the real thing. So we're doing this properly, with a MUSHRA test.
What Is a MUSHRA Test?
MUSHRA is a standardized method used in broadcast and audio research to evaluate perceptual audio quality. Rather than comparing two samples side by side, you rate multiple samples simultaneously on a 0-100 scale. It's the same approach used by organizations like the BBC and EBU to assess audio systems.
How It Works
You'll rate how closely different guitar or bass clips match a Reference sample (the original recording).
Before you begin:
- Find a quiet listening environment
- Use high quality headphones or studio monitors. Avoid laptop speakers, phone speakers, or earbuds.
During the test:
- Each page starts with a Reference sample, followed by several clips to rate
- Use the Reference to set your volume, then keep it consistent throughout
- Rate how closely each clip matches the Reference on a 0-100 scale
- Listen to each clip as many times as needed
- Take breaks to avoid fatigue
What's Next
A2 is launching soon, including retraining all 300K tones on TONE3000 to the new architecture. Stay tuned.
