Long time ago, I collected some information about durations and no of sample
items:
http://tools.ietf.org/html/draft-hoene-codec-quality-01#page-8
For MUSHRA: In 30 minutes, about 42 stimuli can be tested.
For ACS : Listeners can judge about 54 stimuli within 30 minutes.

That's what I have been told.

ACS is faster but less precise.

With best regards,

 Christian


-----Ursprüngliche Nachricht-----
Von: Paul Coverdale [mailto:[email protected]] 
Gesendet: Mittwoch, 17. April 2013 16:30
An: 'Jean-Marc Valin'
Cc: 'Jean-Marc Valin'; 'Christian Hoene'; [email protected];
[email protected]; [email protected]
Betreff: RE: [codec] Opus comparision test plan

>MUSHRA is actually the least time-consuming...


This isn't what my colleagues in ITU-T and 3GPP, who are heavily involved in
codec subjective testing, are telling me. Typically, you can only test about
12 conditions at a time.

...Paul


_______________________________________________
codec mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/codec

Reply via email to