Good points. However, I would submit that an ABX test situation is pretty un-natural to begin with, and will (1) likely require critical listening (2) look for consistently detected *differences* between A and B. As for non-critical listening, the proof is in the market pudding --if low-bitrate MP3s were so bad, they would not be so ubiquitous. (Same as with compact cassettes, VHS, or telephone receivers) However, I agree with you that using the umbrella term "MP3" (or "LP" or "tape" for that matter) is not very useful --the bit-rate is extremely important. Which is why I prefer "extreme" tests, which test the limits of both the format and our perception.
Glad you enjoyed the sources!
Marcos Sueiro Bal
Senior Archivist, New York Public Radio
646 829 4063
From: Association for Recorded Sound Discussion List [mailto:[log in to unmask]] On Behalf Of Eli Bildirici
Sent: Tuesday, May 16, 2017 4:19 PM
To: [log in to unmask]
Subject: Re: [ARSCLIST] Mp3 is officially dead
These are good critiques - the sample size also struck me right away - though it is not the only study to come to such conclusions (I don't have them handy unfortunately). Phrasing the question as you did though sounds problematic - if people are asked to pick out artifacts it might put them in a 'critical' frame of mind that could skew results somehow, right? I don't doubt people would begin 'hearing things' in pristine, uncompressed sources.
These resources seem pretty cool though - looking forward to perusing them, thanks for sharing!