mp3 vs. WAV quiz from NPR Music

[zhimbo]

A pretty nice quiz/demonstration of listening for differences in sound quality for various file types:
http://www.npr.org/sections/therecord/2015/06/02/411473508/how-well-can-you-hear-audio-quality

 

[Natrix]

Listening through my JBL's at work I got 3 out of 6 correct and the 3 I got wrong were all the 320kps MP3's. I can definitely tell the compressed MP3 from the other 2 files, but depending on the music, the WAV and 320kps MP3 are pretty similar to my ears. I found the same to be true listening at home with the Linkwitz CD.

Pretty interesting and thanks for posting.

 

[BoredSysAdmin]

Listed with my AKG K712Pro and AE D1 dac. Except Jayz sample which I could be be forced to listed even if my life depended on it, I didn't got even one right. Once even picked 128mp3.
But then again I already was convinced that 320kbps is indistinguishable from source or my hearing loss is too far gone

 

[zhimbo]

And the aggregate results are...people are just *barely* better than chance. Out of three choices, people picked the uncompressed .WAV file 36% of the time (instead of 33% one would expect by random guessing).

Full story:
http://www.npr.org/sections/therecord/2015/06/09/412271433/audio-quality-quiz-results-you-did-slightly-better-than-guessing-randomly

 

[zhimbo]

Interestingly, out of 3 different times I took the quiz (18 total trials), only once did I ever pick low quality (the Coldplay track, which according to the story was the only track where people picked the low-quality file the most...hmmm), so I'm a bit surprised that low-quality did as well as it did.

I also picked the Mozart symphony correctly 3/3 times, but there's no indication that was a general trend (people did best on the a capella Suzanne Vega track).

 

[Johnny2Bad]

5/6 but it wasn't easy. The particular samples they chose were not, in my opinion. good examples.

 

[zhimbo]

The particular samples they chose were not, in my opinion. good examples.

I think they were good examples in a "broadly representative of general listeners", but not in the sense of "to show the effects of compression". Well-recorded live symphonic music, for example, may show off the benefits of lossless file formats, but not really be relevant for a lot of people.

But, yeah, fully half of these were pretty bad quality sources (the recent pop ones).