> As an aside, here is an interesting online study, in which many responden=
ts actually preferred the MP3 versions.
>
> http://archimago.blogspot.ca/2013/02/high-bitrate-mp3-internet-blind-test=
_2.html
Robin,
I've read the comments and have more to add re. blind testing.
As suggested, there may be a familiarity to an "mp3 sound" as with vinyl as=
I have previously mentioned.
The other comment is that there was no control over the listening devices
and the decoding may not have been perfect. What would have been better
would be to encode each sample to lossless and MP3 and then to re-encode
that to an identical lossless format to listen to.
Other factors may be in-ear modulation distortion of the upper frequencies.=
This could be tested by applying a LPF roll-off to each sample.
What would also have been informative would have been a comment on why
certain tracks sounded different.
So saying, it seems like an interesting experiment.
I've tried subtracting an original lossless (apart from Nyquist filtering)
recording with its mp3 equivalent. There are temporal difference in all
frequency bands which the ear is not supposed to hear, but our ears can be=
very clever at times.
David Brinicombe
|