"Jim Lesurf" wrote in message
...
In article , Michael Chare
wrote:
"Jim Lesurf" wrote in message
...

There are a few snags with the 'test' reported above.

I suppose my question really was can you hear a difference between the
two formats?

Personally, as yet I've not made a serious systematic attempt to do so
since I'm not sure the answer would be of much use unless you add a number
of qualifying conditions to the above sweeping question about 'formats'.
Hasn't been a primary interest to me since I essentially don't use mp3
myself. Some other reasons outlined below...

Firstly it seems quite likely that the choice of bitrate would affect how
audible 'effects' might be.

Secondly, it would probably depend on other settings used when encoding.
If
you look at lame you can find a list of these as long as yer arm...

Thirdly, it would probably depend on the type of music. And how familiar I
was with it. And the replay equipment and listening situation. And no
doubt
other things which could be listed as they occurred to people!

Finally, it would depend on if simply listening to an mp3 with no access
to
the source from which it was made, or if it was possible to do
comparisons.
One aspect of such lossy schemes is that they may seem OK if you can't do
a
direct comparison to detect alterations which - in isolation - may seem
OK.
Similarly you can sometimes 'train' yourself so you can hear effects that
initially passed you by e.g. years ago learning how - at the time - to
hear
what speaker cones were made from from the 'quack' or 'paper cup' effects.
Alas, doing that can spoil listening to music if you are still stuck with
using such things, so I don't advise people to try such things unless they
want to risk becoming distracted. :-)

So trying to come to sweeping conclusions about the 'formats' may fall
foul
of intepreting indivual examples and deciding how 'typical' they are of
what the 'format' could do if optimised for the specific cases. And of
then
generalising from one person to another. Having a few people listen to a
few files isn't much of a basis for that, let alone just one person, if
you
expect the results to mean much to other people in other cases.

Given that I've not used mp3 for much listening I can't say trying to a
sweeping comparison has attracted me much. Particularly as I can't make
any
particular claims for having 'golden ears' so my findings may well not
tell
others much about what they might be able to hear that I might miss.

However I've certainly heard audible problems at times. Even for 256kbps
mp3 on occasion. And it is easy enough to show measurable differences.

I've also found measurable differenced between *decoders* playing the same
mp3 source. But have no idea which might be judged 'better' or 'closer to
the original' in any general sweeping terms due to all the other variables
of the kind I mention above.

Hence at present I'm quite happy to accept that some people can hear some
problems or differences that may not be noticed by others, and this varies
from person to person, file to file, etc, etc. Beyond that I have no real
idea at present. Not done the experiment. Other things have been more
interesting. :-)

FWIW I have been looking at this from a more measured perspective and have
some results which may appear sometime. But converting that into deciding
what it means in audible terms is not simple for the above reasons.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics
http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html

When MP3 perceptual codecs first became popular, I made up a couple of CDs
each with the same track firstly as a straight copy off a commercial CD,
then MP2 and MP3 encoded at different bit rates, some joint stereo, some
true stereo. I was surprised at how good they all actually were, down to
96kbps even MP2. My experience was that doing an AB comparison with the
original, it was fairly clear which was which up to something like (it was
some time ago, and I no longer have my notes) 192kbps MP2 and 128k MP3, but
after that, it became increasingly difficult to tell. However, what I found
most surprising, given all the hype about data-compressed audio in the HiFi
press at the time, was that in the absence of an AB comparison with the
original, how all of them, even the lowest bit rates were perfectly
acceptable for non-critical listening, and weren't obviously flawed.

When I got to above 256kbps for both MP2 and MP3, I couldn't hear any
difference with the original, on the two or three pieces of music I'd
chosen, however carefully I listened, and even knowing which was which, I
couldn't hear a difference. Maybe that says something about my ears, but it
points that TO ME, these algorithms do what they're supposed to do. I don't
often use MP3s as I don't use portable music players, but on my holiday
system, which consists of a laptop with a 500Gb drive and a Digigram sound
card, I have several hundred CDs ripped as 320kbps MP3s, and am not aware of
any lack of sound quality.

S.

S.