MP3 Quality

OK, April 1 joke :)

Thought more people would fall for it...

To my ears there is a clear difference between Lossless and compressed - but you do need high end equipment to hear it.  A decent pair of headphones and an amp/DAC should do the trick

As a genuinely polite enquiry, have you done the comparison as an ABX test* (ideally double-blind), or did you know to which format you were listening?  

If the latter, you may well be hearing differences that aren't there (not your 'fault', of course, just the way the human mind works).

* see, eg, http://en.wikipedia.org/wiki/ABX_test

Possibly true - but not actually relevant, since it's not the artistic interpretation that is the subject of the comparison, but the 'detail' of the sound. 

An analogy might be comparing versions of a painting.

If I wanted to compare their artistic quality and then overall interpretation of the subject matter, I'd want to look at the whole painting separately in each case.

But if I wanted to compare the detail of brush strokes, or the texture of the paint, I'd want to look at very small parts of them under a comparison microscope, flicking between the two pretty often to detect the differences. 

I obviously respect your professional experience, but would be interested in how you overcome the inevitable psychological bias of knowing what the source is. 

"Through a mobile phone while you are jogging" seems a bit of a straw man/red herring.  Plenty enough people genuinely feel high bitrate mp3 is quite good enough (to the point of being indistinguishably from the original) to listen "seriously".

Incidentally, I should point out that I'm indulging in no small amount of devil's advocacy here. 

Whilst I generally find - speaking personally - high bitrate mp3 indistinguishable from the source (maybe due to a deterioration in my hearing with age, maybe not), I also have a large amount of music as lossless files (flac), and I'd be very loathe to convert them into mp3s! Irrationally or otherwise, and no matter how convinced I may be that I can't tell the difference, I feel I'd be missing something :-)

Best regards

No - too busy re-ripping my CDs to Lossless to have the time ;)

When you do A/B/X you don't know which is which, that is supposed to be the deal.  I am far from alone among professionals in being very unhappy with overdependence on A/B/X which has two huge problems.  First, it bears no resemblance to how one listens to music and in this instance we are not evaluating audio we are evaluating how faithfully music is communicated.  Second there is more than one mathematical solution to the equation which is produced when the difference is deemed minimal or non-existent.  The difference may indeed be minimal or non-existent, equally the test and its procedure might be significantly flawed to have made the perception of any difference impossible or severely masked.

Living with uncompressed high resolution audio as people like me do makes the proposition that mp3 would be indistinguishable signify one of several other far more obvious explanations - inadequate source material, inadequate reproduction equipment, or possibly some kind of substance dependence on the part of the listener(s) during the listening process. As it happens, as I am typing this email, our workstation is encoding an 88k2/24 recording of Bach flute sonatas into 44k1 mp3's to email to the artists for a quick listen.  I don't know exactly how many glasses of wine it would require for me to be unable to tell the difference between the mp3 and the original, but it would be enough to get me banned from driving if I even sat in the car with the ignition key in the lock, if I could get to the car without falling over on the way.

This sort of discussion is a red rag to a bull as is talking cheap compact digital cameras with plastic lenses with a serious photographer.  Audio has always been available in various dilutions and distortions to suit listening preferences whether cassette, AM radio, DAB, LP, CD or whatever.  We live with the loss in quality in exchange for convenience.  Only a nutcase would have said that a cassette or AM radio were equal to master quality.  He/she might well say they are happy with the compromises which do not bother them in context, and I have no problem with that. 

Mp3 screws up the sound, and there is much more than anecdotal evidence to confirm it.  The most surprising loss of resolution for me is not just the acidity, squashed nature of the sound and muzzy stereo detail, it is the fact that within the encoding time-window for each block of data the audio gets moved around in time.  That maybe sounds like mumbo-jumbo, but the upshot is that I had a long to&fro with a cello/piano duo who rejected a first edit of Beethoven Cello Sonatas on the basis of 128kbps mp3 files we emailed to them.  They kept saying that various moments in the music were not together between the players and I must do extra editing using alternative takes.  When I checked the original material here I could hear no problem.  When I checked the mp3 encoded files they were not together.  In the end I had to Fedex CDs to the artists and wait for revised comments.

With 'pre-digital' media the losses were basic curtailment of frequency response, addition of noise, maybe some dynamic compression of high sound levels, maybe some wow&flutter.  Digital lossy compression does all sorts of mathematics with filters and masking which produce different side-effects which are not so easy to measure - not because the differences are near non-existent, but because we are measuring them in the way we would have measured the old-style distortions of past analogue carriers and are missing the wood for the trees.

Yes, but the vast majority of us are not (nor are we capable of) listening to 88/24 recordings.  As I've said elsewhere, the most realistic music I've heard recently was a 24/192 recording, and I have no doubt that realism was in no small part due to the high-resolution nature of the source, but my contention is that the difference between a nicely-ripped 320k AAC file and that of the original 16/44 red book is very small indeed, and unlikely to trouble the majority of us mere mortals who have merely mortal stereos.