Mp3 v CD no diff at 320

There have been plenty of debates here about this over the past couple of years but here's a pretty interesting (and seemingly solidly scientific) take on the whole Is compression audible furore over MP3s. The piece is in the May edition of Hi-Fi Choice magazine and is by James Clark who spent many years working at Arcam and now with Bluetooth technology.

He says: "Technology has come to our aid in the form of ITU recommendation BS.137. This describes a computer-based analysis model. It's like this: model human ears on a computer, so allowing the computer to determine how real people would rate the quality of an audio sample. If I give it the right software (I use 'Opera' from German company Opticom) the computer can do the listening tests for me. Opera claims to correlate at 98 per cent with the results from large groups of human listeners.
I sat down in front of my computer with a pile of CDs, my favourite MP3 compression algortihm and a big mug of coffee. We worked out way through several different music genres, a dozen different bit rates and a significant chunk of my hard disk before we came up with the answer. Not the definitive answer you understand for that would depend on the exact listener, but a good guideline.
And the answer, as it turns out, is 224. That's to say a very large majority of people will be unable to to tell the difference between audio compressed in MP3 at 224kbps and the original CD, no matter what type of music is used. Also, unless you're blessed with 'golden ears', almost nobody can tell the difference at 320kbps after all!"

So there you have it. There is no audible difference. I'm no fanatical debunker of hi-fi snobbery (though I do find it pointlessly elitist, wilfully esoteric and self-deluding in direct proportion to size of bank account) but this seems to be some kind of proof that all the bluster about MP3 being the devil's work is rubbish and that after all the posturing from the hi-fi snobs that it's alright for "commoners" but the elite know better, the elite have been shown to be like the emperor, parading around in their birthday suits.

Read it again: "a very large majority of people will be unable to to tell the difference between audio compressed in MP3 at 224kbps and the original CD"
A very large majority. So that's probably between 60 and 80 per cent then.

And:
"unless you're blessed with 'golden ears', almost nobody can tell the difference at 320kbps"
Almost nobody. And honestly, genuinely, how many actually have those "golden ears"? Very, very few I would imagine.

I've always found that burned CDs, high bit rate MP3s and original CDs sound exactly the same. Even through my i-pod and cheap Sennheiser in-ears the music sounds, rich full and informative. I honestly cannot tell the difference between 320 and CD.

In fact the only format I've come across that slays 'em all is vinyl. I'm listening to some Vangelis now and it sounds better than any CD ever will.

But if you must go digital, it seems that any way (as long as the bit rate is high enough) will do - and ignore the snobs who tell you different.




Edited by arcer - April 16 2007 at 18:16

arcer wrote: In fact the only format I've come across that slays 'em all is vinyl. I'm listening to some Vangelis now and it sounds better than any CD ever will.

Sorry, but IMO that's pure nostalgia. Technically the CD is superior ... which doesn't mean that we should stop listening to vinyl - in fact I even bought a record player last year.

BTW: Heavyfreight posted an interesting link a while ago ... it explains in detail why vinyl suffers from many of the problems of "numeric" digital audio. I'll see if I can "dig it up".

The sound of vinyl is different.

Certif1ed wrote: The sound of vinyl is different.   The only superiority that CD holds over vinyl is in terms of accuracy - and accuracy is not as essential or even 100% desirable. I thought that the ultimate goal of audiophiles was to preserve the original performance as accurately as possible.   Much of the "feeling" in music, particularly 1960s-1970s music comes from the noise in the equipment used to record the music, and the noise introduced by the reproduction system (the Hi-Fi) - hence vinyl from this time actually can sound better - strip away the noise by digitally remastering and you change the sound. The recorded noise is usually preserved on the remaster ... I agree that it wouldn't make much sense (nor would it be in the audiophile spirit) to try to make it sound like it was recorded in the 1990s. Only the limitations of the vinyl are removed ... and that indeed changes the sound ... for better IMO, as it is closer to the original. In the 1970s many vinyl mixes/masters were optimized for the equipment that was used to play them ... when Olivier talks about "overbumped lows" for example it may be due to the "underbumped" mixes of the 70s.    Ultimately, it's in the ears of the beholder - digital music is more accurate, but not all CDs are created equal, and some really do sound worse; e.g. early Genesis CDs sound awful in comparison to their first press vinyl counterparts. Many early CD releases of vinyl classics were simply copies of the vinyl mixes ... and thus suffered from bad dynamic resolution. But I think that when you talk about "vinyl vs. CD" in general you should compare the best vinyl mixes to the best CD mixes.   It's the same with encoding - 320Kbs is a slow bitrate compared with raw audio, and will necessarily lose data (since mp3 is a lossy compression). But your ears might not be able to tell the difference. Even vinyl suffers from "numeric" data loss ... the polymer molecule clusters which the vinyl disc consists of are larger than one would think (I really have to dig up Heavyfreight's post which explains it in detail). Even with analog tapes you have single particles which store the information ... *everything* is numeric, just with different resolutions.   However, ears are not created the same either - when a computer monitor goes on the blink and emits a very high-pitched whistle, I can actually hear it (and identify the pitch), while others in my office (who are mostly younger, but less exposed to music) cannot. Me too. My ears are not as good as they were 10 years ago, but I can still hear very high pitch. I also detect "jitter" during digital playback ... earlier today I listened to music on the computer and it sounded a bit strange ... I checked the Task Manager and saw a process utilizing the computer I/O so much that it slightly affected the playback (causing slight "harshness" in high pitched waveforms like violins).   As I said - ears of the beholder (but vinyl IS better ). I like both ... I keep vinyls for nostalgia, but prefer CDs most of the time.

320 is VERY high.  As I've said before, 128's sound the same as CDs for me.  

Certif1ed wrote: MikeEnRegalia wrote: Certif1ed wrote: The sound of vinyl is different. Much of the "feeling" in music, particularly 1960s-1970s music comes from the noise in the equipment used to record the music, and the noise introduced by the reproduction system (the Hi-Fi) - hence vinyl from this time actually can sound better - strip away the noise by digitally remastering and you change the sound.   To clarify - sometimes the colouration introduced by the HiFi can be pleasing to the ears - and with digital sources it's a different ballgame. When digital distorts, it's just nasty - but when analogue distorts (as most low-mid end HiFis do), it can enhance the sound. Arguably... The recorded noise is usually preserved on the remaster ... I agree that it wouldn't make much sense (nor would it be in the audiophile spirit) to try to make it sound like it was recorded in the 1990s. Only the limitations of the vinyl are removed ... and that indeed changes the sound ... for better IMO, as it is closer to the original. In the 1970s many vinyl mixes/masters were optimized for the equipment that was used to play them ... when Olivier talks about "overbumped lows" for example it may be due to the "underbumped" mixes of the 70s.  I don't know if it still is, but for a while, it was common practice to remove "white noise" and possibly other forms of noise in order to "clean up" the recordings during the remastering process - some noise is not nicely reproduced by digital systems.   Audiophile remastering is different again. 1. I know that good tube amps create harmonic distortion which can be pleasing to the ear (usually described as "warmth"), and for people who are used to this effect a good digital amp can sound cold and maybe even "harsh". Nevertheless the added warmth is something which true audiophiles should not want ... the original signal doesn't have that warmth, so the digital amp is a more accurate reproduction of the original. And clipping (digital distortion) doesn't usually occur when you use a digital amp properly, so it's hardly a point against digital amps. 2. I agree that remasters/mixes which are heavily filtered or compressed are bad ... especially the latter is a crime against music IMO.   Ultimately, it's in the ears of the beholder - digital music is more accurate, but not all CDs are created equal, and some really do sound worse; e.g. early Genesis CDs sound awful in comparison to their first press vinyl counterparts. Many early CD releases of vinyl classics were simply copies of the vinyl mixes ... and thus suffered from bad dynamic resolution. But I think that when you talk about "vinyl vs. CD" in general you should compare the best vinyl mixes to the best CD mixes.   Try comparing the first vinyl press of DSOTM with any CD mix.     Led Zep II is a cheaper alternative (about £50 for a reasonable 1st press - don't pay more!).   Thanks for the recommendation ... how do I tell these first presses from re-issues? It's the same with encoding - 320Kbs is a slow bitrate compared with raw audio, and will necessarily lose data (since mp3 is a lossy compression). But your ears might not be able to tell the difference. Even vinyl suffers from "numeric" data loss ... the polymer molecule clusters which the vinyl disc consists of are larger than one would think (I really have to dig up Heavyfreight's post which explains it in detail). Even with analog tapes you have single particles which store the information ... *everything* is numeric, just with different resolutions. Not all vinyl is created equally... Audiophile pressings are made from higher quality vinyl, as well as being cut at half speed in order to increase the amount  - the attention to detail is incredible. I'm sure Heavyfreight knows what he's talking about - but there are still audiophile recording engineers who make stunning quality products, and whatever the physical facts are, my ears are the final judge (either that, or I'm just a hopeless vinylholic who simply doesn't care!). I have a few 180gr vinyls, and I also examined some audiophile vinyls in the record store ... they sure look (and feel) impressive. Ultimately I think that they're simply too fragile, and the audio quality depends much on the playback equipment. And even if you have a top record player, the vinyl disc degrades each time you listen to the album ... even if you know how to handle vinyl discs. And lastly the "attention to detail" is seriously limited compared to the CD both in terms of dynamic resolution and frequency range. Heavyfreight explains in detail why even high quality vinyl discs/players cannot reproduce frequencies higher than ~18khz, and how the "granularity" of the vinyl material introduces a quantisation effect quite similar to the digital "numeric" quantisation near the Nuquist frequency (20.5khz). I don't want to question your judgement or your ears ... I'm just very sure that the CD is better suited to accurately reproduce a recording than the vinyl disc. Of course even when I'm convinced that CD "is better" it doesn't mean that all CD remasters are superior to the vinyl editions. However, ears are not created the same either - when a computer monitor goes on the blink and emits a very high-pitched whistle, I can actually hear it (and identify the pitch), while others in my office (who are mostly younger, but less exposed to music) cannot. Me too. My ears are not as good as they were 10 years ago, but I can still hear very high pitch. I also detect "jitter" during digital playback ... earlier today I listened to music on the computer and it sounded a bit strange ... I checked the Task Manager and saw a process utilizing the computer I/O so much that it slightly affected the playback (causing slight "harshness" in high pitched waveforms like violins).   Isn't that the effect caused by digital quantisation? I read somewhere what the difference in sound was between jitter and quantisation - but both are products of digital systems. Oddly, my ears actually feel like they're more sensitive than they were 10 years ago - maybe it's the constant battering my headphones give them, coupled with advanced tinnitus... Jitter usually means that samples are skipped ... the system fails to reproduce the samples of the original signal in the same sequence and timing. Mild jitter sounds really awful and can be quite "sublime" ... you just know that something's wrong with the signal - sounds "unstable" and "dirty". Quantisation simply means loss of detail due to low sampling rate ... high pitch sounds will lose detail or simply disappear. As I said - ears of the beholder (but vinyl IS better ). I like both ... I keep vinyls for nostalgia, but prefer CDs most of the time. Track down a UK "Plum and Orange" copy of Led Zep II (you should find one for £10 or thereabouts). First pressings are best, obviously (the £50 and up copies - the really expensive ones are the mispressings - don't bother with those, they don't sound any different!) - but all the plum Zep albums sound amazing. If you still think the CD sounds better, then I don't know what else to say I know what Olivier would say to me in this case: "Buy a better system". First he would blame the record player, then the cable connecting record player and amp, then the amp, then the speakers, then the speaker cables, then the amp power cables ... if all else fails, it's the room!

Depending on the source and the decoder used, lower bitrate mp3's can sound either very good or extremely nasty.

Personally, I always rip to FLAC. Sometimes I don't hear a significant difference between a 192 kbps mp3 and a FLAC, but I still know that there's data missing from the mp3, and that doesn't feel right to me. I want to be sure that I have exactly the same music on the computer, because it gives me peace of mind so to speak, and because I can replace the original CD with an exact copy in case something would happen to it.

arcer wrote: There have been plenty of debates here about this over the past couple of years but here's a pretty interesting (and seemingly solidly scientific) take on the whole Is compression audible furore over MP3s. The piece is in the May edition of Hi-Fi Choice magazine and is by James Clark who spent many years working at Arcam and now with Bluetooth technology.He says: "Technology has come to our aid in the form of ITU recommendation BS.137. This describes a computer-based analysis model. It's like this: model human ears on a computer, so allowing the computer to determine how real people would rate the quality of an audio sample. If I give it the right software (I use 'Opera' from German company Opticom) the computer can do the listening tests for me. Opera claims to correlate at 98 per cent with the results from large groups of human listeners.I sat down in front of my computer with a pile of CDs, my favourite MP3 compression algortihm and a big mug of coffee. We worked out way through several different music genres, a dozen different bit rates and a significant chunk of my hard disk before we came up with the answer. Not the definitive answer you understand for that would depend on the exact listener, but a good guideline.And the answer, as it turns out, is 224. That's to say a very large majority of people will be unable to to tell the difference between audio compressed in MP3 at 224kbps and the original CD, no matter what type of music is used. Also, unless you're blessed with 'golden ears', almost nobody can tell the difference at 320kbps after all!"So there you have it. There is no audible difference. I'm no fanatical debunker of hi-fi snobbery (though I do find it pointlessly elitist, wilfully esoteric and self-deluding in direct proportion to size of bank account) but this seems to be some kind of proof that all the bluster about MP3 being the devil's work is rubbish and that after all the posturing from the hi-fi snobs that it's alright for "commoners" but the elite know better, the elite have been shown to be like the emperor, parading around in their birthday suits. Read it again: "a very large majority of people will be unable to to tell the difference between audio compressed in MP3 at 224kbps and the original CD" A very large majority. So that's probably between 60 and 80 per cent then.And:"unless you're blessed with 'golden ears', almost nobody can tell the difference at 320kbps"Almost nobody. And honestly, genuinely, how many actually have those "golden ears"? Very, very few I would imagine. I've always found that burned CDs, high bit rate MP3s and original CDs sound exactly the same. Even through my i-pod and cheap Sennheiser in-ears the music sounds, rich full and informative. I honestly cannot tell the difference between 320 and CD. In fact the only format I've come across that slays 'em all is vinyl. I'm listening to some Vangelis now and it sounds better than any CD ever will.But if you must go digital, it seems that any way (as long as the bit rate is high enough) will do - and ignore the snobs who tell you different.

And BTW, if you had a working system, you'd realize that a computer-burned Cd sounds awful.

Snow Dog wrote: oliverstoned wrote: arcer wrote: There have been plenty of debates here about this over the past couple of years but here's a pretty interesting (and seemingly solidly scientific) take on the whole Is compression audible furore over MP3s. The piece is in the May edition of Hi-Fi Choice magazine and is by James Clark who spent many years working at Arcam and now with Bluetooth technology.He says: "Technology has come to our aid in the form of ITU recommendation BS.137. This describes a computer-based analysis model. It's like this: model human ears on a computer, so allowing the computer to determine how real people would rate the quality of an audio sample. If I give it the right software (I use 'Opera' from German company Opticom) the computer can do the listening tests for me. Opera claims to correlate at 98 per cent with the results from large groups of human listeners.I sat down in front of my computer with a pile of CDs, my favourite MP3 compression algortihm and a big mug of coffee. We worked out way through several different music genres, a dozen different bit rates and a significant chunk of my hard disk before we came up with the answer. Not the definitive answer you understand for that would depend on the exact listener, but a good guideline.And the answer, as it turns out, is 224. That's to say a very large majority of people will be unable to to tell the difference between audio compressed in MP3 at 224kbps and the original CD, no matter what type of music is used. Also, unless you're blessed with 'golden ears', almost nobody can tell the difference at 320kbps after all!"So there you have it. There is no audible difference. I'm no fanatical debunker of hi-fi snobbery (though I do find it pointlessly elitist, wilfully esoteric and self-deluding in direct proportion to size of bank account) but this seems to be some kind of proof that all the bluster about MP3 being the devil's work is rubbish and that after all the posturing from the hi-fi snobs that it's alright for "commoners" but the elite know better, the elite have been shown to be like the emperor, parading around in their birthday suits. Read it again: "a very large majority of people will be unable to to tell the difference between audio compressed in MP3 at 224kbps and the original CD" A very large majority. So that's probably between 60 and 80 per cent then.And:"unless you're blessed with 'golden ears', almost nobody can tell the difference at 320kbps"Almost nobody. And honestly, genuinely, how many actually have those "golden ears"? Very, very few I would imagine. I've always found that burned CDs, high bit rate MP3s and original CDs sound exactly the same. Even through my i-pod and cheap Sennheiser in-ears the music sounds, rich full and informative. I honestly cannot tell the difference between 320 and CD. In fact the only format I've come across that slays 'em all is vinyl. I'm listening to some Vangelis now and it sounds better than any CD ever will.But if you must go digital, it seems that any way (as long as the bit rate is high enough) will do - and ignore the snobs who tell you different. And BTW, if you had a working system, you'd realize that a computer-burned Cd sounds awful.   We're back to that again are we?

If the sound of vinyl is appropriate for that album, then record it onto vinyl and then back onto CD for release