Lossy vs lossless

[yetanotherid]

EDIT: Posts from thread What Audio Level (http://forum.doom9.org/showthread.php?t=159709)

How is it done losslessly to a lossy track? It can't be done losslessly to anything, but to say changing the volume of a lossy track is lossless is just bonkers. Whenever anything is edited there is a loss of information from quantization. I find the volume knob easy enough adjust.

Bonkers is calling something bonkers without knowing what you're talking about.
I'm pretty sure there's an explanation on the MP3Gain website. There's no re-encoding, you're simply changing the volume of each frame in the MP3 and it's a process which can be completely reversed, also losslessly. As far as I know most other formats don't store volume information the same way as MP3 so it only apples to MP3.
It's also useful when using programs such as MP3DirectCut. You can use it to edit or fade in and fade out MP3 tracks etc without re-encoding.

[ramicio]

It's bonkers because mp3 is almost the gayest format known to man.

[yetanotherid]

It's bonkers because mp3 is almost the gayest format known to man.

Now I know not to take you seriously.

But no, you said "to say changing the volume of a lossy track is lossless is just bonkers." Nothing to do with your sexual fantasies.

[ramicio]

Anything relating to mp3 in any way is retarded.

[yetanotherid]

Anything relating to mp3 in any way is retarded.

Okay, you obviously need the attention so I'll bite..... Why?

[Ghitulescu]

MP3 was the solution for space problems. These are gone. The multisession idea for CDR was also caused by space problems (at that time the CD had 2x or 3x the size of the largest HDD available). There's nothing left for MP3, other than the players that cannot accommodate WAV or FLAC or other lossless formats.

[ramicio]

Because it's lossy, and an archaic scheme, at that. AAC is much better, but still lossy. Just not a fan of lossy. But it's interesting to find out about this loudness data in each frame. I only find lossy useful for portables, because their storage capacity hasn't caught up enough yet to mirror my lossless library.

[yetanotherid]

I only find lossy useful for portables, because their storage capacity hasn't caught up enough yet to mirror my lossless library.

So it's not completely retarded?

Well because I still keep a lossless copy anyway , and because I'm not one of those people who's convinced I have magic ears and can hear the difference between the original audio and a decent MP3 encode, and definitely not in 99% of the situations where I'm going to be listening to music anyway, I don't see why there's a reason not to be a fan of space saving lossy formats, especially when they allow the cost of portable devices to be kept down while still holding lots of songs, whether they be MP3, AAC or otherwise.

[ramicio]

Anyone without damaged hearing under the age of 80 can easily hear the difference. It's easier to lie about not being able to hear a difference than lying about hearing a difference. If you take a test you can make yourself look mentally retarded even if you know all the answers. But I can see no use, personally, for lossless in a portable environment. Portable means you are going to be exposed to external noise. When I sit down at home in a certain mood and want to listen to some good sounding music through headphone, I will never choose lossy. Same goes for audio tracks in video. I'm even one of those nuts who claim audible difference between Redbook and higher resolution stuff.

[kypec]

Anyone without damaged hearing under the age of 80 can easily hear the difference.
No, statistically 80%+ of population won't be able to hear a difference. Those who can must be exposed to test environment with very high-quality audio components anyway - source DAC, amplifier, speakers - price of such audio chain is somewhere $10,000+ and therefore in real world lossy formats are acceptable and wanted by common people, I wouldn't call anyone using MP3/AAC/Vorbis retarded for that reason.

[Ghitulescu]

90% of the population can detect MP3 if one is kind enough to draw their attention to what exactly are the differences. Like for TVs, most people can distinghish between "egg heads" and normal people, yet they are so lazy to switch the 16:9 mode or hate more the pillarboxes that they got used to them.

It's not "not being aware of" but more "don't care".

[ramicio]

And most people have no idea what to listen for, as far as what compression or digitization artifacts sound like. And again, as for ABX testing, DBT, or whatever you want to call them, they are easily led about to show something is identical. It's called lying. You could show me something blue and something red and I can tell you that I see no difference. You wouldn't even need a $1,000 chain to hear the differences between PCM and lossy stuff. And most people just don't care. Despite internet being fast as hell and storage being plentiful people would still rather download lossy tracks in a few seconds than go buy a CD.

[tebasuna51]

... And again, as for ABX testing, DBT, or whatever you want to call them, they are easily led about to show something is identical. It's called lying. ...

Please stop with this old discussion, is off topic in this thread and you know there are different opinions.

[ramicio]

Why? Is this HydrogenAudio?

[tebasuna51]

Why?

For instance because:
Anyone without damaged hearing under the age of 80 can easily hear the difference.

isn't true for me, and don't want you call me a liar.

[yetanotherid]

Anyone without damaged hearing under the age of 80 can easily hear the difference. It's easier to lie about not being able to hear a difference than lying about hearing a difference.

You've got it backwards. It's human nature to convince yourself you can hear a difference if you think you should be able to hear one.

90% of the population can detect MP3 if one is kind enough to draw their attention to what exactly are the differences.

At what bitrate? Using which encoder?

Like for TVs, most people can distinghish between "egg heads" and normal people, yet they are so lazy to switch the 16:9 mode or hate more the pillarboxes that they got used to them.

You've got to love a meaningless generalisation which probably has no basis in fact.

It's not "not being aware of" but more "don't care".

Rubbish. Even if the average person can easily pick out a decent quality MP3 over the original, which I doubt, there'd still be a willingness to accept some quality loss for convenience, especially when the equipment being used to play the audio makes hearing any quality difference that much harder.
Think back to when we all used cassette tapes in our cars, or even in our walkmans. Or listened to the radio when out and about. Did we care they didn't have the same fidelity as an LP or that they were hissy etc? Of course we did, but they were widely used for their convenience.
The same would hold true for MP3 today. If the average person could hear the difference they'd probably care even if the difference is only small, but given the choice between only being able to store a few CDs on their portable device and being able to store hundreds of them most people would accept a small quality loss for the convenience.
No doubt if they day comes when a portable device can hold as much uncompressed audio as it can hold compressed audio today, and the audio can be transferred to the device just as quickly, most people will listen to uncompressed audio. Why? because they do care. Even if they are only telling themselves they can hear the difference simply because they think they should be able to.

[ramicio]

No, people will have the attitude then that they can only wait .1 second to download a song versus maybe 1 second now. If you can't hear the difference then your hearing is damaged or you do not know what you are listening for. Lossy compression uses tricks to fool the brain, not fool the ears, so by someone saying they can tell the difference doesn't mean they are claiming to have "golden ears", it means they are claiming to have a real brain. Do some LSD and you will hear that if you get involved in a highly compressed song you won't hear artifacts.

[yetanotherid]

No, people will have the attitude then that they can only wait .1 second to download a song versus maybe 1 second now. If you can't hear the difference then your hearing is damaged or you do not know what you are listening for.

Who mentioned downloading? It's got nothing to do with the number of songs which you can store on a device and the fact that so many people download large 720p versions of video from naughty sites when there's a much smaller AVI available contradicts your theory that most people won't wait a little longer for quality if they want it. Some people do care for the extra quality, some prefer a smaller download, either way it proves nothing as to whether the people downloading either version can see the difference between them.

And no, I use my ears for a living and they take some abuse so I get them tested regularly and they're within normal range for my age, and yes I know exactly what I'm listening for. I also know how easy it is to hear things when you want to hear them.

Lossy compression uses tricks to fool the brain, not fool the ears, so by someone saying they can tell the difference doesn't mean they are claiming to have "golden ears", it means they are claiming to have a real brain.

Your eyes see images and send them to your brain. If they see a succession of still images go by fast enough the brain will perceive it as continuous motion.
Lossy audio uses different tricks which fool the brain in different ways, and to claim you can hear those tricks and therefore have a real brain while other people don't contradicts it's lack of involvement in putting forward your viewpoint. Which lossy audio tricks is your brain able to hear which mine can't?

Do some LSD and you will hear that if you get involved in a highly compressed song you won't hear artifacts.

No mention of bitrates in your argument, no mention of the type of compression artifacts you might be able to hear yourself, just a drug reference....

Who said anything about a highly compressed song? What's your definition of a highly compressed song? At least offer some specifics. At which bitrates do lossy audio "tricks" or artifacts start to become really obvious to you?
My original comment, which is the one which I think started this discussion, was that I can't hear the difference between the original audio and a decent MP3 encode, not between the original audio and a low bitrate MP3. And I can't.... at least not when using the equipment I use to listen to audio. Maybe if I had a lot of money and could afford the best equipment I might be able to hear some differences, but if you're going to claim you can hear the difference how about some specifics?

Better still, why not prove what you say to shut me up? Why not upload a small example of an original audio track, a decent MP3 encode of that audio track and a description of the differences you can hear and allow those of us with slightly less than real brains to educate them as to what we're not hearing? I'm keen to learn. In fact it's probably the best way to continue this discussion, otherwise we'll just be arguing theory back and forth and never convince anyone of anything. By doing it that way, we can all listen to the same audio, all listen for the same differences, and all discuss what we can or cannot hear in a rational manner.
Maybe Ghitulescu would like to help educate those of us with uneducated brains in a similar fashion. I'd guess as 90% of people can hear the difference between uncompressed audio and an MP3, there's a pretty good chance I'd be one of the 90%, or several other people reading this thread will fall into that 90% category and might like to participate in the experiment/discussion too, and he'd also be able to teach us what to listen for. It'd even be interesting to see if the two of you hear the same differences or whether both of you have brains which are real, but in different ways.
Maybe one of you could even start a new MP3 Education thread in which to conduct the MP3 lessons and to keep the mods happy, and so we're not disrupting this thread any further. An education on MP3 audio such as that would deserve it's own thread anyway. Or if you like, I'll start a new thread and upload a few different samples, as long as I know I won't be wasting my time and Ghitulescu and yourself are willing to participate.

[ramicio]

Hardcore enthusiasts actually only wait to download 720p stuff. People will still and always favor the 700 MB xvid rips, because if they are too cheap to buy it, then they are too cheap to have gone to the theater to see if they even like the movie, so they aren't going to waste the time to download DVD5 or 9 sized file when they can download a CD sized file. ANY mp3 is highly compressed in my viewpoint. AAC is slightly better. Get a life. Don't turn this place into HydrogenAudio. Let technology progress.

[yetanotherid]

Hardcore enthusiasts actually only wait to download 720p stuff. People will still and always favor the 700 MB xvid rips, because if they are too cheap to buy it, then they are too cheap to have gone to the theater to see if they even like the movie, so they aren't going to waste the time to download DVD5 or 9 sized file when they can download a CD sized file. ANY mp3 is highly compressed in my viewpoint. AAC is slightly better. Get a life. Don't turn this place into HydrogenAudio. Let technology progress.

I'm impressed that you can speak for the entire population in respect to what they'll always favour in the way of downloads and the reasons for it, even if you're probably completely wrong.

"Get a life." "Don't turn this place into HydrogenAudio." I wondered what the excuses might be.
I've got some samples ready to upload so you and Ghitulescu can teach us uninformed people what to listen for, but it doesn't surprise me you're not prepared to put your ears where your mouth is.
I guess that puts an end to this discussion, at least for me. You don't seem to be willing to offer anything which would make me think I should take your claims seriously. Maybe Ghitulescu will come back to perform the task of teaching my semi-real brain what to listen for in a new thread, although I kind of doubt that'll happen either.

[ramicio]

Find me a method of testing that isn't flawed and I will take it. You know anyone can take a test and claim not to hear a difference between 2 things. It's a two-way street. How is this so hard to understand?

[yetanotherid]

Find me a method of testing that isn't flawed and I will take it. You know anyone can take a test and claim not to hear a difference between 2 things. It's a two-way street. How is this so hard to understand?

It's not a test as such, although if other people participate it might be interesting to see how many claim they can hear a difference, but both Ghitulescu and yourself have stated that pretty much anyone can hear a difference between an MP3 and the original audio ("anyone without damaged hearing under the age of 80 can easily hear the difference"), and you both also claimed if people can't hear it, it's because they don't know what to listen for. I can't hear a difference, therefore I must not know what to listen for. It's not so hard to understand and if you tell me what to listen for and I can hear it I won't be claiming otherwise.

I'd like to upload some different uncompressed and MP3 samples in a new thread, have yourself, Ghitulescu and anyone else who cares to participate download them and listen for differences, and then explain to me what I should be listening for in those samples in order for me to hear the difference between an MP3 and the original audio too. It shouldn't be too hard given it's so easy for you to hear the difference yourself, and I'm sure there'll be plenty of others who'll read the thread and also be willing to learn.

[Ghitulescu]

@yetanotherid:
If you can't hear a difference that means, statistically, that you belong to the 20%. Someone should be enrolled for the 20%, don't you think? And if you can't hear the difference, how can one explain to you that there is a difference?

[yetanotherid]

@yetanotherid:
If you can't hear a difference that means, statistically, that you belong to the 20%. Someone should be enrolled for the 20%, don't you think? And if you can't hear the difference, how can one explain to you that there is a difference?

20%?? It was only 10% a little while ago. To refresh your memory, here's what you said:

90% of the population can detect MP3 if one is kind enough to draw their attention to what exactly are the differences.

All I'm asking is for you to be kind enough to draw my attention to the differences. If I happen to be one of the 10%, so be it, but I'm sure I won't be the only one reading the thread and there will be others who fall into the 90% category who'll be happy for you to draw their attention to the differences. Lots of people view this forum.

Why the excuses? This shouldn't have to be hard. You made a claim which you say applies to 90% of the population, ramicio went even further and claimed it applied to anyone under the age of 80 without hearing damage, and while admittedly at this point I don't believe either of you, I'm giving you what should be an easy opportunity to prove me wrong.

[ramicio]

If you can't hear differences, you can't. One can only describe what artifacts sound like, not directly make your senses know what they sound like. Everyone's perception is different, and everyone's senses are slightly different. Some people can see different wavelengths outside the visible spectrum. Some infrared, some UV. Why is it so miraculous or taboo that hearing can be the same? No one is claiming to have bat-like hearing, just that their bias of sensitivity may be in the higher frequencies versus mids or lower. If everyone's ear pieces, canals, external ears, sinuses, etc. were all identical down the atom then people might be able to even begin to claim that everyone hears the same way. It's why people prefer different sounds. Some people like bass-heavy stuff, some like mids, some highs, and some prefer balanced sound.

My hearing tends to smear mids all into one sound, be lacking in bass sensitivity, but I can hear a lot of detail in the highs. I can hear quite high frequencies 20 khz+ (not so much hear, but be able to detect their presence.) I listen to classic rock, and the vocals and guitars always smear together, meaning if the singer is singing my brain can't focus on what the guitars are doing anymore. I need a lot of boost to get bass booming. Anything that is a cymbal I can detect easily if it's been lossily compressed. They sound like they were run through one of those speaker-spinning effect machines from back in the day, like some heavy phasing is done to them, hard to describe.

[yetanotherid]

Okay, I'll take the continuing list of excuses as an unwillingness to prove your claims.

It's fairly simple, you claimed virtually everyone can here the differences between an MP3 and the original audio if they have those differences pointed out to them. Ghitulescu claimed it was 90% of the population. Those are the claims I am disputing. Nothing else.

I'm willing to start a new thread in order to upload some samples, have the two of you listen to them and report back the differences almost everyone should be able to hear once they're made aware of them (as you both claimed). I have no doubt I'll not be the only one viewing the thread and downloading the samples for comparison. I'm sure I won't be the only one who'll respond regarding whether they can hear the differences Ghitulescu and yourself can hear. If 90% of the people who reply to the thread can hear the differences Ghitulescu claims they'll hear once he points those differences out, I'll stand corrected. If everyone under the age of 80 without hearing damage who responds also agrees they can hear the differences you point out, I'll stand corrected once again.

Those are the claims you both made. No furious back-peddling now as you offer qualifications regarding differences in people's hearing, or tales of your own special ability to hear frequencies normally reserved for dogs will change those claims. Once again, I'm stating I don't believe either of you but I'm giving you the opportunity via a new thread and the responses of myself and other posters to prove me wrong.... and hopefully I'll be able to hear those differences and learn myself. Let me know when you're ready to stop back-peddling and ready for me to take you seriously and I'll start the new thread.

[ramicio]

How are frequencies reserved for dogs? Dogs don't hear much beyond human capability anyway. Now cats and bats hear insanely high up. How are we going to be so sure your sources are honest? Most people's hearing is damaged. It's not permanent. Most people spend most of their time around noise, therefore their brain just ignores a lot of information. Go spend some time, like months, in pure mother nature and then take a hearing test. Most people clean their ears by forceful or liquid means, which is also harmful. I just let my earwax come out on its own.

[yetanotherid]

How are we going to be so sure your sources are honest?

Well if you think suspecting me of lying is another excuse worthy of exploring (I must admit I never thought the list would include earwax), feel free to upload the samples yourself. I'm happy to either learn using my samples, or to learn using yours.

[tebasuna51]

New thread with estracted post from What Audio Level thread.

[Groucho2004]

AAC is much better
AAC is slightly better.

Interesting how you change your opinion in less than 3 days.

[ramicio]

Interesting to see that I don't care, because it doesn't change my opinion that lossy is worthless for real listening.

[shon3i]

I am fan of lossless to, but aslo i can't say that lossy encode with proper encoder is worthless especially with codec such AAC. Difference is most visually if you watch spectograms or peaq, while i can bet nobody really can't hear differenece between AAC@320 and lossless, unless you are not human.

[Groucho2004]

Anything relating to mp3 in any way is retarded.
Anyone without damaged hearing under the age of 80 can easily hear the difference.
Most people's hearing is damaged.
lossy is worthless for real listening.

I thought it would be worthwhile to collect some highlights of your blatant generalizations for everyone's pleasure.

[Ghitulescu]

Comments on the summary above:
1. MP3 is obsolete. It unified a telecom technology with the acute lack of space of the early PCs and portable devices (when the CD appeared the largest HDD was some 80MB, when DAE-enabled drives appeared they jumped to ~320MB). Now we have the technology to provide unaltered quality (FLAC, APE etc.) and the needed space. An aggressively compressed FLAC is of ~the same size as a 320kbps MP3, so why not satisfy the placebos, too?

2. If s/he knows what to sense. There are people that don't.

3. The hearing is adaptive (because what we hear and see and smell is processed in the brain). Nobody can sense her/his own smell (because we get used to, yet we can sense it after a bath because it changes), yet the "sensors" are not technically broken. The hearing is what the brain understands, not what the ear actually hears. The image is reversed, so we have equilibrium problems in the very first year, but we learn how to "mirror" it. Are we defects?

4. The better the sound system one has (and paid a lot of money), the easier is to spot the differences. And yes, a 320kbps MP3 sounds extremely close to the original, sometimes better (it filters some bands), but in the end MP3 remains a lossy codec.

I think he is right, just exaggerate a bit, as you said, in generalisation.

PS: to 3. and 4. - I said once in this forum that during the very first MP3 years, many people tried to convince me that MP3 (max. 128kbps was the standard that days) sounded "indistinguishable" from the CD it originated. I invited one of those guys to me and let him hear the MP3 and the WAV through the AWE64-Technics system. He changed his mind, but still said that on his system (PC with cheap consumer card - PC speakers) the MP3 was the same as the WAV. Many years later people realised this, now 320kbps seems "indistinguishable" from the WAV it originates. So the hearing is perfectible and therefore damaged, to reuse some words.

[yetanotherid]

I'd like to say thank you to tebasuna51 for moving this discussion to a new thread. Originally tebasuna51 closed the old thread which seems to happen a bit around here and it's terribly frustrating when obviously in a discussion forum, discussions are going to take place, and sometimes the topic of discussion takes a different direction to the original thread.
In this case the original question asked in the old thread had been answered and the OP was long gone so it's terribly frustrating to have answered questions asked, only to be cut off later in the middle of a discussion which is being carried out in a civil manner and not really hurting anyone.
So thank you to tebasuna51.

[GodofaGap]

An aggressively compressed FLAC is of ~the same size as a 320kbps MP3, so why not satisfy the placebos, too?
Not a chance. I have never seen a lossless codec that averages out under 50% compression for music (~700 kbps for CD) and FLAC certainly doesn't.

4. The better the sound system one has (and paid a lot of money), the easier is to spot the differences. And yes, a 320kbps MP3 sounds extremely close to the original, sometimes better (it filters some bands), but in the end MP3 remains a lossy codec.

Not necessarily true either. Psycho acoustic models make assumptions about audio properties (like frequency response) that are usually better met in good audio systems.

[Ghitulescu]

Not a chance. I have never seen a lossless codec that averages out under 50% compression for music (~700 kbps for CD) and FLAC certainly doesn't.

It might be 200 vs 300 (vs 600), but in order for any MP3 to be perceived as "CD quality" (with the standards of 2010, ie 320, not those of 1992, ie 128) its size more than doubled, ie no longer 80-100 for an album but 200+. That was my point.

MP3 started to loose from its killer advantage, the space savings. It was also the favourite audio codec in DIVX era (again for space reasons not for its quality), now more and more AC-3 audio tracks are found into these files.

I know this debate can be continued until the doomsday, but nobody ever will convince me that MP3 at 320kbps sound identical (is transparent) to its original WAV. It's only a debate about the trade-off sound-space. Which in the end, like quality, is a personal thing.

[yetanotherid]

Some comments on Ghitulescu's previous posts. As with ramicio's later posts there seems to be some back-peddling going on in the form of adding qualifications to the original claims.

90% of the population can detect MP3 if one is kind enough to draw their attention to what exactly are the differences.

@yetanotherid:
If you can't hear a difference that means, statistically, that you belong to the 20%.

The percentage of people who can't hear the difference just doubled.

1. MP3 is obsolete.
The point is a red herring whether the MP3 is obsolete or not as it has absolutely nothing to do with whether someone can hear the difference between an MP3 and the original audio, and that is the subject which is being discussed.

An aggressively compressed FLAC is of ~the same size as a 320kbps MP3, so why not satisfy the placebos, too?

It's blatantly untrue for a start. Even an aggressively compressed flac file is around twice the size of a 320kbps MP3. Which in turn can be around twice the size of a decent VBR MP3 while the quality may still remain indistinguishable from the original.
Look at the samples I'll be uploading in my next post. The VBR MP3 versions total about 15MB. The 320kbps samples total about 25MB. The flac versions total around 60mb. If you're going to offer another red-herring, at least make sure it's accurate one.

2. If s/he knows what to sense. There are people that don't.

Yet as you've pointed out, most of those people can apparently be taught what to sense. Which is why I originally offered to upload samples in order for yourself and ramicio to teach the rest of us what to listen for, yet so far you've ignored that offer.

3. The hearing is adaptive (because what we hear and see and smell is processed in the brain). Nobody can sense her/his own smell (because we get used to, yet we can sense it after a bath because it changes), yet the "sensors" are not technically broken.

True, which is why even if there's tiny difference between an MP3 and an original it's probably almost irrelevant. While ramicio and yourself have claimed to simply be able to tell if a source is MP3 when you hear it, I'd be theorising that the majority of us would need to listen to the MP3 and original side by side before we'd be able to tell the MP3 does sound different, if we're "lucky" enough to be able to hear the difference at all.

4. The better the sound system one has (and paid a lot of money), the easier is to spot the differences. And yes, a 320kbps MP3 sounds extremely close to the original, sometimes better (it filters some bands), but in the end MP3 remains a lossy codec.

Is this a qualification you're now adding to your original statement that 90% of people can easily hear the difference if they're taught what to look for? Are you now saying they can only hear the difference when using a top quality sound system? If that's the case and the majority of people don't use a top quality sound system, wouldn't it mean there's effectively no difference when using the type of sound system the average person would use?
Before you answer that question (unless of course you're going to ignore the ones which don't suit you as has been your previous preference) let me remind you what you said on this same subject in another thread a while ago:

I recognize MP3 from WAV on my car, it's neither a silent car (Diesel) nor I have good ears: the trick is to recognize compression artefacts: once you got them you can't overlook them. It happens also on TVs, I can't stand LCDs :)

I just thought I'd remind you now of your previous claim of not having good ears yourself, yet you can pick an MP3 on an average sound system in a noisy environment, so I can only assume the 90% of the population who could also pick an MP3 from the original, once they're told what to listen for, will also be able to do the same and won't require hideously expensive sound systems to do so? Would the average MP3 player and a decent set of headphones suffice?
Does your comment in a previous thread nullify any further argument regarding the quality of the sound system used?

to 3. and 4. - I said once in this forum that during the very first MP3 years, many people tried to convince me that MP3 (max. 128kbps was the standard that days) sounded "indistinguishable" from the CD it originated. I invited one of those guys to me and let him hear the MP3 and the WAV through the AWE64-Technics system. He changed his mind

That's all very well, and I doubt anyone would argue it's not possible to pick the difference between an original recording and a 120kbps MP3 on a decent sound system, however that's not the topic being discussed, rather it's another diversion.

My original statement was that I can't hear the difference between a decent quality MP3 and the original, a statement which both ramicio and yourself claimed would be untrue for almost everyone. Can we stick to the original subject and not deviate onto the topic of low quality compression?

Many years later people realised this, now 320kbps seems "indistinguishable" from the WAV it originates. So the hearing is perfectible and therefore damaged, to reuse some words.

Another logical fallacy of the kind which you're so fond of offering to support your argument. Many years later did you offer for the same person to compare a 320kb MP3 with the original using the same sound system and have him also agree he could hear the difference, or are you assuming the rest of use will accept that simply through inference the same results would apply to an MP3 today?

And of course the argument completely ignores the fact that the MP3 hasn't remained stagnant over the years. Despite it's "obsolete" label the LAME MP3 encoder at least, has constantly been developed and improved. What may have been true in terms of comparisons back when the MP3 first emerged as a common compression format may be completely untrue when using an MP3 encoder today.

[yetanotherid]

I know this debate can be continued until the doomsday, but nobody ever will convince me that MP3 at 320kbps sound identical (is transparent) to its original WAV. It's only a debate about the trade-off sound-space. Which in the end, like quality, is a personal thing.

Which is why I offered an opportunity for yourself and ramicio to prove the claims you've made in the form of sweeping generalisations. An offer which so far has constantly been ignored.

Nevertheless I've uploaded some samples to use as comparisons. If what you and ramicio say is true, you'll have no problem identifying the differences between the original audio and the MP3 versions and explaining those difference to the 90% of the rest of us who should also be able to hear them too. If others download the samples and can also hear the differences you point out, then you're correct and I stand corrected.

Here's a link to a zip file containing seven 1 to 2 minute samples taken from typical pop songs.
I used flac files which were on my computer and which I know I ripped myself directly from the CDs. The audio has not been altered in any way aside from the application of ReplayGain when converting the flac files back to wave files (I did this in order to balance the levels and to make sure none of them would clip a pre-amp).
I then took 1 to 2 minute samples from each song, trying to ensure each one was different in some way in order to ensure there'd be a wide variety of compression artefacts to point out. For instance one is a live recording containing audience applause, others range from only acoustic guitar and vocals to full pop tunes to hard rock etc.
Once I'd edited out the samples I converted those wave files to 320kbps MP3 and also to a VBR MP3 using the standard LAME V2 preset. It's the method I use for all of my own MP3 conversions. I then converted the edited wave files back to flac in order to save space while including the original audio in the samples.

I'll eagerly await the differences you can hear without having good ears yourself. Differences and artefacts which you can still hear on a car sound system without even having the original audio to compare them to. I look forward to you pointing out those differences and artefacts in the hope I may be able to re-listen to the samples and find myself able to hear them too.

http://ifile.it/40wzm3i/Music%20Samples.zip

[GodofaGap]

It might be 200 vs 300 (vs 600), but in order for any MP3 to be perceived as "CD quality" (with the standards of 2010, ie 320, not those of 1992, ie 128) its size more than doubled, ie no longer 80-100 for an album but 200+. That was my point.
Then you have a particular lousy way of making your point, because what you just said is so incoherent I'm not even going to try to understand it.

[Ghitulescu]

I don't own low quality gear, it happens to own also studio gear. People that got used to hear music in good conditions feel the pain of listening the same on low fidelity gear. The same is true for video.

And since not many people do care about the quality today (or in the past) as much as they care about the price, I can give a comparison everyone can test it or have already tested it: passing from ***** hotels to ** ones is painful (it's much easy to get from ** to *****), and anyone passing from ** to ***** hotels (for a period of time long enough to get used to), can find it difficult to downgrade back to ** (despite they got used to ** before and they were quite happy with **). Every single human sense acts this way. Because it's not the sense but the brain.

Why don't you simply stay at ** and live happy, pretending ***** is the same just more expensive? Matthew 5:3. I am happy with my toys and I pretend to have what I paid for. You are happy (maybe) paying 1$ per MP3 at Amazon.
I did once a test (with the help of some friends involved in wedding videos): not even the worse user of camrips and divx/AVI files of the latest movies wanted to have his own wedding as Divx. They claimed before that divx/xvid is the same as a DVD, just space oriented. So, they knew it, they knew that's not true, yet they pretended the contrary, to ease their feelings as they couldn't (or wanted not to) pay for the DVD/BD.
This LoFi = HiFi just cheaper axiom is very common to people having access to LoFi gear only. And it's speculated by marketing to an extent few can dare to think, in both ways, of course.

[GodofaGap]

From now on I'm going to call you Mr. Anecdote.

[Ghitulescu]

Thanks,

but remember, the inventors of MP3 stated that the compression algorithm exploits some weaknesses of the human hearing. Many sound engineers do not want to hear about noise reducers in their studios, yet consumer types of NRs are plenty (DNL, HighCom, Dolby B/C etc.) and their results explain why they are avoided in studios (even the Dolby A is avoided if possible, as it may more degrade than cure if incorrectly set) - because all of them exploit human hearing "weaknesses"/features. The best NR is to have LN gear.

Imagine that not all people are born the same, and some hear, some don't. Those that don't hear don't care, those that hear do care. And they cannot explain the other part where the differences are. Try imagine explaining what colour RED is to a daltonist. Or the beauty of a butterfly to a blind.

[yetanotherid]

From now on I'm going to call you Mr. Anecdote.

I think you're being too generous. For one thing, the anecdotes would have to be coherent. ;)

The DivX argument is of course another fallacy. DivX was widely used because on a CRT TV it's virtually impossible to distinguish a DivX encode form the original, yet it doesn't negate the desire to have a better quality copy as the "master" recording.
The question would be, did any of those people who didn't want a DivX copy as the master recording of their wedding even think to discuss whether the audio track would come in the form of uncompressed audio, AC3 or MP3?? :)
I'd bet nobody expressed a preference for uncompressed audio.

It seems Ghitulescu is unwilling to download the samples and put his ears where his mouth is. ;) I'll admit I don't hold any high hopes for ramicio being willing to either, but I guess I'll wait and see.

[Ghitulescu]

I don't need samples for what I know. And I definitively for something that I'll never use.

Divx wasn't for CRT TVs, was for computer monitors, as Divx is progressive, like monitors are. CRT TVs were interlaced, I'm not sure I know any progressive consumer TV (CRT), and only HD studio monitors (CRTs) could have an option for progressive.

Divx was given as an example how low quality devices can make two things "identical". I assume anyone playing the same Divx and DVD on the new 22" or more PC monitor or LCD/Plasma TV can see there are differences. Wow, isn't that nice?
If one can digest the Divx example can pass to the next test: differentiating an MP3 from its original using first a Durabrand player (bought for 1.99$ at Walmart) and a Mark Levinson one (sorry that you have to sell your house to buy one). On such setups the "golden ears" can sense differences between 2 CDR with the exactly same data.

[Groucho2004]

Comments on the summary above:
1. MP3 is obsolete. It unified a telecom technology with the acute lack of space of the early PCs and portable devices (when the CD appeared the largest HDD was some 80MB, when DAE-enabled drives appeared they jumped to ~320MB). Now we have the technology to provide unaltered quality (FLAC, APE etc.) and the needed space. An aggressively compressed FLAC is of ~the same size as a 320kbps MP3, so why not satisfy the placebos, too?

2. If s/he knows what to sense. There are people that don't.

3. The hearing is adaptive (because what we hear and see and smell is processed in the brain). Nobody can sense her/his own smell (because we get used to, yet we can sense it after a bath because it changes), yet the "sensors" are not technically broken. The hearing is what the brain understands, not what the ear actually hears. The image is reversed, so we have equilibrium problems in the very first year, but we learn how to "mirror" it. Are we defects?

4. The better the sound system one has (and paid a lot of money), the easier is to spot the differences. And yes, a 320kbps MP3 sounds extremely close to the original, sometimes better (it filters some bands), but in the end MP3 remains a lossy codec.

I think he is right, just exaggerate a bit, as you said, in generalisation.

PS: to 3. and 4. - I said once in this forum that during the very first MP3 years, many people tried to convince me that MP3 (max. 128kbps was the standard that days) sounded "indistinguishable" from the CD it originated. I invited one of those guys to me and let him hear the MP3 and the WAV through the AWE64-Technics system. He changed his mind, but still said that on his system (PC with cheap consumer card - PC speakers) the MP3 was the same as the WAV. Many years later people realised this, now 320kbps seems "indistinguishable" from the WAV it originates. So the hearing is perfectible and therefore damaged, to reuse some words.

Wow. You really missed the the point entirely, didn't you?
His statements are either his personal opinions or hypotheses wrapped as facts.

You're now doing it as well:
1. MP3 is obsolete.

According to what or whom? Do you have any numbers to back this up? How many people are using flac or similar instead of mp3?

[yetanotherid]

I don't need samples for what I know. And I definitively for something that I'll never use.

Well if you're now going to claim that was the point of the exercise, I guess I'll give up expecting anything more from you than sweeping generalisations and unsubstantiated claims. The point of the samples was to allow you to teach the rest of us what you already know.

Divx was given as an example how low quality devices can make two things "identical".

I recognize MP3 from WAV on my car, it's neither a silent car (Diesel) nor I have good ears: the trick is to recognize compression artefacts: once you got them you can't overlook them. It happens also on TVs, I can't stand LCDs :)

I'll wait till you've stopped arguing yourself into a hole and you're maybe willing to download the samples and teach the rest of us how the hear the compression artefacts you've claimed 90% of us should be able to easily hear once we're taught what to look for.
Until then, I doubt you'll be able to offer anything more than further anecdotes to lower your credibility even further.

[kypec]

I've been watching this thread very closely even before it got splitted, downloaded the samples provided by yetanotherid and will post my hearing experience shortly.
On such setups the "golden ears" can sense differences between 2 CDR with the exactly same data.
Wow, the best way to express my feelings about your statements is to quote Martin Crane when he said to his son Frasier:

M.C.: I've underestimated you.
F.C.: [feeling flattered] Really?
M.C.: Yeah, you're making even bigger jackass of yourself than I thought.

Now you're trying to make us believe that you're able to tell apart two identical records burned on separate CD-R. :rolleyes:
How come? Ah, I forgot - you can hear beyond the Error Correction Codes mechanism and distinguish which bytes were re-constructed in affected disc sector, God bless your ears! :p

[Ghitulescu]

I still don't see how I can use the samples to prove that you don't hear any difference. And if the samples are supposed to show me that there's no acoustic difference, well, there are indeed cases where an MP3 sounds the same, like 1kHz sine or square wave.

I said that quality is a matter of personal choice, experience, native gifts and expectations. I am happy with mine.

[yetanotherid]

I still don't see how I can use the samples to prove that you don't hear any difference.

Of course you can't prove what I can hear. Why do you continue to go to such great lengths to avoid the point (rhetorical question)?
You can use the samples to explain the differences you can hear. Whether I can hear them is irrelevant. Using your original claim as the basis, if I can't hear the differences then statistically the next 9... or depending on which of your claims is used.... 8 out of the next 9 people who download the samples will be able to hear the same differences you do. Therefore, your point is proven and I am wrong in claiming you're wrong. (feel free to add unnecessary highlighting at your discretion).

I said that quality is a matter of personal choice, experience, native gifts and expectations. I am happy with mine.

I don't recall you saying that. Do you imagine you said that or do you honestly believe it?

90% of the population can detect MP3 if one is kind enough to draw their attention to what exactly are the differences.

The above quote is the topic under discussion. Are you planning on getting back to it eventually?