FLAC VS MP3

[Cerb]

Originally posted by: ixelion
Is there a really big diffrence between FLAC and MP3, should I spend the time to re-encode all my CDs to FLAC?

Basically is lossless that much better that compressed?

Uh, yes. It's done. Unless you need a copy that can CDDB, going to FLAC, with a good ripper, will be the end of your ripping CDs.

It's not FLAC vs. MP3. You can make MP3 from FLAC. The comparison is FLAC vs. available HDD space. MP3 can give you good quality sound to listen to (especially LAME at preset standard or extreme)...but what if you need it in another format? Rip again. What if you get a system where you can tell a difference? Rip again.

With FLAC, that's not an issue. Especially if you use EAC's secure mode with the proper read offset (not absolutely needed, but it only takes one CDR to get it), you will simply never need to rip the CD again, barring massive hardware failure. So um, with that in mind, back things up, too .

I rip straight to FLAC now, transcoding to MP3 for my flash player. AFter awhile, hearing little ripping or encoding errors gets tiring. Rip to FLAC and don't encounter them.

 

[Cerb]

Originally posted by: ixelion
OK thanks for the advise, I see now that ripping AC3 from a cd makes no sense at all.

However, many have said they use VBR MP3s while others use Ogg.

My question is, if file size is not an issue, what will be a better audio quality (not technically but subjectively i.e. your opinion) VBR MP3 or whatever birate people use for Ogg.

Foobar2000 has a ABX test plugin. It also has the capability to fairly easily transcode.

To quote myself:

Yawn why not just listen? Transparency above around 160kbps, with an output good enough that you might tell a difference, is highly subjective. It really does not take that long, especially w/ FB2K, to make various quality MP3, Ogg, or whatever else, from a lossless source, and listen to them.

From my own ABXing, FI, I seem to like stereo more than detail, and on several tracks, can pick out Ogg up to around q9 (70-80%, typically, so not perfectly conclusive), while not able to distinguish most MP3 of --preset standard, and only a couple songs at --preset extreme.

With all the time taken ripping, encoding, and then listening, is it not worth a few hours (and not more than 15 minutes at a time, usually) to figure out where your threshhold for transparency is?

adaptable to any, if you aren't familiar w/ FB2K's function-over-form transcoding

 

[The Boston Dangler]

FLAC will generally compress to 55-70% of original CDA size, with no loss of quality. It's useable in any PC player software, and any portable that does OGG. The final file is a faithful reproduction of the original music, which is the goal of any serious music enthusiast. When have you seen a CRC checksum for an mp3?

If you're happy with mp3, fine. Just don't tell me it's just as good. The numbers- and more importantly, my observation- say otherwise.

 

[svi]

No, the reason I haven't seen you support your arguments isn't because I haven't done any research; it's because you haven't presented any evidence. Strangely enough, you've accused me of the same impropriety,

I just gave you a link to evidence. If you don't know what a listening test is or don't understand why it constitutes evidence against your point, you have no business wasting one more second of anyone's time.

I've accused you of this because I know for a fact that the argument you're making is one that has been disproven many times before you were even born.

The human ear can detect small variations in sound. Just do a simple Google search, and you'll find a slew of instructional pages like this one, or this one, which unquestioningly verify my contention.

Dynamic range of hearing has little to do with lossy codecs. You can't link a page about the human ear and say "that's evidence for my point!" Similarly, I can't link a page about cars and say "that's evidence that high-quality lossy codecs are comparable to lossless!" Why? Because there is no direct connection. If you think there is a direct connection, either research psychoacoustics until you understand why your point is illogical and unrelated to those links

I've tried very hard to make this as clear as possible. I have illustrated in plain language the following:

1. If you can't hear something, it's not audible.
2. If you think you can hear something where few others can, and you haven't done double-blind testing, you have probably bitten the sugar pill.
3. You should not put forth arguments unless you are willing to supply evidence.

Which of these points are you having trouble with?

But think about what that means. You've heard how people can identify tiny variations in the sound they hear, but you can't do it yourself, right? Well, the truth is, unless you've had some sort of accident, you can hear these tiny variations. You just don't realize it.

No, if you can't hear them, you can't hear them. That's what it means if you can't hear something. If this is not clear to you, I strongly advise seeing a doctor. You can repeat "the truth" all you want, but it will still be nothing more than an unfounded allegation that, just as last time, you will argue to the death.

If you're happy with mp3, fine. Just don't tell me it's just as good. The numbers- and more importantly, my observation- say otherwise.

It is just as good when encoded properly. If you feel otherwise, one or more of the following is true:

1. You are a victim of the placebo effect.
2. You don't know how to encode mp3s.

Neither of these is particularly shameful. #2 is difficult, and #1 is something that every human being is to some point. However, in the future, please try not to make sweeping, vaguely condescending statements about codecs.

 

[Cerb]

Originally posted by: The Boston Dangler
FLAC will generally compress to 55-70% of original CDA size, with no loss of quality.

The low end varies much more than that. My lowest has been 43%. But expect an average of about 65%.

It's useable in any PC player software, and any portable that does OGG.

No, it is not. FLAC fits in an Ogg container, but the player must still support FLAC, and may even need to support the Ogg format as a container, rather than assuming Ogg = Vorbis.

The final file is a faithful reproduction of the original music, which is the goal of any serious music enthusiast. When have you seen a CRC checksum for an mp3?

Never, because it's pointless. LAME will do it, though.

If you're happy with mp3, fine. Just don't tell me it's just as good. The numbers- and more importantly, my observation- say otherwise.

 

[The Boston Dangler]

Originally posted by: svi

No, the reason I haven't seen you support your arguments isn't because I haven't done any research; it's because you haven't presented any evidence. Strangely enough, you've accused me of the same impropriety,

I just gave you a link to evidence. If you don't know what a listening test is or don't understand why it constitutes evidence against your point, you have no business wasting one more second of anyone's time.

I've accused you of this because I know for a fact that the argument you're making is one that has been disproven many times before you were even born.

The human ear can detect small variations in sound. Just do a simple Google search, and you'll find a slew of instructional pages like this one, or this one, which unquestioningly verify my contention.

Dynamic range of hearing has little to do with lossy codecs. You can't link a page about the human ear and say "that's evidence for my point!" Similarly, I can't link a page about cars and say "that's evidence that high-quality lossy codecs are comparable to lossless!" Why? Because there is no direct connection. If you think there is a direct connection, either research psychoacoustics until you understand why your point is illogical and unrelated to those links

I've tried very hard to make this as clear as possible. I have illustrated in plain language the following:

1. If you can't hear something, it's not audible.
2. If you think you can hear something where few others can, and you haven't done double-blind testing, you have probably bitten the sugar pill.
3. You should not put forth arguments unless you are willing to supply evidence.

Which of these points are you having trouble with?

But think about what that means. You've heard how people can identify tiny variations in the sound they hear, but you can't do it yourself, right? Well, the truth is, unless you've had some sort of accident, you can hear these tiny variations. You just don't realize it.

No, if you can't hear them, you can't hear them. That's what it means if you can't hear something. If this is not clear to you, I strongly advise seeing a doctor. You can repeat "the truth" all you want, but it will still be nothing more than an unfounded allegation that, just as last time, you will argue to the death.

If you're happy with mp3, fine. Just don't tell me it's just as good. The numbers- and more importantly, my observation- say otherwise.

It is just as good when encoded properly. If you feel otherwise, one or more of the following is true:

1. You are a victim of the placebo effect.
2. You don't know how to encode mp3s.

Neither of these is particularly shameful. #2 is difficult, and #1 is something that every human being is to some point. However, in the future, please try not to make sweeping, vaguely condescending statements about codecs.

1. I have observed a difference in audio quality, and mp3 sounds lousy. For the record, I prefer the sound from originals on my Denon DVD-2900 -> Denon AVR-2802 over analog input. I believe the better sound is due to the DAC's in the player. FLAC on WinAmp -> A2ZS -> AVR-2802 over digital coax is a close second. The player supports mp3, and the sound was just as bad, even with it's DAC's.

2. I'm very familiar with several types of transcoding, compression, modulation, amplification, et cetera.

mp3 is similar to Sony's horrible ATRAC, in that supposedly inaudible information is discarded. Then, the bandwidth is reduced (cut of below X Hz and above Y Hz), as well as the dynamic range (the difference between the softest and loudest sounds). Anything left is then compressed.
I will continue to poo-poo all lossy codecs, thanks.

 

[svi]

1. I have observed a difference in audio quality, and mp3 sounds lousy.

Repeat: placebo. You haven't done a double-blind test, you expect mp3 to sound worse, so it does. It's the same thing that makes people claim to hear a difference between $200 speaker cables and 12AWG lamp cord.

2. I'm very familiar with several types of transcoding, compression, modulation, amplification, et cetera.

To make things clear, I didn't mean to insult your expertise, I was only providing a possibility. At any rate, if that is the case, then it's definitely placebo.

mp3 is similar to Sony's horrible ATRAC, in that supposedly inaudible information is discarded. Then, the bandwidth is reduced (cut of below X Hz and above Y Hz), as well as the dynamic range (the difference between the softest and loudest sounds). Anything left is then compressed.

mp3 is similar to any lossy codec in that supposedly inaudible information is discarded. At sufficiently high bitrates, said information actually is inaudible according to DBTs.

I will continue to poo-poo all lossy codecs, thanks.

Sure, it's absolutely your decision what to do with your audio and your system. Just please don't act as if you have better ears or taste than the rest of us because you've taken a sugar pill.

SVI: I'm done with you, so, if you want, you can cease your blather.

Well, that's not very nice. I was rather hoping you'd agree to cease putting forth your opinion on this matter until you found a shred or two of scientific evidence for it, or maybe even admit that you didn't know what you were talking about (as in the last few times-- in a bad mood this time or something?), but I guess an end to this inane argument is the next best thing.

 

[hurtstotalktoyou]

Originally posted by: The Boston Dangler
It's useable in any PC player software, and any portable that does OGG.

I don't know about all that. What about Windows Media Player, which, sadly, everyone and their brother uses? Sure, there's CoreFLAC, but it's pretty darn buggy.

SVI: I'm done with you, so, if you want, you can cease your blather.

 

[The Boston Dangler]

Originally posted by: hurtstotalktoyou

Originally posted by: The Boston Dangler
It's useable in any PC player software, and any portable that does OGG.

I don't know about all that. What about Windows Media Player, which, sadly, everyone and their brother uses? Sure, there's CoreFLAC, but it's pretty darn buggy.

SVI: I'm done with you, so, if you want, you can cease your blather.

A plugin will take care of any player that doesn't do it natively.

svi: I don't need a sugar pill, I'm sweet enough. Pr!ck.

 

[Cerb]

Originally posted by: hurtstotalktoyou

Originally posted by: The Boston Dangler
It's useable in any PC player software, and any portable that does OGG.

I don't know about all that. What about Windows Media Player, which, sadly, everyone and their brother uses? Sure, there's CoreFLAC, but it's pretty darn buggy.

I don't think anyone who uses WMP cares. If you do care, you can easily find a player to do Free codecs, like FLAC; Winamp and FB2K being the most popular.

 

[Concillian]

Originally posted by: The Boston Dangler
I dunno 'bout you, but that swooshy metallic sound in just about any mp3 is easily apparent to me.

Good for you!

Does that have ANY bearing on the original poster's issues?

This thread is a total mess. Nobody's golden ear has any bearing on the original poster's issue at all. Everybody hears differently, and everybody is comparing audio encoded with different algorithms.

What matters is whether the OP would fail such a blind test as I did and my semi-audiophile brother did (he was an audiophile, but he seems to have outgrown it). Any test where you know which is .mp3 and which is .wav is pretty useless in my opinion. It's important to listen in this manner first in order to learn how the compression alters original sounds, but when choosing your final compression method you need to perform your own blind test where you don't know which is original and which is .mp3.

 

[Cerb]

Originally posted by: Concillian

Originally posted by: The Boston Dangler
I dunno 'bout you, but that swooshy metallic sound in just about any mp3 is easily apparent to me.

Good for you!

Does that have ANY bearing on the original poster's issues?

This thread is a total mess. Nobody's golden ear has any bearing on the original poster's issue at all. Everybody hears differently, and everybody is comparing audio encoded with different algorithms.

What matters is whether the OP would fail such a blind test as I did and my semi-audiophile brother did (he was an audiophile, but he seems to have outgrown it). Any test where you know which is .mp3 and which is .wav is pretty useless in my opinion. It's important to listen in this manner first in order to learn how the compression alters original sounds, but when choosing your final compression method you need to perform your own blind test where you don't know which is original and which is .mp3.

QFT * eleventy billion plus one.

 

[Pariah]

Originally posted by: ixelion
Basically is lossless that much better that compressed?

If you can't hear the difference, then, no, lossless isn't better at all. Don't pay attention to what anyone else says. Do your own test encodings. If you can't tell the difference, don't waste your time reencoding.

Any player supporting OGG can can play FLAC files too. When encoding, put the FLAC in an OGG wrapper. Presto!

Not true. Your ability to post disinformation appears to know no bounds here.

 

[hurtstotalktoyou]

Originally posted by: Concillian

Originally posted by: The Boston Dangler
I dunno 'bout you, but that swooshy metallic sound in just about any mp3 is easily apparent to me.

Good for you!

Does that have ANY bearing on the original poster's issues?

Sure it does. He asked if it was "worth it" to re-encode his CDs. Knowing the benefits of lossless over lossy should help him decide.

This thread is a total mess.

You can blame that on SVI. I got sucked into his argumentative nonsense, too, so I guess I'm partially at fault. I don't feel bad, though, because that's how threads usually are.

Nobody's golden ear has any bearing on the original poster's issue at all.

Yes, it does. I know you probably don't agree with me, but double-blind tests don't mean anything.

Everybody hears differently, and everybody is comparing audio encoded with different algorithms.

Everybody hears differently, but there are certain common elements. Everyone can hear the difference between mp3s and FLACs, for example.

What matters is whether the OP would fail such a blind test as I did and my semi-audiophile brother did (he was an audiophile, but he seems to have outgrown it).

Double-blind tests are inherently flawed. Hearing doesn't always translate into identifying. In other words, people do hear a difference; they just don't realize it. So failing a double-blind test doesn't mean you can't hear the differences; it just means you can't consciously identify them.

Think about some ordinary-looking person you only met once. Could you pick him out of a lineup of similar-looking people? Maybe, maybe not. But if you couldn't, would that mean you didn't see the differences in their faces? No, it just means you couldn't recognize them. And that's faces--with mp3s the difference is much more subtle.

 

[Cerb]

Originally posted by: hurtstotalktoyou
Double-blind tests are inherently flawed. Hearing doesn't always translate into identifying. In other words, people do hear a difference; they just don't realize it. So failing a double-blind test doesn't mean you can't hear the differences; it just means you can't consciously identify them.

Think about some ordinary-looking person you only met once. Could you pick him out of a lineup of similar-looking people? Maybe, maybe not. But if you couldn't, would that mean you didn't see the differences in their faces? No, it just means you couldn't recognize them. And that's faces--with mp3s the difference is much more subtle.

Your analogy has nothing to do with a doublt-blind test. Here's a double-blind test:
There are two similar-looking people next to each other, and a divider. On the other side of the divider are two similarly looking people. These people are four brothers. Two sets of identical twins. In the test, you must match up which twin on one side is which twin on the other side. However, you do not have to know which is the older set, or which is the younger set; just that they are not the same.

A line-up is completely different. With a ABX test, being able to tell that there is a difference is enough. You don't have to pick the 'right' one, you just have to match two of them as being the same.

Your analogy puts you into listening one sample once, then identifying later. A double-blind AB test should put both samples forth immediately, and allow any amount of listening to either sample, in whole or in part. If you can tell the difference, then you should be able to repeatably get a decent amount of them right w/ a ABX test, as long as it isn't rushed (fatigue is an issue, and why you shouldn't do many tests at once, but spread them out).

 

[svi]

Well put, Cerb. In my experience, much of the refusal to accept double-blind tests stems from ignorance of their nature. The vast majority of the audiophiles that denounce them don't actually know what they involve, and demonstrate their ignorance when they try to use examples to show their points. Perhaps they reject the method out of ignorance, perhaps out of a lack of belief in the scientific method (which generally entails ignorance of said method as well, IMHO and IME), perhaps they just want to believe really hard that their $300 speaker cables make a difference.

You can blame that on SVI. I got sucked into his argumentative nonsense, too, so I guess I'm partially at fault. I don't feel bad, though, because that's how threads usually are.

If you insist on making unfounded, false, extremely general arguments, then others will usually feel obliged to object as so to ensure that you mislead as few people as possible.

Yes, it does. I know you probably don't agree with me, but double-blind tests don't mean anything.

Says someone who not only is completely unqualified on the matter, but doesn't even know what a DBT is. There are just so, so many things that have to be wrong with your reasoning for you to continually argue this at this stage. A few are outlined below.

Everybody hears differently, but there are certain common elements. Everyone can hear the difference between mp3s and FLACs, for example.

Yes, you say that a lot. You are not an audiologist. You are not an audio engineer. You have no experience or knowledge in psychoacoustics or psychology. Most important of all, you have no evidence whatsoever to back up that statement. If this doesn't illustrate why you should stop repeating yourself, I don't know what will. (What happened to "I'll shut up until I can find someone qualified who agrees with me"? Did you suddenly realize no such person existed, or did you think it was more important to repeat yourself than to attempt to verify your argument?)

It's bad enough that you claim there's an audible difference between ultra-high-bitrate VBR mp3 and FLAC. To claim that everyone should hear such a difference is... well, words fail me.

Double-blind tests are inherently flawed. Hearing doesn't always translate into identifying. In other words, people do hear a difference; they just don't realize it. So failing a double-blind test doesn't mean you can't hear the differences; it just means you can't consciously identify them.

Actually, if a difference is audible at all-- has ANY effect on the brain-- then it would be picked out in a DBT. You would know this if you knew anything at all about DBTs. As it is, you can read the definition in Wikipedia if you're not comfortable with relying on my word.

svi: I don't need a sugar pill, I'm sweet enough. Pr!ck.

Ouch, my feelings. Ah, sugar pills aren't actually sweet.. they usually just taste starchy and awful in the mouth, just as (almost) any pill. I think the sugar is in them to make sure your body is fooled into accepting the crap as food.

And anyway, your semi-nonsensical retort has nothing to do with my point. You are a victim of the placebo effect, and are unable to argue otherwise due to that your testing does not rely on the scientific method (and, probably, that you do not fully understand the placebo effect or said scientific testing).

 

[Concillian]

Originally posted by: hurtstotalktoyou

Double-blind tests are inherently flawed. Hearing doesn't always translate into identifying. In other words, people do hear a difference; they just don't realize it. So failing a double-blind test doesn't mean you can't hear the differences; it just means you can't consciously identify them.

This is exactly why I mentioned that you should first listen to both knowing which is which so you can learn what poor compression does to audio. I notice you managed to not quote the part of my post where I explained that. I implied, but didn't state clearly that it should be done first at low bitrate, then progressively higher so that you can identify the characteristics. Educating yourself on the differences will help you not to accidentally encode at a rate where you develop the ability to distinguish the differences later because you learn what to listen for. That will annoy the piss out of you if you do that.

Think about some ordinary-looking person you only met once. Could you pick him out of a lineup of similar-looking people? Maybe, maybe not. But if you couldn't, would that mean you didn't see the differences in their faces? No, it just means you couldn't recognize them. And that's faces--with mp3s the difference is much more subtle.

So what? Does it bother you simply knowing that you're listening to an .mp3 instead of .wav? That seems pretty messed up. But in that case I guess you just bite the bullet and go with FLACs. All I'm saying is that it's a personal decision. I think you should let your own ear decide rather that take people's advice on what sounds good and what doesn't. I mean we all have ears, we should be able to make that judgement for ourselves. Listening to someone else's opinion on what sounds good and bad is kind of an odd concept to me.

Back to the lineup example... If you can't recognize the differences, then why do the differences even matter? Using a lineup as an example puts an emotional charge to a non-emotional issue. It puts images in people's heads that there are some consequences for making the 'wrong' assessment. There is no wrong assessment in a blind test. The .wav file will not get punished unjustly if you choose incorrectly. You will not feel guilty, because your decision is not going to potentially send the selection for 25 to life in a federal 'pound me in the ass' prison. You are simply choosing which you think is the losless and which is the lossy and testing your own threshold for noticing artifacts. How can there NOT be value in testing what you can identify and what you can't?

Again, who cares if the differences are so subtle that you can't tell the difference. It's not like you are going to sit in your listening chair and cringe at the difference you can't hear. I mean, you can't hear the difference. If you can hear the difference, then it's a different story. But it's important to know whether or not you can.

Everybody hears differently, but there are certain common elements. Everyone can hear the difference between mp3s and FLACs, for example.

I can't. At least not the way I encode my .mp3s.

Sure, if I use crappy encoding I can, but I'm not going to archive my entire collection using POS encoding. Different encoding algorthms are VERY different. People aren't talking about not being able to tell the difference from some 128k garbage encoding with a half-assed CODEC, they are talking about high variable bitrate encoding with CODECs of above average quality. If you're talking about 128k or 192k straight rates I would agree that just about anyone (who isn't deaf) can tell the difference, but I guarantee that those arguing that they can't tell the difference are not using this kind of low quality compression... at least I'm not.

 

[hurtstotalktoyou]

Originally posted by: Cerb
Your analogy has nothing to do with a doublt-blind test. Here's a double-blind test:
There are two similar-looking people next to each other, and a divider. On the other side of the divider are two similarly looking people. These people are four brothers. Two sets of identical twins. In the test, you must match up which twin on one side is which twin on the other side. However, you do not have to know which is the older set, or which is the younger set; just that they are not the same.

A line-up is completely different. With a ABX test, being able to tell that there is a difference is enough. You don't have to pick the 'right' one, you just have to match two of them as being the same.

Your analogy puts you into listening one sample once, then identifying later. A double-blind AB test should put both samples forth immediately, and allow any amount of listening to either sample, in whole or in part. If you can tell the difference, then you should be able to repeatably get a decent amount of them right w/ a ABX test, as long as it isn't rushed (fatigue is an issue, and why you shouldn't do many tests at once, but spread them out).

You seem to have misunderstood me. My analogy wasn't referring to double-blind tests. I know how those work, and they are not relevant to my point.

The analogy is meant to show how subtle differences are not always noticed. That's why it doesn't matter whether or not you can identify lossy and lossless audio.

Consider this: some listeners can identify mp3s, and others cannot. Why do you suppose that is? Is it because some listeners' ears are unable to hear the audio well enough? No; their hearing is fine. They just don't consciously understand and identify the different sounds.

 

[Jeff7]

Checks back, cautiously looks around and assesses the situation.


Well, add another one to the list.
Intel vs AMD
nVidia vs ATi
Windows vs Linux
.
.
.
.
.
FLAC vs MP3

 

[hurtstotalktoyou]

Originally posted by: Concillian

This is exactly why I mentioned that you should first listen to both knowing which is which so you can learn what poor compression does to audio. I notice you managed to not quote the part of my post where I explained that.

It still doesn't matter. Demonstrating the differences before the test doesn't guarantee the listener will be able to identify them, even if they are picked up by their own ears.

Does it bother you simply knowing that you're listening to an .mp3 instead of .wav? That seems pretty messed up. But in that case I guess you just bite the bullet and go with FLACs. All I'm saying is that it's a personal decision. I think you should let your own ear decide rather that take people's advice on what sounds good and what doesn't. I mean we all have ears, we should be able to make that judgement for ourselves. Listening to someone else's opinion on what sounds good and bad is kind of an odd concept to me.

Not once in this thread have I said you shouldn't listen to mp3s or other lossy audio. I simply pointed out a much-misunderstood fact, because the guy asked about the difference between lossy and lossless.

If you can't recognize the differences, then why do the differences even matter?

Since you're hearing different sounds, your brain certainly will respond differently. It's quite possible that means you enjoy the music a little bit less.

Again, I'm only pointing out a fact which I happen to find interesting. It's not a dealbreaker, but I think it's relevant to the initial question.

 

[Cerb]

Originally posted by: hurtstotalktoyou
Consider this: some listeners can identify mp3s, and others cannot. Why do you suppose that is? Is it because some listeners' ears are unable to hear the audio well enough? No; their hearing is fine. They just don't consciously understand and identify the different sounds.

You're right. And if you don't know how to tell them apart, then for you, there is no difference. That in no way invalidates good testing prodecures. If it were not the case, then it would be a very simple matter to say X (lossy) is transparent, and Y (lossy) is not.

However, it is not that simple, because we do not all hear the same things the same way. If you cannot tell the difference, then the difference is inaudible. If you do not consciously hear it, it is not there for you.

 

[Baked]

Only if you have the equipment to backup the sound. If you're just gonna use the integrated sound card w/ a pair of $5 headphones, stick w/ MP3.

 

[Cerb]

Originally posted by: hurtstotalktoyou

Originally posted by: Cerb
Your analogy has nothing to do with a doublt-blind test. Here's a double-blind test:
There are two similar-looking people next to each other, and a divider. On the other side of the divider are two similarly looking people. These people are four brothers. Two sets of identical twins. In the test, you must match up which twin on one side is which twin on the other side. However, you do not have to know which is the older set, or which is the younger set; just that they are not the same.

A line-up is completely different. With a ABX test, being able to tell that there is a difference is enough. You don't have to pick the 'right' one, you just have to match two of them as being the same.

Your analogy puts you into listening one sample once, then identifying later. A double-blind AB test should put both samples forth immediately, and allow any amount of listening to either sample, in whole or in part. If you can tell the difference, then you should be able to repeatably get a decent amount of them right w/ a ABX test, as long as it isn't rushed (fatigue is an issue, and why you shouldn't do many tests at once, but spread them out).

You seem to have misunderstood me. My analogy wasn't referring to double-blind tests. I know how those work, and they are not relevant to my point.

The analogy is meant to show how subtle differences are not always noticed. That's why it doesn't matter whether or not you can identify lossy and lossless audio.

However, if choosing whether or not to go lossless, those differences do not matter--it's the differences that you can hear that matter, if your decision is based mostly on sound quality.

With a 250GB HDD at $120 or less, though, and DVDs at $.40 or under (even for good quality ones) I don't think there are any good arguments for using lossy formats to save space.

 

[wpeng]

Originally posted by: hurtstotalktoyou
Read my initial post in this thread, at the top of p2.

The human ear can detect small variations in sound. Just do a simple Google search, and you'll find a slew of instructional pages like this one, or this one, which unquestioningly verify my contention.

But think about what that means. You've heard how people can identify tiny variations in the sound they hear, but you can't do it yourself, right? Well, the truth is, unless you've had some sort of accident, you can hear these tiny variations. You just don't realize it.

And there-in lies the problem with your argument. You assume everyone has the same definition as you. If I don't realize it, then by my definition, I can't hear it. Even if my neuronal ions pick up the slight differences, when my brain processes the information and decides what I hear, that's the end result.

 

[Jeff7]

Ok, I shall summarize that which has been said and flamed:


FLAC is lossless. Re-encode it as much as you want, it'll still be the same as what comes off of the CD. It has a worse compression ratio than MP3 for that reason, and so it will take up more disk space than MP3.

Whether or not it'll sound different is subjective, much as monitor refresh rates. Some people don't mind 60Hz. I personally prefer 85Hz or higher; 60Hz gives me a headache quickly.

So, to the OP:
Re-encode one of your CDs to FLAC, and listen to it. If there's enough of a difference to you to justify the additional hard drive space usage, the time of re-encoding all your discs, and the knowledge that your CD collection is archivd losslessly, then do it. Otherwise, don't.


Concerning the butterfly effect of minor distortions, that's probably best left to a philosophy discussion. Yes your mind may well perceive it. But it just might not matter, and so you really won't remember or notice it. The brain uses very lossy processing and compression. Data gets thrown out all the time without your conscious approval or knowledge.