Lossless audio formats, why didnt WAV take off?

[JSt0rm]

Take a lossless audio file and a lossy audio file into a audio workstation. The lossy file should be converted back into wav.

Now invert the phase of one of the files 180 degrees and line it back up and listen. You are now hearing the differences and anything that is the same has been canceled out.

If lossy files sounded the same as lossless then they should null into silence. Thats science. Everything else is subjective.

 

[Cogman]

Take a lossless audio file and a lossy audio file into a audio workstation. The lossy file should be converted back into wav.

Now invert the phase of one of the files 180 degrees and line it back up and listen. You are now hearing the differences and anything that is the same has been canceled out.

If lossy files sounded the same as lossless then they should null into silence. Thats science. Everything else is subjective.

The question isn't "are there differences" the name "lossy" implies that. The question is "Can humans hear the difference?"

What you've posted isn't science.

 

[JSt0rm]

The question isn't "are there differences" the name "lossy" implies that. The question is "Can humans hear the difference?"

What you've posted isn't science.

If there are differences within the range of human hearing then they will sound different. Now have you trained yourself to pick up on those differences?

How is posting a way to reproduce a method to test if a lossy file and lossless file null NOT science? I have no control over it. You can take that information and try and prove your claim.

 

[Cogman]

If there are differences within the range of human hearing then they will sound different. Now have you trained yourself to pick up on those differences?

How is posting a way to reproduce a method to test if a lossy file and lossless file null NOT science? I have no control over it. You can take that information and try and prove your claim.

Because when dealing with audio/visual feeds, the difference between two feeds is not important. The human mind doesn't interpret differences based off of differences even if the difference is perceivable (using your method of subtraction).

The best example I can think of for this is in the x264 project. There are several measurements that say how mathematically close two sources are (PNSR and SSIM are very common), yet one of the biggest improvements to the x264 encoders causes those scores to decrease significantly. The result, however, is an image that LOOKS like it is closer to the source to the human eye. Even performing a difference results in noticeable differences.

That is science. Realizing that humans aren't some infallible machines and taking advantage of it.

Oh, and for reference the settings are PsyRD and PsyTrellis

 

[JSt0rm]

Because when dealing with audio/visual feeds, the difference between two feeds is not important. The human mind doesn't interpret differences based off of differences even if the difference is perceivable (using your method of subtraction).

I have no idea what you are trying to say here.

Also most anyone can tell the difference between h264 and bluray in the right situation (yeah on a 13" screen of course not) so I dont know what you brought that up. I like to look at h264 but its not some imperceptive difference between it and full 1080p.


I would have to ask though - Would a colorist be able to tell the difference even on a 13" screen?

 

[Cogman]

I have no idea what you are trying to say here.

Pretty much exactly what I just said.

Also most anyone can tell the difference between h264 and bluray in the right situation (yeah on a 13" screen of course not) so I dont know what you brought that up. I like to look at h264 but its not some imperceptive difference between it and full 1080p.

You do realize that blurays commonly USE h264 as their codec, correct? (they can use VC-1 as well, it is also lossy).

 

[JSt0rm]

Pretty much exactly what I just said.


You do realize that blurays commonly USE h264 as their codec, correct? (they can use VC-1 as well, it is also lossy).

I dont know very much about video. Lets keep talking about audio

I'm not saying your an expert either though. I just dont have the knowledge to comment on your claim.

 

[Cogman]

II dont know very much about video. Lets keep talking about audio

I'm not saying your an expert either though. I just dont have the knowledge to comment on your claim.

My point with video translates directly over to audio. Lossy compression can easily reach levels where no human can tell the difference between the two sources. Not only that, when you take the difference of the two sources like you where suggesting, it doesn't necessarily translate into "this is what you can hear/see" as a difference.

It is true, you can train yourself to see compression artifacts, and even hear them, but that only helps to some extent and usually only at the lowest of bitrates.

AAC at 128kb/s is extremely good. AAC at 64kb/s roughly translates to MP3s between 128 and 196kbs to paint a picture of how well it works. And maybe that is where the confusion comes in as most people don't deal with AAC and they prefer their MP3s (which you can hear distortions at 128kb/s).



As for the edited part of your post.

No, I'm not an expert, but I do know more than most on the subject. You can verify what I've said if you like by going to forums.doom9.org or www.doom10.org . The developers of the best h264 codec (x264) are fairly active there and will pretty much repeat what I've said in regards to lossy video compression. (you could google it as well if you like).

 

[JSt0rm]

It is true, you can train yourself to see compression artifacts, and even hear them, but that only helps to some extent and usually only at the lowest of bitrates.

ok dude I think this is as close to agreement on this issue as we will have.

 

[Cogman]

ok dude I think this is as close to agreement on this issue as we will have.

I think we can solidify it a little more.

Do you agree that at some point, some bitrate, a lossy compression algorithm become indistinguishable from its lossless cousin by the human ear, correct? (You should realize that there is a bitrate with all lossy algorithms at which point the output will be a bit perfect match with the lossy algorithms)

 

[0roo0roo]

Really? I just explained why. Refute my points if you think they are wrong.

No, how about you prove your extraordinary claim, if you could it would be worth millions to the audiophile industry to finally have proof that their products are more than snake oil.

Taking bits of science which are superficially related and using them to support unfounded conclusions is snake oil 101.

 

[JSt0rm]

I think we can solidify it a little more.

Do you agree that at some point, some bitrate, a lossy compression algorithm become indistinguishable from its lossless cousin by the human ear, correct? (You should realize that there is a bitrate with all lossy algorithms at which point the output will be a bit perfect match with the lossy algorithms)

Well I would say even lossless is still lossy and there is a quest to get closer and closer to perfection all the time. What that perfection is is in the eye of the beholder.

Some people would say perfection is lavry

http://www.lavryengineering.com/productspage_pro_ad122_96mk.html

and others would say that is burl

http://www.burlaudio.com/products/b2-bomber-adc

And they are both right.

As for your question only time will tell. I'm not going to say that in any room you cant hear the difference. If we were a/bing compressed formats in this room we would probably be hearing a lot of things we didnt know were there.

 

[JSt0rm]

No, how about you prove your extraordinary claim, if you could it would be worth millions to the audiophile industry to finally have proof that their products are more than snake oil.

Taking bits of science which are superficially related and using them to support unfounded conclusions is snake oil 101.

Oh I'm sorry you are confused. I laugh at audiophiles all the time. We are talking about compressed digital formats here. We arent talking about $3000 cables. Big difference. If the barrier to being a audiophile is now that you don't listen to compressed audio we are fucked.

 

[0roo0roo]

I dont know very much about video. Lets keep talking about audio

I'm not saying your an expert either though. I just dont have the knowledge to comment on your claim.

That is pretty obvious....anyone interested in compression knows h264 is part of bluray....and has nothing at all to do with quality being inherently worse.

Your subtraction theory doesn't work, these things work on the limits of perception and do not rely on recreating the same wave form, and if you percieve two things are identical they are. Unless you can prove that you can repeatedly pick out the lower quality source in a blind test no claims can be made.

 

[0roo0roo]

Oh I'm sorry you are confused. I laugh at audiophiles all the time. We are talking about compressed digital formats here. We arent talking about $3000 cables. Big difference. If the barrier to being a audiophile is now that you don't listen to compressed audio we are fucked.

No, you aren't wiggling out of this one.

Same thing still applies, if you have proof of your extraordinary claims they would be worth millions to the industry that relies on claims about sound quality.

Says a lot that this industry has not been able to come up with anything scientifically solid that would support their claims isn't it. They've had all the time in the world, and all the incentive, but they can't deliver.

 

[JSt0rm]

No, you aren't wiggling out of this one.

Same thing still applies, if you have proof of your extraordinary claims they would be worth millions to the industry that relies on claims about sound quality.

Says a lot that this industry has not been able to come up with anything scientifically solid that would support their claims isn't it. They've had all the time in the world, and all the incentive, but they can't deliver.

I'm not going to defend that industry. I'm not advocating crazy power cables with huge magnets and other voodoo.

I would say you have probably not experienced good audio in a good room if you think its all bullshit.

 

[JSt0rm]

At what point is it no longer snake oil?

Are 20k a speaker B&W in the above picture snake oil? are the weiss digital eq's in his rack that cost $7000 snake oil?

Is that room snake oil?

Is made in china in a untreated room as good as it gets?

I'm really trying to understand your mindset here.

 

[0roo0roo]

It is no longer snake oil when you can scientifically prove your claims.
Let alone pass a simple blind listening test.

These are simple requirements, the fact that you are ducking them shows you need to believe more than you want to know whether it is true or not.

You can be technically impressive all you want, if you could write a book on a grain of rice that would be impressive, just don't pretend it is quality you can percieve in any real way with your given senses.

 

[JSt0rm]

It is no longer snake oil when you can scientifically prove your claims.
Let alone pass a simple blind listening test.

These are simple requirements, the fact that you are ducking them shows you need to believe more than you want to know whether it is true or not.

You can be technically impressive all you want, if you could write a book on a grain of rice that would be impressive, just don't pretend it is quality you can percieve in any real way with your given senses.

I'm not ducking anything. I dont have time to do a double blind and are you gonna fly out to LA to administer the test? I mean really. We cant have a discussion about compressed vs not compressed without some forum warrior flying across the country?

If its no longer snake oil when I can scientifically prove my claims then I present the null test and you say no not good enough because of subjectivness.

I'm already saying its all subjective anyhow.

Can I sit in my car on the highway and listen to the difference between compressed and lossless? Probably not. Could I sit in that room above and do it? probably 90% of the time.

 

[JSt0rm]

Oh and thank you for calling me technically impressive

 

[LikeLinus]

Post a better test that disproves what I'm saying. In other words, put up or shut up. It should be easy, after all, every audio aficionado and his dog claims this super sonic hearing that can tell the difference between any lossy and lossless.

Prove to me that someone can tell the difference between 128 AAC and lossless just using their own hearing.

Ok, provide me funding for the research and I'll conduct the study. Your "theory" is based on Lossy vs. Lossy and you're saying that lossy=lossless.. I feel for you and you're a prime example of why education in the U.S. is failing. I have far more to say on this, but honestly I'm hanging out with my wife and this is SO not worth my time.

 

[Howard]

If its no longer snake oil when I can scientifically prove my claims then I present the null test and you say no not good enough because of subjectivness.

You shouldn't even talk about the results since it's not going to happen.

Also, you never answered Cogman's question. The answer is a "Yes, because" or a "No, because".

 

[Howard]

Can I sit in my car on the highway and listen to the difference between compressed and lossless? Probably not. Could I sit in that room above and do it? probably 90% of the time.

You do realize how lame you sound by continuing to make these unprovable claims?

Hey, I can fly, but it doesn't work if it's videotaped so you're gonna have to come here to see it.

 

[JSt0rm]

You do realize how lame you sound by continuing to make these unprovable claims?

Hey, I can fly, but it doesn't work if it's videotaped so you're gonna have to come here to see it.

Ok we can go round and round. How can I describe what red looks like to blind people? Thats how I feel right now.

 

[JSt0rm]

You do realize how lame you sound by continuing to make these unprovable claims?

Hey, I can fly, but it doesn't work if it's videotaped so you're gonna have to come here to see it.

Oh and I told you what I would need to do to give you your test. I'm not doing that. you guys are robots. I'm not arguing as an audiophile. For fucks sake I know half this shit is recorded with a sm57 thru a mogami cable. Why the hell would I spend 1200 on a rca cable?