Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: REALLY lossless? (Read 29743 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

REALLY lossless?

Reply #75
Voodoo.

@ daPyr0x:  You've already said your teacher is the kind of person who "always has to be right".  If this is the case, pushing to prove him wrong may only hurt your own position (and potential grade, even though you said this isn't an issue).  Trying to convince a "believer" is most often pointless.

My recommendation is to smile, nod, take the test, give the answers he told you to give, and move on in life knowing that you know more than he does on this subject.  You're right, and he's wrong.  But the most important concern is the risk-vs-benefit of trying so hard to prove your point.  Seems to me the risk would be medium-to-high and the benefit almost non-existent.
Sometimes you have to jump off the cliff and build your wings on the way down.

REALLY lossless?

Reply #76
This is getting a little weird. He's obviously put some thought and research into this, however poorly done, so what I'm kind of suspecting here is some stupid confusion of terminology or CD read mechanisms. Or he's drank the Kool-Aid from some audio vendor.

EDIT: Wait, are there two teachers espousing this?

REALLY lossless?

Reply #77
I'm willing to bet twenty thousand Korean won that he can't explain properly what subsample data is, or how lossless compression of Word documents is different from that of PCM audio.

If you win the bet and come to Seoul, Korea, I'll hand you the cash. If you can't come, I'll send you a Korean CD.

edit: *I* get to decide what's proper or not.

REALLY lossless?

Reply #78
You really need to break it down into two separate arguments.

First, show that lossless compression really means freakin' lossless. FLAC is really just like ZIP is, except that it's designed specifically for audio. If you can't make him understand this, then the other argument isn't going to matter. If he really doesn't understand the concept behind lossless compression, then you're not going to get anywhere.

Second, the ripping angle is probably the harder one, as this guy sounds like a psuedoaudiophile. You'll need to make him understand that the data on the cd is digital and that a ripping process which is also all digital will end up with the same bits. He probably doesn't really understand what "digital" means (in comparison with analog) and thinks of the CD as something like an LP, with grooves in it. Once you can convince him that it's not an analog thing, that it is an on/off process, then you might make some headway in getting him to grasp that if you take those 1's and 0's and duplicate them exactly, then you have the entire thing. "Subsample" data doesn't make any sense, because there is no such thing. There's a single stream of data on an audio CD. That's it. Getting that stream exactly is very difficult with certain hardware, but very easy with other hardware. So it's not really a matter of getting more or less detail off the CD.

REALLY lossless?

Reply #79
The guy is probably talking about CD sub-channel data (clonecd has an option to read it, but I don't really know what it is, perhaps related to reed-solomon codes).

FLAC losslessly compresses a WAV file, but there's no guarantee that the WAV file is in fact bit identical to the stream on the CD. But it does compress the WAV losslessly, and that's what you should argue.

REALLY lossless?

Reply #80
I was under the impression sub-channel data was something used more for copy protection and the like, but what do I know?

I didn't get to talk to him today, he wasn't around like I thought he would be.  There's always tomorrow, though...

As far as WAV being bit identical to what's on the CD; why not?  I mean, what good reason is there that it wouldn't be when using EAC on paranoia mode?  I mean, caching and ECC are pretty much moot with those settings, and if you're going from pcm cda to pcm wav, I don't see any reason why it couldn't be bit identical (as far as the audio stream itself goes, anyways).

REALLY lossless?

Reply #81
Quote
The guy is probably talking about CD sub-channel data (clonecd has an option to read it, but I don't really know what it is, perhaps related to reed-solomon codes).
Subchannel data on Red Book CDs only contains positioning information, the MSB and the ISRC.

Quote
As far as WAV being bit identical to what's on the CD; why not?  I mean, what good reason is there that it wouldn't be when using EAC on paranoia mode?  I mean, caching and ECC are pretty much moot with those settings, and if you're going from pcm cda to pcm wav, I don't see any reason why it couldn't be bit identical (as far as the audio stream itself goes, anyways).
You can test this yourself. Write a PCM file to a CD, rip it with EAC and then compare it to the original using foobar2000's bitcompare component. Provided you compensate for read and write offsets it should be bit identical.

REALLY lossless?

Reply #82
Quote
Quote
Don't believe somebody, who's telling you, that burning speed has any impact on sound quality or that a copy of a CD sounds different then an original. It isn't possible, because the CD player gets always the same data.

Unfortunately, that isn't neccesarily the case.  Burning speed (along with CDR type, burner, burn type, etc) do affect the intensity of the ink blots (that simulate pits on the CD), and in lower end readers (those found in consumer electronics) they can misread data easier on them as opposed to a stamped CD.  Also lends to the overall longevity of the CD.


Yes, you're right. But the problem is, taht you never know, what is the ideal burning speed. Modern burners and CD-Rs are designed for high speed burning and many tests showed, that using 8x burning with a 48x CD-R is usually worse then 24x for example. Each combination of CD and burner is unique. But as long as the CD is burned correctly every good CD player shouldn't have problems reading it. And on low-level electronics you won't tell the difference anyway.

REALLY lossless?

Reply #83
Quote
My recommendation is to smile, nod, take the test, give the answers he told you to give, and move on in life knowing that you know more than he does on this subject.
but there are FOUR lights!

 

REALLY lossless?

Reply #84
Quote
As far as WAV being bit identical to what's on the CD; why not?  I mean, what good reason is there that it wouldn't be when using EAC on paranoia mode?  I mean, caching and ECC are pretty much moot with those settings, and if you're going from pcm cda to pcm wav, I don't see any reason why it couldn't be bit identical (as far as the audio stream itself goes, anyways).
[a href="index.php?act=findpost&pid=322713"][{POST_SNAPBACK}][/a]


Yes it can be, but there's no guarantee that for every person (specially those don't using EAC) it will be.

You're missing the main point thou, FLAC compresses WAV files and it does it losslessly. Arguing with somebody that is "always right" and have no clue about the subject is more difficult if you try to prove extra things (i.e. FLAC compresses bit identical to CD), remain simple.

Once he gets that FLAC indeed is lossless when compressing any WAV, you can address the subject that a WAV can be a perfect copy of a track on a CD.

Greetings.