Finally. Converting my stuff

mmadden28

Just do the test yourself...Rip a file twice or more, then run a File checksum utility on the resulting files and see if the checksums are the same or different. If one single bit is different the checksum will be radically different.
Here's a basic one from MS - FCIV.
There are many many checksum verifiers out there, both command line and GUI based.

mmadden28

heiney9 wrote: »

That is a grossly untrue statement. Lossless Compression DOES NOT equal loss of data. Compression or conversion or ripping may not give you a bit perfect duplicate but it has nothing to do with loss of data; PERIOD!

+1

It should probably be noted that the compressed lossless files are transcoded back to their original uncompressed form on the fly as they are being played. Maybe that's why there is some disconnect in grasping this concept.

It's like opening a compressed zip file, on the fly and transparent to the end user, it's decompressed and opened in its full original glory before being sent on to the Application (i.e. DAC)

This would explain why, with the Wadia review, that the output stream was the same as the original WAV file.

mmadden28

mmadden28 wrote: »

Just do the test yourself...Rip a file twice or more, then run a File checksum utility on the resulting files and see if the checksums are the same or different. If one single bit is different the checksum will be radically different.
Here's a basic one from MS - FCIV.
There are many many checksum verifiers out there, both command line and GUI based.

I did a quick test on a ~29s long track.
In all cases I ripped to Lossless WAV using the same track with the same drive on the same computer.
1.

Media Player
File size is 5,158,302 (4.91MB)
The checksum is 7a73fa2aa131d1fd8674e3248f6ebcbd29eb68ce

2.

EAC
File size is 5,157,980 (4.91MB)
Checksum is 73252ff22daf49d503929c1be4d78fb504119ac1

3.

WinAMP
File size is 5,157,980 (4.91MB)
Checksum is 10eb772138360e51447dd46e9c1b2d35086c5777

4.

MediaMonkey
File size is 5,158,186 (4.91MB)
Checksum is 21bffe7ec9b5673297919ea421e480d68c8276ce

Note the checksum between all 4 rippers is different - thus they are not bit for bit identical to each other.
Also the file size between 2 & 3 are the same, the others are different by a small amount.

So it appears that different applications result in different output files by a small degree. I would imagine that is related to the error detection engines in use (if at all) as well as the WAV encoding engine used.
I also used the EAC 'Compare WAV' utility, and in all cases the EAC file differed from the others by '6 samples'.

To ensure the results were consistent, I ripped the same track again with all the apps and the results were exactly the same. The checksums exactly matched the original ripped files. So the resulting file should be the same everytime it extracted when using the same application. This makes sense, making the AccurateRip online checksum database possible.
This also shows that its application dependent, and not the laser reading the disc differently each time.

And for kicks and giggles, I took a WAV file and converted it to FLAC with MediaMonkey, and then converted that FLAC back to WAV. The original .wav file and the new one converted from flac back to wav file checksums were identical. Thus no loss of information at all.

So, as H9 mentioned, the only real loss possibly occurs in the original extraction method and application used, but that's not any fault of the compression algorithm so still lossless. It's likely that the same possible extraction losses occur when just straight up playing the CD raw, so really when comparing apples to apples, there is no real loss.

mmadden28

tcrossma wrote: »

If the .WAV file is created the exact same each time (and I don't know if they are or aren't, which is why I'm asking) then how can we think that it is not a perfect bit-for-bit copy being ripped?
.

In addition to what i wrote above:
Since EAC uses AccurateRIP and an online database of checksums, and the majority of people are getting the same checksums across a widely varying subset of hardware and OSs, its logical (to me) to think that the EAC method of extraction is fairly accurate, if not consistent, and so probably the most bit for bit accurate to the original CD.

wayne3burk

Over the last 4 years i've converted 110 LPs to digital... and unthinking i converted them all to 128kb MP3... and didnt write them as wave files to CDs (for no apparent reason). so in hindsight and for the last 15 or 20 i have written them as 44.1HZ 16bit .wav CDs. I guess the moral to this is to spend the 20cents per CD to compliment your Hour per LP. A worthwhile investment.

vc69

Thanks mmadden28 for the test.
I had a feeling that was true, but I never did the legwork. I think the moral of the story here is use EAC (it's free) and FLAC (it's also free) and you can be confident in your outcomes.

Maybe one of these days I'll get back to ripping my collection. For now, I'm listening to discs.

Pauly

Thanks for the input guys. Im already familier with EAC so my next project is to start ripping everything again now that i have the room to do it properly.

Pauly

bigaudiofanatic

mmadden28 wrote: »

+1

It should probably be noted that the compressed lossless files are transcoded back to their original uncompressed form on the fly as they are being played. Maybe that's why there is some disconnect in grasping this concept.

It's like opening a compressed zip file, on the fly and transparent to the end user, it's decompressed and opened in its full original glory before being sent on to the Application (i.e. DAC)

This would explain why, with the Wadia review, that the output stream was the same as the original WAV file.

From what I have seen the ripped format of flac or lossless equals the same space taken up on the disk. So thus far is an exact copy of what is on the disk. NOTHING gets compressed and nothing is thrown away.

vc69

bigaudiofanatic wrote: »

From what I have seen the ripped format of flac or lossless equals the same space taken up on the disk. So thus far is an exact copy of what is on the disk. NOTHING gets compressed and nothing is thrown away.

Not so.

FLAC (it's the one I am most familiar with) lets you select a compression value wich directly corresponds with the final file size. Even if you choose a higher value, the FLAC file will be smaller than the original .wav file.

The size on disk is another matter if that is what you are saying.

heiney9

bigaudiofanatic wrote: »

From what I have seen the ripped format of flac or lossless equals the same space taken up on the disk. So thus far is an exact copy of what is on the disk. NOTHING gets compressed and nothing is thrown away.

Bigaudio,

Not a slight towards you, but you need to spend a little more time researching so you can understand these concepts better .

H9

reeltrouble1

hey guys I was away for a few days both playing and listening to music.......gotta say I enjoyed pumping out some bass on my Fender Roscoe Beck at a music thing we did on Saturday night, getting to play with other musicians is a kick and a half for me, really enjoyed doing bass lines in The Joker ala Steve Miller..........and of course I spent alot of time in the Rabbit Hole....I bagged four more OOP SACD for the weekend, getting deep into CowBoy Junkies, Deep Purple live, The Who. Got a Springsteen still need to hear that one.

anyway I have learned quite a bit from this thread about "lossless" and appreciate the geek speak placed into readable understandable text......not likely to make any changes right now in the rig but when I do, I will be informed..........its great to see a thread like this one, instead of whining and calling people as******.

RT1

mmadden28

This thread actually got me kickstarted into finally ripping ALL of my CDs into FLAC. Its been on my TODO list for some time.

In 5 days, I've ripped 108 CD's and counting - its an obsession now to complete the job. I have a ways to go yet...

John30_30

mmadden28 wrote: »

In addition to what i wrote above:
Since EAC uses AccurateRIP and an online database of checksums, and the majority of people are getting the same checksums across a widely varying subset of hardware and OSs, its logical (to me) to think that the EAC method of extraction is fairly accurate, if not consistent, and so probably the most bit for bit accurate to the original CD.

trivia corner:
EAC has been the defacto standard for ripping for the past 10 or so years. It's so anal, MAC users wish they had something that good. FLAC is not the only lossless codec, nor most efficient ( see monkeysaudio ape files) but it's open-source and works under linux with no problem.
I think the 1st actual working lossless codec was Shorten (Shn), which has been pretty much discarded. Most of the popular codecs achieve approx. 50% size compression with no data loss. That is conceded to be about the best we'll ever see, and it's not such a problem anymore what with bigger, cheaper hard-drive and solid-state drives in players.
Other lossless codecs include Windows WMA lossless and Apple's M4A. If you're a big iPod or Zune user, you'd also want to think in terms of what codec runs on your player.

heiney9

Correct

Shorten (SHN) was widely used on many of the first file swapping sites, mostly at that time for trading live bootlegs, etc.

APE files are great as well, I still usually convert those to FLAC. I believe the SQueezebox will play APE files too.

Very few media players play SHN files. I have a lot of Hendrix bootlegs in SHN, those were from a few years ago.

coolsax

how much compression does a the FLAC format do? most cds are approx 780 Megabytes.. how much would a fully recorded CD ripped to FLAC take up? 400MB? 500MB? obviously much more than an albums worth of MP3s which is usually 50-100MB.

bobman1235

Looks like it's almost as high as 50%

http://flac.sourceforge.net/comparison.html

vc69

Yeah, it's variable but about 50% is the best you can hope for. Higher compression equals more CPU intensive decode. Kind of irrelevant as there is plenty of horsepower these days.

coolsax

kewl.. good to know.