x
Stabb 3 January 2004

Hi All,
I wanted to get good ref tracks so I:

Extracted the aiffs from production CDs using Toast.

Then used an audio editor to separate the L and R tracks into 2 mono.

Then loaded the files into my digital recorder (MX-2424) via file transfer (not real time).

Then played back on 2 adjacent channels for stereo (MX is connected to my 02r via TDIF).

My question is:
Do you guys think that I'm playing back what the mastering engineer created with no degradations?

Thanks for your thoughts.
Stabb

I very much appreciate the existence of this site, BIG Thank You to all responsible.

[ January 03, 2004, 02:29 PM: Message edited by: Stabb ]

Comments

MisterBlue Sat, 01/03/2004 - 12:12

Yes and no (how is that for a great answer? :D ).

From the way you describe it you should have a perfect representation of what the mastering engineer released as the final audio CD version.

It is entirely possible, however that the mastering engineer worked with 24-bit formats, higher sample rates or even in the analog domain altogether in which case there might be a small but nevertheless existing difference between what the mastering engineer worked with while doing his/her magic and what finally ends up on CD.

Still, the mastering engineer will of course check the final CD to make sure it is as close as possible to what can be achieved.

Not sure if this answers your question ...

MisterBlue.

Stabb Sat, 01/03/2004 - 12:49

Hi MB,
Thanks for the post.
I was going to say I guess the answer is in the nuts and bolts of the file conversion applications that I used.

But "upon further review", I thought about the variables in what happens after the mastering process and before I get my mitts on the CD.
For example - CD makers being businesses, won't be using the most expensive media they can find, right?

Anyway I just wanted to know if I was missing something in my thought process.

Stabb

Ammitsboel Sat, 01/03/2004 - 13:03

Hi All,
I wanted to get good ref tracks so I:

Extracted the aiffs from production CDs using Toast.

Then used an audio editor to separate the L and R tracks into 2 mono.

Then loaded the files into my digital recorder (MX-2424) via file transfer (not real time).

Then played back on 2 adjacent channels for stereo (MX is connected to my 02r via TDIF).

My question is:
Do you guys think that I'm playing back what the mastering engineer created with no degradations?

Why don't just play the CD on a system exactly the same as the ME that did the master, then you have what he did.

If you what to test on different reference systems or just your own then just play the CD and listen.
You will move more away from what the ME did if you make all those convertions and play it at a second device with another digital out and so on... you will then hear more of your gear that interacts with the original signal. ;)

Best Regards.

Stabb Sat, 01/03/2004 - 13:55

Hi Henrik,
Playing the CD in my player was the start of this.

My player's output was connected to channel inputs of my 02r.
The inputs have what, 20bit ADC's?
Maybe it was my player, maybe it was the cables, maybe it was the 02r's inputs,
but the CD player going into input channels did not sound good.

So I wanted to get master recordings that haven't left the digital domain played from my recorder.

Besides my CD players tray won't open no more!
Thanks for your input.
Stabb

ps - Would it be safe to assume that most mastering engineers use digital connections for their playback?
(Is that a can of worms?)

[ January 04, 2004, 02:23 AM: Message edited by: Stabb ]

MisterBlue Mon, 01/05/2004 - 15:14

If a CD-ROM/RW or whatever extracts a file from audio CD and stores it onto a hard disk you have a perfect copy of the 44.1kHz 16-bit audio file. It is my understanding that not a single bit should be flipped or corrupted as there is extensive error correction and check sum testing involved. These are the same CD reader units that read programs of data disks and software simply doesn't run if even a single byte is corrupted.

Using the digital output of a CD player could lead to different results as some players will use interpolation should they encounter corrupt bytes. This could definitely lead to quality degradation.

In conclusion, I can believe that a "digitally recorded" file is different from a "digitally extracted" one, although they should in theory be exactly the same. From my understanding I would always assume that the "digitally extracted" one is a perfect copy and thus the one for me to use.

Now sending the file through A/D and D/A converters is obviously an entirely different story - these copies can be really messed up. I would not use those as references.

Hope this helps,

MisterBlue.

Michael Fossenkemper Mon, 01/05/2004 - 15:31

I disagree, an extracted file goes through a conversion process. there are two different kinds of error correction, one for data and one for audio. most computer roms are cheap too. you can try it and see for yourself. I thought it would be better myself to extract it until someone, a client, called me up thinking he was hearing a difference. I then did a full day of a/b ing and found that extracted files did not sound as good as digitally dumped files. I found that most mastering engineers I know found the same thing. All I can say is do it both ways and take a listen and decide for yourself. it is most noticable in the upper freqs. they will sound a bit blury and not as detailed.

Michael Fossenkemper Mon, 01/05/2004 - 16:01

I had 3 refs, one was the original file, the other was the same thing burned to cd and then extracted, and the 3rd was the same cd dumped in digitally. Then I set all 3 up and pumped them out of the same D/A and bouned between them. I was a little shocked myself and thought something had to be wrong. i spent all day comparing. maybe someone knows exactly what happens to a file extracted and the effects of the different error corrections that are going on.

MisterBlue Mon, 01/05/2004 - 19:24

Well, if we are trying to find out whether there are real differences between an original recording, a digitally extracted and a digitally recorded audio file from an audio CD we certainly can compare those files byte by byte once they are all on the hard disk. It's a rather quick exercise - I can probably whip up a computer program to do that in a few minutes or so.

Am I still not getting it, Michael?
I don't mean to be a pain ... it's the engineer in me who wants to get to the bottom of this interesting problem ... :roll:

MisterBlue.

falkon2 Tue, 01/06/2004 - 06:21

I've tried this a couple of times... Burn a 16/44 .wav, rerip it (I'm using EAC for PC, by the way - it tells you if *any* bits are misread by the CD drive and interpolated) and run CRCs and difference testing between the original and the rip.

They've been identical all these while. Granted, quite a lot of this stuff is pretty hardware dependant, but if CRC says two files are identical...

Albert Tue, 01/06/2004 - 08:48

My question is:
Do you guys think that I'm playing back what the mastering engineer created with no degradations?

I know you guys are looking for perfection, but isn't the extracted track going to be *good enough* for use as a reference mix? The subtle variations are not going to change the overall mix/master of the album beyond the point where it can be of use as an educational tool, or beyond the point of where it is a fair representation of the work of its creators.

Stabb Tue, 01/06/2004 - 11:16

Thanks all for the great discussion.

What I'm getting is this-

An extraction of an audio file is more of a file 'creation' than a bit for bit copy.
If this is true then it would follow that since all ripping applications are not created equal, their respective extractions will not be equal.

Also a big variable is the amount of errors and error correction - in a real time transfer, file copy or extracting to disk (HD or CD).
(Hey, I used to use cheap audio RCA cables to connect s/pdif - I was told some of my 1's and 0's probably got flipped)

It is likely there are errors embedded in production CDs.

Albert,
I was hoping for perfect bit for bit copies of masters coming out of my recorder.

Now it looks like I might be getting perfect bit for bit copies w/ errors.

In this case I will settle for good enough, but wait a minute -

My CD players' busted, but hmmm, my cheap DVD player has a coax out, I wonder if that will work?

realdynamix Tue, 01/06/2004 - 13:24

Originally posted by Michael Fossenkemper:
...some drives will do an exact copy, errors and all. and others will correct errors. this may be where the differences lie.

:) Ya, technology working for your benefit, like the occasional SERVICE lite flashing in my wife's car. Is something really wrong, or should we pay the 350.00 just to find out it needs an oil change?

Thanks for finding that Michael, I was wondering what and who might solve this mystery.

--Rick

Michael Fossenkemper Tue, 01/06/2004 - 18:59

I'm going to keep my eyes out for any more info on this. The best way I know of telling if two files are identical is the phase reverse test. sample align both files and flip the phase of one. If you have some really sensitive meters, your can see if there are any differences as well as hear them, but don't crank your monitors too loud and blow something.
I also don't agree with good enough, I think it's really important to hear the detail that can be achieved in a good mastering job and try to match that. the difference between a good mastering job and a great mastering job is in the detail.

x

User login