MPC 4000 with adat card & Protools (who has this set up I need u to test something)

Guys, if anyone here has an MPC 4k with adat card and protools
please run the following test for me and tell me your results.


Take a file straight from CD (say Rkelly or something)
IMPORT this into PTOOLS

Save as a wav. file and load it into the MPC

Record via the Adat outs on the 4k back to protools

Is there a sound difference? Are they the same?

[ILL]

I don't have a 4k but my guess is that since it's all digital transfer there would be absolutely no change in sound.

Well duhhh I coulda guessed that but it wouldnt make me sure of what each audio engine does to the sound

[TonyBelmont]

It's quite possible there could be a change in sound. Although from one pass it would likely be minimal.. The adat protocol "can" change the sound and suffers from it's own demons (clocking issues, jitter, etc.). Make ten passes with the same piece of audio back and forth and tell me if you hear a difference.

Although, with a good clock it would be minimal.

I only play to make one pass when tracking a song so that would be cool if its a minimal difference that only shows after several generations

[dussel]

Originally Posted by TonyBelmont

It's quite possible there could be a change in sound. Although from one pass it would likely be minimal.. The adat protocol "can" change the sound and suffers from it's own demons (clocking issues, jitter, etc.). Make ten passes with the same piece of audio back and forth and tell me if you hear a difference.

Although, with a good clock it would be minimal.

No Tony, this is not true! I do not know how it is with the MPC (although I got one :-)) But ADAT Itself gives a 100% identical Bit 4 Bit copy of the signal. The Clocking and Jitter und so on is only an issue if you convert on the way. If you do not - digital copies are perfect!

regards
Dussel

[TonyBelmont]

Originally Posted by dussel

No Tony, this is not true! I do not know how it is with the MPC (although I got one :-)) But ADAT Itself gives a 100% identical Bit 4 Bit copy of the signal. The Clocking and Jitter und so on is only an issue if you convert on the way. If you do not - digital copies are perfect!

regards
Dussel

Actually, in researching it and reading through some AES papers... It seems all digital transfers can suffer from a degredation in the sound. But, it's typically not noticeable unless you do multiple passes. Or, so the literature I've read states.

But, clocking and jitter can most definitely effect the sound over the lightpipe format. It's not only when you convert from A/D. It's actually whenever you make a digital connection between gear. That's why word clock is used when you connect one digital piece of gear to the next.

[dussel]

Hello Tony,

Can you point me at some papers stating this? I am almost shure there must be some degradation (as with data on computer harddisks) due to the physics of the lightpipe and the emitter/reciver circuits. But the probability of a bit beeing transmitted wrongly is probably extremely low. Apart from that I think if the recieving clock is able to lock on the frequency of the sender, there should be no loss in digital signal transmission. (Otherwise you would not be able to read this text right?).

Even if there was a loss in ADAT pipes, the result of a wrong or lost bit would probably be clearly audible (We all know the sound of a D/A that is not clocked correctly).

But maybe I miss something. I would be very interested in your sources.

best regards
Dussel

[TonyBelmont]

Originally Posted by dussel

Hello Tony,

Can you point me at some papers stating this? I am almost shure there must be some degradation (as with data on computer harddisks) due to the physics of the lightpipe and the emitter/reciver circuits. But the probability of a bit beeing transmitted wrongly is probably extremely low. Apart from that I think if the recieving clock is able to lock on the frequency of the sender, there should be no loss in digital signal transmission. (Otherwise you would not be able to read this text right?).

Even if there was a loss in ADAT pipes, the result of a wrong or lost bit would probably be clearly audible (We all know the sound of a D/A that is not clocked correctly).

But maybe I miss something. I would be very interested in your sources.

best regards
Dussel

We aren't talking about missing bits of audio... We are talking about the degredation of the signal due to jitter, etc. This has been covered in several AES technical papers over the years. I believe they have an online library..

[dussel]

Hello Tony,

I think either you or I get something wrong. I have read more or less every AES Paper on D/A and A/D Conversion I could get my hands on in the last two years and from what I understand Jitter is only an issue when a signal conversion is involved.

As far as I understand Jitter is the effect produced by small changes in the clock frequency. As we "reclock" our digital Data each time we play it back (there is no timing information stored, just the bits encoding single samples). Timing is no issue as long as every bit is recieved correctly. This is different if we convert the recieved digital signal 2 analog with the jittery timing. The analog signal "stores" the timing of the samples.

regards
Dussel

[7ciel]

Originally Posted by dussel

from what I understand Jitter is only an issue when a signal conversion is involved.

That's what I understood of it as well. I don't get how you can have a jitter phenomen when staying in the digital realm.

[ILL]

You can't, jitter is just a signal conversion term. When you are running something trough an ADAT lightpipe it's like copying from one hard disk to another.

Damn I didnt mean to start a war. LOL would someone test it and let me know the answer to the original ?

Despite whether or not lite pipe has an effect, what ffx does the audio engine have.

AND let's NOT Assume it's nothing. Try doing the same thing in reason and tell me it's not effect

[TonyBelmont]

Originally Posted by ILL

You can't, jitter is just a signal conversion term. When you are running something trough an ADAT lightpipe it's like copying from one hard disk to another.

It's not like copying for one hard disk to another... It suffers from degradation, but you wouldn't notice until after multiple passes. It has something to do with the clock/jitter... I can't remember exactly what the reasoning was off the top of my head, but it was in an AES tech note from 2001 or 2002.. On a side note, you will probably never notice unless you do 10 or 20 passes, so I doubt it matters. But, if you want to try it out, go for it..

There was also another AES paper that talked about how digital copies of CD's suffer from degradation after multiple generations of the same disk are copied (I'm not making this up)...

[dussel]

Originally Posted by TonyBelmont

It's not like copying for one hard disk to another... It suffers from degradation, but you wouldn't notice until after multiple passes. It has something to do with the clock/jitter... I can't remember exactly what the reasoning was off the top of my head, but it was in an AES tech note from 2001 or 2002.. On a side note, you will probably never notice unless you do 10 or 20 passes, so I doubt it matters. But, if you want to try it out, go for it..

There was also another AES paper that talked about how digital copies of CD's suffer from degradation after multiple generations of the same disk are copied (I'm not making this up)...

Oh, I believe you, but it would be nice to know why. Maybe we can get one of the experts (Nika, Dan ....) to chime in?

As far as I understand it there is only one possibility of signal degradation from

11010101111000011010101 What the File in the Computer reads.

to

11010101111000011010101 What the recorded File in the MPC reads.

That would be one or several Bit Flip. There is no reason to believe it would always happen to low bits so you would definitely hear it in the first pass.

To answer the initial Question I will do the phase test with my MPC this weekend (I have a MP4k w. ADAT and an RME HDSP9652).

regards
Peter

P.S. The CD example is a different Story because there is error correction going on and 1to1 copies pass the errors on. Whereas in our example corrected errors do not appear in the recorded Data.

BUMP

[dussel]

Originally Posted by no ssl yet

BUMP

oh yes, I almost forgot the problem. I will do the experiment this week. I had a really busy time - sorry.

regards
Dussel