R/E/P Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 2 [All]   Go Down

Author Topic: CD Players transport differences?  (Read 11049 times)

pipelineaudio

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
CD Players transport differences?
« on: September 08, 2006, 03:34:25 PM »

I am hearing there are differences in sound quality between transports.

I call shennanigans on this and here's my reason

I can take a file, zip it, and burn it on a 20 dollar cd burner

If I open that file on another PC, thru its own 20 dollar cd burner, I can unzip it and it will open and run.

SO unless even modern CD players are trying to play in realtime instead of buffering, I dont see how the transport itself should make any difference. Can someone help?

Sahib

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 429
Re: CD Players transport differences?
« Reply #1 on: September 08, 2006, 04:43:21 PM »

I worked with numbers of CD mechanisms. There is not much to say about this.

There is no relation between the sound quality and the mechanics of the transport system. Also reading the CD directly or through a buffer is irrelevant. You are reading a digital data. I is one 0 is zero.

Cemal

Logged

pipelineaudio

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
Re: CD Players transport differences?
« Reply #2 on: September 08, 2006, 05:29:51 PM »

Sahib I think the question is, IF there are going to be errors. Lets say the cd player plays back at 1X speed in realtime, with a very short buffer. In this cae could there be less errors between transports?

I'm just trying to determine if there is ANY possible mechanism of difference to the output audio between a 20 dollar transport and a 6000 dollar transport, if they are feeding the same DAC

Ronny

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2739
Re: CD Players transport differences?
« Reply #3 on: September 08, 2006, 08:23:31 PM »

pipelineaudio wrote on Fri, 08 September 2006 17:29

Sahib I think the question is, IF there are going to be errors. Lets say the cd player plays back at 1X speed in realtime, with a very short buffer. In this cae could there be less errors between transports?

I'm just trying to determine if there is ANY possible mechanism of difference to the output audio between a 20 dollar transport and a 6000 dollar transport, if they are feeding the same DAC



No not in the sense of tonal and dynamic sonic degradation. The cheaper burners and readers, cd player etc. can have focus and tracking anomalies directly related to the mechanics of the device, that will give more errors on the write and read than a high end burner or player, but errors represent a lack of information, not an altering of the positions of ones and zeros, as Cemal puts it, that it would take to affect sound field, stereo image, tonal balance or dynamic levels. So you may experience dropouts and skips, audio jittering (not to be confused with line jitter), but not degradation associated with say analog generational loss of frequencies. IOW, errors on a cd playback are just a lack of tonal and dynamic information not an aleration of these elements and you can have 10,000 CU's on a cd and still not hear any audible difference if they are spread out far enough, so that the uncorrectible errors pass on the timeline faster than the human can perceive them.  
Logged
------Ronny Morris - Digitak Mastering------
---------http://digitakmastering.com---------
----------Powered By Experience-------------
-------------Driven To Perfection---------------

pipelineaudio

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
Re: CD Players transport differences?
« Reply #4 on: September 08, 2006, 08:46:08 PM »

shouldnt they just buffer largely anyway so like on a cd holding computer programs, it could be surely played with NO errors? Or is that wishful thinking

Sahib

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 429
Re: CD Players transport differences?
« Reply #5 on: September 09, 2006, 05:17:51 AM »

Hi Aaron,

I support Ronny's reply on the sonic issue.

In terms of increasing the buffer capacity, it would not make much difference unless you move into complete hard disk system.

First you have the influence of the overall time frame that the transport mechanism operates in; circular motion of the CD and the linear motion of the reading head hence tracking.

You also have the influence of the time frame of continues operation of data reading - error checking - storing - putting out to play - going back to reading - error checking etc.

Therefore whichever way you look at it you need a certain time frame. Even if you work on one song at a time buffering, then you still have to read the entire data content of the song, error check that entire data and put it out and that again takes a long time.


Logged

Jon Hodgson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1854
Re: CD Players transport differences?
« Reply #6 on: September 10, 2006, 04:44:27 AM »

pipelineaudio wrote on Sat, 09 September 2006 01:46

shouldnt they just buffer largely anyway so like on a cd holding computer programs, it could be surely played with NO errors? Or is that wishful thinking


It is wishful thinking because there is insufficient redundancy for that. The error correction in Red Book Audio was specced on the basis that a few errors here and there which are detected but cannot be corrected are ok, since you can use interpolation between the previous and next samples.

Data CDs require zero errors, so they have an extra level of error correction on there, which is why a standard length CD holds 650MB, and not over 700 as you might expect if you calculate the amount of data in 70 minutes of audio.
Logged

mdemeyer

  • Newbie
  • *
  • Offline Offline
  • Posts: 41
Re: CD Players transport differences?
« Reply #7 on: September 12, 2006, 12:48:50 AM »

The original post didn't say (clearly, at least to me) if he was talking about difference in CD playback when the transport was driving a DAC directly or some kind of PC setup.  If the former, the DAC clock is being picked up from the S/PDIF (or AES/EBU) signal and, depending on the implementation, the signal from the transport can impact the sound not because it has bad data, but because of jitter in the recovered clock.  This is a well known and measurable item.

Want to clarify the question?
Logged

pipelineaudio

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 379
Re: CD Players transport differences?
« Reply #8 on: September 12, 2006, 03:02:52 AM »

the transport would be driving the dac directly.

Jon Hodgson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1854
Re: CD Players transport differences?
« Reply #9 on: September 12, 2006, 03:30:14 AM »

mdemeyer wrote on Tue, 12 September 2006 05:48

The original post didn't say (clearly, at least to me) if he was talking about difference in CD playback when the transport was driving a DAC directly or some kind of PC setup.  If the former, the DAC clock is being picked up from the S/PDIF (or AES/EBU) signal and, depending on the implementation, the signal from the transport can impact the sound not because it has bad data, but because of jitter in the recovered clock.  This is a well known and measurable item.

Want to clarify the question?


Even if the transport is your master you don't use the clock recovered from the disc itself as your master, that would be foolish since it is dependent not only on any imperfections in the CD, but also on any variation in spin speed... we'd be back in the world of wow and flutter.

Instead you use a nice solid crystal clock as your master and synchronize the mechanism to that, at least that's what we did on the mechanism I worked on and I cannot se any sensible reason to do it differently. So it is  lot less variable than some people seem to think.

Though I can imagine that depending on the quality of that clock, the circuitry to drive the interface, and the quality of the receiver and PLL in the DAC you might be able to detect a difference... but I don't know how bad each would have to be to be noticable (Dan probably does), and therefore how likely it is.

It is theoretically possible for different transports to sound different, they could have differing quality of data reading off the disc, leading to more unrecoverable errors, they could have differing quality of interpolation for those errors (linear, polynomial) and they could possibly have enough variability on their clock to manifest itself in the DAC, especially one with a poor PLL.

However I have to note that firstly things would probably have to get pretty bad before you noticed these things, with the exception of the jitter most errors will most likely be few and far between and be gone in 20 microseconds. Secondly most in most audiophile reviews I've seen of these things the claims made as to what differences they can hear between transports make absolutely no sense to me considering what the nature of such errors would be and how they might manifest themselves audibly, which leads me to believe that mostly they are just self delusion by the reviewers.. with a thought process something like "It's a 20 thousand dollar mechanism with a marble base and platinum wiring, it must be better, and with my golden ears I'm sure I can pick it uy.... oh yes, there it is, the third violin is coming through better in this 300 piece orchestra, I'm sure of it"


Logged

Sin x/x

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 72
Re: CD Players transport differences?
« Reply #10 on: September 12, 2006, 07:26:15 AM »

mdemeyer wrote on Mon, 11 September 2006 23:48

The original post didn't say (clearly, at least to me) if he was talking about difference in CD playback when the transport was driving a DAC directly or some kind of PC setup.  If the former, the DAC clock is being picked up from the S/PDIF (or AES/EBU) signal and, depending on the implementation, the signal from the transport can impact the sound not because it has bad data, but because of jitter in the recovered clock.  This is a well known and measurable item.

Want to clarify the question?


http://recforums.prosoundweb.com/index.php/m/180621/0/#msg_1 80621

The quality of digital audio depends only on the quality of the converters.
If you hear sound difference between cd-transports, there's a design fault in the dac.
Logged

Jon Hodgson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1854
Re: CD Players transport differences?
« Reply #11 on: September 12, 2006, 08:49:47 AM »

Sin x/x wrote on Tue, 12 September 2006 12:26


The quality of digital audio depends only on the quality of the converters.
If you hear sound difference between cd-transports, there's a design fault in the dac.


Dan has a good paper on jitter and phase locked loops on his website.

A perfect transport would have zero jitter on the output.

A perfect PLL would have 100% jitter rejection, a perfect PLL is not easy, which is why Lavry products have CrystalLock.

If you can hear a difference due to jitter on the output of the transport, then they are BOTH imperfect.

And that is only covering the subject of jitter, if the transport is inferior enough that it has more read errors, and/or the interpolation method is inferior, then you might also hear a sound difference.... though to be honest I suspect it's a fairly moot point, even cheap mechanisms can usually manage 100% data recovery once combined with the extra level of redundancy which is used on a data CD, which indicates that their read errors are kept under control.

What we need is an "unrecoverable error" counter on the transport, then we could see how different they really are.


Logged

Sin x/x

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 72
Re: CD Players transport differences?
« Reply #12 on: September 12, 2006, 01:40:58 PM »

Jon Hodgson wrote on Tue, 12 September 2006 07:49

Sin x/x wrote on Tue, 12 September 2006 12:26


The quality of digital audio depends only on the quality of the converters.
If you hear sound difference between cd-transports, there's a design fault in the dac.


Dan has a good paper on jitter and phase locked loops on his website.



I can't find the paper on jitter and phase locked loops on Mr.Lavry's website.

Jon Hodgson wrote on Tue, 12 September 2006 07:49


A perfect transport would have zero jitter on the output.

A perfect PLL would have 100% jitter rejection, a perfect PLL is not easy, which is why Lavry products have CrystalLock.

If you can hear a difference due to jitter on the output of the transport, then they are BOTH imperfect.



No, only the pll's in the converter needs to be good enough. Then all jitter is rejected.

Jon Hodgson wrote on Tue, 12 September 2006 07:49


And that is only covering the subject of jitter, if the transport is inferior enough that it has more read errors, and/or the interpolation method is inferior, then you might also hear a sound difference.... though to be honest I suspect it's a fairly moot point, even cheap mechanisms can usually manage 100% data recovery once combined with the extra level of redundancy which is used on a data CD, which indicates that their read errors are kept under control.



If the read errors are within a certain range, or don't exceed a certain range, then all errors are corrected.

Logged

Jon Hodgson

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1854
Re: CD Players transport differences?
« Reply #13 on: September 12, 2006, 02:13:57 PM »

Sin x/x wrote on Tue, 12 September 2006 18:40

Jon Hodgson wrote on Tue, 12 September 2006 07:49

Sin x/x wrote on Tue, 12 September 2006 12:26


The quality of digital audio depends only on the quality of the converters.
If you hear sound difference between cd-transports, there's a design fault in the dac.


Dan has a good paper on jitter and phase locked loops on his website.



I can't find the paper on jitter and phase locked loops on Mr.Lavry's website.


Look on the support page, "White Papers"->"On Jitter"
Quote:

Jon Hodgson wrote on Tue, 12 September 2006 07:49


A perfect transport would have zero jitter on the output.

A perfect PLL would have 100% jitter rejection, a perfect PLL is not easy, which is why Lavry products have CrystalLock.

If you can hear a difference due to jitter on the output of the transport, then they are BOTH imperfect.



No, only the pll's in the converter needs to be good enough. Then all jitter is rejected.


And how does that contradict what I just said? For you to hear a difference caused by jitter requires that there be jitter in the first place, AND that your converter does not reject it completely.

However if Dan's paper is correct then rejecting ALL jitter is not as easy or as common as you seem to think. In fact if it was then Dan wouldn't be so adamant about using the clock in the ADC as the master whenever possible, he'd just be saying "clock it from whatever you want, the PLL will reject all jitter anyway".

Quote:

Jon Hodgson wrote on Tue, 12 September 2006 07:49


And that is only covering the subject of jitter, if the transport is inferior enough that it has more read errors, and/or the interpolation method is inferior, then you might also hear a sound difference.... though to be honest I suspect it's a fairly moot point, even cheap mechanisms can usually manage 100% data recovery once combined with the extra level of redundancy which is used on a data CD, which indicates that their read errors are kept under control.



If the read errors are within a certain range, or don't exceed a certain range, then all errors are corrected.


Quite true, however this does not change the fact that it is possible for different transports to have different rates of unrecoverable errors given the same CD. For example a few years ago many CD transports were very sensitive when it came to Writeable CDs, often a writeable CD would play in one machine but not in another. These days they are a lot more tolerant.

Having said all that, I suspect that if we actually quantified these differences with most modern CD transports then they would probably be insignificant.
Logged

Ronny

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2739
Re: CD Players transport differences?
« Reply #14 on: September 12, 2006, 03:52:55 PM »



Yes, but is any converter actually totally 100% jitter free? Ala what we old timers used to call the holy DAC. Jitter immunized maybe, jitter free, I'm not so sure. If it's not audible but still measurable, that doesn't make it a 100% elimination of all jitter. For all practical audio purposes the reduction of jitter well below audible should be sufficient, should it not?
Logged
------Ronny Morris - Digitak Mastering------
---------http://digitakmastering.com---------
----------Powered By Experience-------------
-------------Driven To Perfection---------------

Sahib

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 429
Re: CD Players transport differences?
« Reply #15 on: September 12, 2006, 05:33:23 PM »


I actually picked up a Leader CD jitter meter from e-bay about six months ago. But never tried it. It looks pretty straight forward and this could be a good time to get to learn it. I'll tell the wife I have urgent work to do over the weekend and spend some time with it. Anybody has used one before?
Logged

danlavry

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 997
Re: CD Players transport differences?
« Reply #16 on: September 12, 2006, 08:21:53 PM »

Ronny wrote on Tue, 12 September 2006 20:52



Yes, but is any converter actually totally 100% jitter free? Ala what we old timers used to call the holy DAC. Jitter immunized maybe, jitter free, I'm not so sure. If it's not audible but still measurable, that doesn't make it a 100% elimination of all jitter. For all practical audio purposes the reduction of jitter well below audible should be sufficient, should it not?


Yes, if you can get jitter to be low enough, but what is low enough? The problem is that the ear can hear very low jitter, and it is not all that easy to accommodate such low jitter levels.

BTW, when talking about jitter, it is too simplistic to think of one number. The ear sensitivity varies greatly when you change the frequency content of the jitter.

regards
Dan Lavry
http://www.lavryengineering.com
Logged

Sin x/x

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 72
Re: CD Players transport differences?
« Reply #17 on: September 13, 2006, 02:09:04 AM »

Jon Hodgson wrote on Tue, 12 September 2006 13:13

Sin x/x wrote on Tue, 12 September 2006 18:40

Jon Hodgson wrote on Tue, 12 September 2006 07:49

Sin x/x wrote on Tue, 12 September 2006 12:26


The quality of digital audio depends only on the quality of the converters.
If you hear sound difference between cd-transports, there's a design fault in the dac.


Dan has a good paper on jitter and phase locked loops on his website.



I can't find the paper on jitter and phase locked loops on Mr.Lavry's website.


Look on the support page, "White Papers"->"On Jitter"
Quote:

Jon Hodgson wrote on Tue, 12 September 2006 07:49


A perfect transport would have zero jitter on the output.

A perfect PLL would have 100% jitter rejection, a perfect PLL is not easy, which is why Lavry products have CrystalLock.

If you can hear a difference due to jitter on the output of the transport, then they are BOTH imperfect.



No, only the pll's in the converter needs to be good enough. Then all jitter is rejected.


And how does that contradict what I just said? For you to hear a difference caused by jitter requires that there be jitter in the first place, AND that your converter does not reject it completely.




We don't disagree.
But if you hear a difference between digital cables/transports, don't blame the cables/transports.

Thanks for showing me Mr. Lavry's paper on jitter.
As far as my knowledge goes it's truthfully.

It also agrees with my statement: If the dac is properly designed, all jitter is rejected.



So if a converter reacts to different cables, transports or needs an external jitter free master clock.
Throw it in the bin.





P.s. Put the "Proper word clock implementation" thread back where it belongs.


Logged
Pages: 1 2 [All]   Go Up
 

Site Hosted By Ashdown Technologies, Inc.

Page created in 0.091 seconds with 20 queries.