Home of the Squeezebox™ & Transporter® network music players.
Page 6 of 8 FirstFirst ... 45678 LastLast
Results 51 to 60 of 76
  1. #51
    Senior Member
    Join Date
    Apr 2005
    Location
    Watford, UK
    Posts
    1,651
    Quote Originally Posted by SoftwireEngineer View Post
    Exactly, my point and probably that of others.
    Now lets speculate what could be the reason - different players used as transport sound different. Conclusion - jitter.
    Ah, but now you're talking about two-box setups, where the CD player is used only as a transport. Jitter can indeed be an issue here, but it's not due to the transport - it results from the generation of the SPDIF signal and/or transmission through the interconnect and/or in the external DAC's receiving circuit. And exactly the same set of jitter-generating mechanisms exists if the transport is a Squeezebox.

    Quote Originally Posted by SoftwireEngineer View Post
    BTW, where in the Prism paper does it say, jitter is not affected ?
    Jitter is discussed in many places throughout the paper, but if you want to just get the gist of their findings, read the Conclusions section on page 8. Here is a relevant passage:
    "The effects of disc-related or servo-related sampling jitter have NOT been found in either two-box or one-box players. Sampling jitter has been widely cited as a significant artifact in CD players by writers in both the consumer and professional audio fields. It seems possible that the sidebands produced by amplitude modulation may have been mistaken for the sidebands characteristic of sampling jitter modulation"
    Quote Originally Posted by SoftwireEngineer View Post
    Re: your signature line. Recordings that were mixed in analog and finally converted to CD or DSD sound the best to me. Also, sound engineers who use a good quality clock for their ADCs mention that the sound quality is much better. So reducing jitter either during recording or during playback should be a priority for high-fidelity sound reproduction.
    What I'm trying to say in the sig line is that the way many modern CDs are recorded and mastered destroys so much of the music in the first place, that striving for the best possible reproduction seems a little pointless. It's just a bit of a rant against the music industry.
    Transporter -> ATC SCM100A

  2. #52
    Member
    Join Date
    Oct 2006
    Location
    Beautiful Central Oregon
    Posts
    37
    [QUOTE=cliveb;146937]Unfortunately, the article by Steve Nugent that you cite is utter hogwash. His thesis regarding the superiority of computer based playback over CD is founded on the completely erroneous axiom that the master clock in a CD player is in the transport. It isn't.[QUOTE]

    Hogwash eh?

    How many DAC's and transports have you actually opened-up?

    If you had done this as I have, you would have found that ALL Transports have a master clock that is both a PLL and part of a spindle speed control system. This is why I have installed Superclocks in 100's of customers Transports in order to reduce the S/PDIF signal output jitter. It is also the reason why rewriting a CD with a low-jitter writer such as a Reality-Check makes such an improvement in the Transport output jitter. The pits are more accurately located and easier for the Transport to read, and therefore the PLL and master clock are less effected due to less jitter as the pits are being read.

    On the other hand ONLY DAC's that do upsampling typically contain an oscillator or clock. 99.9% of DAC's that do not upsample do not have ANY oscillator at all. They depend entirely on the recovered clock from the S/PDIF, AES or Toslink input in order to function. The "Receiver" chip does this clock recovery. There are a very few exceptions where the master clock is located in the DAC and a "word-clock" is transmitted to the Transport, but these are rare. Most modern DAC chips are actually clocked on the bit-clock, not the word clock as many of the older DAC chips were, so the "word-clock" paradigm just does not work with most modern DAC chips, including: AD1853, PCM1702, CS4396/7 and most others.

    Perhaps it is the nomenclature of "Transport" and "DAC" that has you confused. In a CDP, they are of course combined.

    Steve N.
    Empirical Audio
    Manufacturer
    Last edited by audioengr; 2006-10-18 at 12:11.

  3. #53
    Steve just to clarify one thing in your response. Is the situation exactly the same in one box CD players as it is in two box transport/DAC? That is, the master clock that controls spindle speed and PLL is located in the transport circuitry, and not in the DAC chip?

    Thanks.

  4. #54
    Member
    Join Date
    Oct 2006
    Location
    Beautiful Central Oregon
    Posts
    37
    Quote Originally Posted by boybees View Post
    Steve just to clarify one thing in your response. Is the situation exactly the same in one box CD players as it is in two box transport/DAC? That is, the master clock that controls spindle speed and PLL is located in the transport circuitry, and not in the DAC chip?

    Thanks.
    Every one that I have seen from Sony, Pioneer, Denon, Musical Fidelity, Electrocompaniet, Shanling, Samsung, Philips ... need I go on?? CDP's dont always generate a S/PDIF encoded signal, but the I2S clocks are derived from the Transport section, not the DAC.

    It is true that many portable CD players buffer the data and spool it out to eliminate the effect of shock and vibration. These have an additional clock to read-out the buffer and the transport is actually like a CD-ROM drive, reading the data at much higher speed than the native speed. It's too bad that these designs are not used in more high-end transports. There are a few "computer-type" CD players out now, notably from Cary and Meridian that are actually CD-ROM drives and computers disguised to look and behave like a CD player.

    Steve N.
    Last edited by audioengr; 2006-10-18 at 14:48.

  5. #55
    Senior Member opaqueice's Avatar
    Join Date
    Feb 2006
    Location
    A place where something is or could be located; a site.
    Posts
    1,815
    If you had done this as I have, you would have found that ALL Transports have a master clock that is both a PLL and part of a spindle speed control system. This is why I have installed Superclocks in 100's of customers Transports in order to reduce the S/PDIF signal output jitter. It is also the reason why rewriting a CD with a low-jitter writer such as a Reality-Check makes such an improvement in the Transport output jitter. The pits are more accurately located and easier for the Transport to read, and therefore the PLL and master clock are less effected due to less jitter as the pits are being read.
    Sorry, I don't understand what you're saying. I thought cliveb's point was the following:

    the bits recorded on a CD bear little resemblance to the bits in the S/PDIF stream, because of interleaving and error-checking codes.

    Therefore, the bits must first be read into a buffer, after which they are processed by some chip which decodes them to S/PDIF and sends them to a DAC chip or a digital out.

    Now, the point was that, since the bits are coming from a buffer and not directly from the CD, there is no real difference between a CD player and a solid state player like the SB. Of course there can be some jitter going into the DAC, or in the DAC clock itself, but that's got nothing to do with the CD (modulo some form of electrical interference caused by the moving parts).

    You're saying that the same oscillator controls the speed of the CD spinning and the DAC, but.... so what? That doesn't seem to invalidate his point at all.

  6. #56
    Senior Member
    Join Date
    Apr 2005
    Location
    Watford, UK
    Posts
    1,651
    Quote Originally Posted by audioengr View Post
    Quote Originally Posted by cliveb View Post
    Unfortunately, the article by Steve Nugent that you cite is utter hogwash. His thesis regarding the superiority of computer based playback over CD is founded on the completely erroneous axiom that the master clock in a CD player is in the transport. It isn't.
    Hogwash eh?

    How many DAC's and transports have you actually opened-up?

    If you had done this as I have, you would have found that ALL Transports have a master clock that is both a PLL and part of a spindle speed control system. This is why I have installed Superclocks in 100's of customers Transports in order to reduce the S/PDIF signal output jitter.
    Of course the clock is part of the spindle control system. That's the point: the clock determines the spindle speed, not the other way round.

    The fact that you've modified hundreds of CD players with different clocks doesn't make any difference to the fact that your article is misleading in the extreme. You say that there are a number of sources of jitter, and include in that list the pits on the CD itself and the reading of those pits by the laser. Nowhere in your article can I find any mention of the fact that these jitter sources are completely irrelevant, since the data that comes off the CD is buffered and then clocked out by the master clock.

    The point is that in any digital source, the samples are fed into a buffer and then sent to the DAC (or SPDIF output) under the control of a free-running clock. The buffer DECOUPLES the DAC (or SPDIF transmitter) from the vagaries of the process which supplies the data. On the source side of the buffer, it doesn't matter whether the data is supplied by reading a spinning optical disc with a laser, receiving a network data stream (at the other end of which is a spinning hard disk read by magnetic heads, of course), or a talented morse-code operator with very fast fingers.

    Your article is clearly worded such that the reader will infer that the jitter in the CD's pits and resulting from reading them with the laser are transmitted through to the DAC (or SPDIF transmitter). This is simply not true. [edit - removed a comment based on my incorrect belief that the Reality Check CD copying device was an Empirical Audio product - I have now discovered that it is from Digital Systems & Solutions, so the suspicions I originally voiced are not valid, and I apologise for them]

    I may have muddied the water a little when I said the the master clock is not in the transport. In this context (a single box CD player), by "transport" I am referring to the opto/mechanical disc reading system. It occurs to me that many people understand the word "transport" to refer to a device which emits an SPDIF signal to an external DAC. In that context, then yes, the master clock is in the "transport", but it's still on the DAC side of the buffer.

    Quote Originally Posted by audioengr View Post
    It is also the reason why rewriting a CD with a low-jitter writer such as a Reality-Check makes such an improvement in the Transport output jitter. The pits are more accurately located and easier for the Transport to read, and therefore the PLL and master clock are less effected due to less jitter as the pits are being read.
    The hypothesis that making CDs easier to read has an effect on the jitter of the master clock has been shown to be unfounded in the Prism/DCA paper referenced earlier in this thread.
    Last edited by cliveb; 2006-10-19 at 02:47.
    Transporter -> ATC SCM100A

  7. #57
    Junior Member
    Join Date
    May 2006
    Posts
    24

    to summarize...

    Thanks to everyone for responding to my thread. I've learned a lot, and there has been relatively little flaming. Here are my thoughts:

    1) We need to be careful about what we are talking about. My original question was basically if there is a fundamental difference between single box optical CD players vs. network/hard disk players like the SB/Transporter (with the hypothesis that it was due to a lack of sufficient buffer to eliminate problems with jitter and errors). So points made about traditional CDPs as "transports" feeding external DACs are off topic and a whole other issue (given that you then have a SP/DIF cable, no clock being transmitted, etc., jitter can easily be introduced in that case).

    2) The points made by Clive and others that the buffering of the bits coming off the optical disc and how they are clocked into the internal DAC of the CDP seem basically valid. I.e. that the bits are clocked out of the buffer by the master clock of the CDP into its DAC must basically eliminate any jitter from the optical drive, ASSUMING the clock is solid, and there are no interactions between the optical drive and the jitter of the clock.

    3) Sean asserts that there are NO fundamental differences between an optically based player and a network/HD player like the SB. I see where he is coming from in one sense. The bits coming out of the optical player's buffer ought to be the same as the ones coming from the network, IF you exclude errors and error correction errors. This is because, as noted above, the stream of bits coming out of the optical buffer are controlled by the master clock that is also driving the DAC.

    4) Nevertheless, my sense is still that there are often signficant sonic problems related to optical drives inside of single box CDPs. Perhaps not inherent, and perhaps high-end CDP makers get around the problems by including separate power supplies for motors and audio circuits, using data drives to have better error correction (which then requires bigger buffers), using higher quality DACs and clocks, etc. But still, my gut feeling is that there are a set of complex engineering compromises when working with a real-time delivery (i.e. not CD_ROM data) optical drives that **usually** compromise the sonics.

    4a) I say this for two reasons. (a) my friends and my (admitedly unscientific) listening experience is that the SB has (while not always better overall) a reliably more pleasant, musical, and different sound than most CDPs (insert your favorite audio descriptors here: relaxed, focused, mo-better bass, smoother, blah, blah, blah), **especially for the money**. I know this will not satisfy the bits-are-bits folks, but my experience working in recording studios is that the ear/brain system is an extremely sensitive (though often unrealiable--hence the need for blind testing) measurement system that can discover differences that can't initially be explained in engineering terms. I think many here will remember when CD players first came out and people who said they didn't think the sound was that great compared to vinyl were ridiculed because CD audio was "perfect". My ears tell me something is wrong with traditional CDPs.

    4b) Second, the whole optical real-time delivery system seems like kind of a mess. It's mechanical, it puts EMI producing, variable speed motors next to analog electronics, it is prone to data errors, it puts a strain on the power supply, etc. (I'm sure those who know more about the insides of this system can describe other problems). I think we've learned that both bits and analog audio are pretty fragile, and many factors can affect both. So moving all that optical junk off to an external computer (and eliminating the real-time getting bits off a crappy optical disc delivery problem completely), seems like a good thing. If nothing else, it makes for a simpler engineering problem and fewer compromises in a box like an SB.

    5) Of course, there are all the convienience advantages, which also increase the pleasure of listening to music.

    Thanks again to everyone!

    .phil

  8. #58
    Member
    Join Date
    Oct 2006
    Location
    Beautiful Central Oregon
    Posts
    37
    Quote Originally Posted by opaqueice View Post
    Sorry, I don't understand what you're saying. I thought cliveb's point was the following:

    the bits recorded on a CD bear little resemblance to the bits in the S/PDIF stream, because of interleaving and error-checking codes.

    Therefore, the bits must first be read into a buffer, after which they are processed by some chip which decodes them to S/PDIF and sends them to a DAC chip or a digital out.

    Now, the point was that, since the bits are coming from a buffer and not directly from the CD, there is no real difference between a CD player and a solid state player like the SB. Of course there can be some jitter going into the DAC, or in the DAC clock itself, but that's got nothing to do with the CD (modulo some form of electrical interference caused by the moving parts).

    You're saying that the same oscillator controls the speed of the CD spinning and the DAC, but.... so what? That doesn't seem to invalidate his point at all.

    My point is that the buffer that is doing the decoding of the bits on the disk is clocked by the PLL clock that is influenced by the jitter on the disk. The clock that clocks the buffer input is identical to the clock that clocks the buffers output. It is one PLL clock and it gets jittery from the variable spacing of the bits on the disk. This buffer is only a holding place in order to decode the bits. It is NOT an elastic buffer for reclocking the bits.

    There is only one PLL that locks to the bit-stream. If the frequency of this clock varies much from the bitstream frequency, there would have to be a very large buffer to take up these short-term frequency differences to prevent overrun or underrun. Therefore, the PLL frequency must track the bit frequency quite closely.

    No matter how you look at it, if there is one PLL and it must track the bit-rate coming off the disk, then it will be affected to some extent by the jitter in the pits and motor speed-variation, loop-control, mechanical resonance of the disk spinning etc...

    Unless you can completely decouple the disk bitstream from the clock that clocks-out the S/PDIF data stream, there will always be jitter in the output signal due to the disk and its mechanical/electrical systems.

    The players that effectively do this decoupling read from the disk at high speed into an elastic buffer. Then it is basically a read from RAM. Both shock-proof portables and computer-based units can do this.

    Steve N.
    Empirical Audio
    Last edited by audioengr; 2006-10-19 at 11:55.

  9. #59
    Member
    Join Date
    Oct 2006
    Location
    Beautiful Central Oregon
    Posts
    37
    "I may have muddied the water a little when I said the the master clock is not in the transport. In this context (a single box CD player), by "transport" I am referring to the opto/mechanical disc reading system. It occurs to me that many people understand the word "transport" to refer to a device which emits an SPDIF signal to an external DAC. In that context, then yes, the master clock is in the "transport", but it's still on the DAC side of the buffer."

    No it's not. The "master clock" is the PLL clock that controls the spindle motor speed and therefore the bit frequency from the disk. This has nothing to do with the DAC.

    Steve N.

  10. #60
    Member
    Join Date
    Oct 2006
    Location
    Beautiful Central Oregon
    Posts
    37
    "The hypothesis that making CDs easier to read has an effect on the jitter of the master clock has been shown to be unfounded in the Prism/DCA paper referenced earlier in this thread."

    This paper is not relevant IMO. Anyone can write a paper and publish it on the web. Does the author have 30 years digital design experience? Was he a design lead on the Pentium 2?

    Steve N.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •