Home of the Squeezebox™ & Transporter® network music players.
Page 2 of 6 FirstFirst 1234 ... LastLast
Results 11 to 20 of 54

Thread: SPDIF is evil

  1. #11
    Senior Member Pneumonic's Avatar
    Join Date
    Feb 2007
    Location
    Toronto, Canada
    Posts
    137
    I too think we are stuck with SPDIF which is unfortunate. It appears the two remedies are to employ an external clock to hopefully counter the woes that the flawed SPDIF protocol may impart on the sound you hear or to put everything into one box in order to avoid SPDIF in the first place.
    Main: Acer Aspire One netbook --> Squeezebox Transporter (slave mode) --> Lessloss 2004 mkII Pre/DAC (master mode) --> Sanders Magtech stereo/Innersound ESL800 mono power amps --> Sanders 10c active speakers

  2. #12
    Member
    Join Date
    Mar 2006
    Location
    Seattle
    Posts
    78

    Holy Clocks Batman!

    My big-rig setup includes a hacked SB3 with all analog circuitry removed that has been converted entirely to a bit-server. Its secondary clock has been removed and the master clock bypassed to an additional connector on the outside of the unit. It cannot run without an outboard clock signal, and was modded with my Lessloss DAC as clock.

    The Lessloss unit can operate in 'slave' mode, i.e. normal spdif mode, or it can act as clock-master with 'data in' and 'clock out'. The DAC replaces the transport superclock signal, these guys are WAY into the question of clocking,jitter, and solutions:

    http://www.lessloss.com/page.html?id=41

    http://www.lessloss.com/page.html?id=33

    http://www.lessloss.com/page.html?id=42

    It takes two digital cables to run, one for data and the other for clock, so a little more complex. The SB3 mod was pretty cheap, a couple hundred bucks of a modder's time - that's the good news. The bad news is the DAC isn't cheap, about $4,500 today. Hey this IS the audiophile forum.

    The results are quite special. I've had quite a group of nice DACs and CDPs through the house and this is the best I've heard. I have a modded Rega CD transport that can take advantage of the feature too, though they now recommend CEC transports which have the clock input feature standard.

    It wouldn't be that hard for manufacturers to include the master-clocking feature Without any change in format. The connectors are the same, the data and clock streams are the same - just separated.

    In my experience it's completely worthwhile to do this but honestly, even the audiophile folks aren't talking about it. My particular SB3 has been completely dedicated to the cause, no analog, can't function without DAC, can't do internet radio. I have two other, more-typical SB3s but it would be great if a single unit could be bought off the shelf that would give either option.
    Last edited by miklorsmith; 2009-11-16 at 12:53.

  3. #13
    Senior Member radish's Avatar
    Join Date
    Apr 2005
    Location
    Red Bank, NJ
    Posts
    5,052
    Quote Originally Posted by miklorsmith View Post
    I have two other, more-typical SB3s but it would be great if a single unit could be bought off the shelf that would give either option.
    You mean a Transporter, right?

  4. #14
    Senior Member
    Join Date
    Sep 2006
    Location
    Danmark
    Posts
    1,361
    Finding a replacement protocol for spdif isn't, and shouldn't, be a complicated technical matter. The only thing needed is a properly asynchronous protocol. There are lots of them, but why not choose something which can be implemented using off-the-shelf components.

    TCP/IP could be one choice, but its probably too complicated and overkill (i.e. does too many things). Asynchronous USB could be another choice, which I believe is already used in some transports/dacs today. Such a choice would also narrow the "compatibility" gap with a lot of present equipement. In terms of needed hardware it would be dead cheap I believe.

    HDMI.... well if becomes a political matter, I'm sure all is set for a catfight.

  5. #15
    Senior Member pfarrell's Avatar
    Join Date
    Apr 2005
    Location
    Wayne, PA
    Posts
    4,251

    SPDIF is evil

    bhaagensen wrote:
    > TCP/IP could be one choice, but its probably too complicated and
    > overkill (i.e. does too many things). Asynchronous USB could be another
    > choice, which I believe is already used in some transports/dacs today.
    > Such a choice would also narrow the "compatibility" gap with a lot of
    > present equipement. In terms of needed hardware it would be dead cheap I
    > believe.


    Why would you suggest an asynchronous protocol? The whole problem, if
    there is any, with jitter is that its not synchronized well enough.

    Just include a clock signal with the existing data signal, and joy and
    happyness.


    > HDMI.... well if becomes a political matter, I'm sure all is set for a
    > catfight.


    Yeah, anytime you get the record industry and the movie industry into
    technology, its very ugly.


    --
    Pat Farrell
    http://www.pfarrell.com/


  6. #16
    Senior Member
    Join Date
    Sep 2006
    Location
    Danmark
    Posts
    1,361
    Quote Originally Posted by pfarrell View Post
    Why would you suggest an asynchronous protocol? The whole problem, if
    there is any, with jitter is that its not synchronized well enough.
    Because carrying the signal asynchronously as far as possible in the chain is a simple solution which by definition avoids any (replay-chain) generated jitter, since there won't be any. It has to get synchronous at some point, but this should be as close to the dac-chip as possible. At least in the "same box" so that the DAC, "source point" and the bus connecting them can be driven off the same crystal. Thats IMO the best solution, since its a solution by design and hardware wise it should be a no-brainer today...

  7. #17
    Senior Member
    Join Date
    Jun 2006
    Posts
    1,690
    Even though I'm generally in the "evil S/PDIF" camp my latest DAC actually uses S/PDIF (coax or optical). But I'm using both an in and an out stream. The DAC generates a stream that is has all zeros for data, this is fed to a soundcard that can sync its output stream to an incoming stream. It doesn't matter how good or bad the cables are etc since the timing on the S/PDIF links is not used for anything. The Local oscillators in the DAC are used to drive the DAC chips.

    By doing this I get to use an asynchronous interface without having to write any drivers, the soundcard manufacturer takes care of that for me.

    The best interface would be for something like the Touch to have clock inputs (NOT a word clock) that drive in instead of its local crystals, nothing other than a switch needed. The data can still go out over the existing digital outs. This automatically syncs it up to the device generating the external clocks and it can still switch between sample rates. The receiving DAC will have to look at its incoming stream and figure out what the sample rate is so it can choose the right clock to use, but that is not particularly difficult.

    As for a new interface I'm actually working on one based on netjack (network protocol for the jack server) which is a very simple protocol using UDP, its simple enough that it can be implemented entirely in a cheap FPGA. It uses standard network infrastructure so you can use regular ethernet cable, switches etc. It doesn't just have to use special hardware, it uses standard low level protocols so a computer can talk this directly out of its existing Ethernet port. The protocol was designed for studio use so it supports many channels and sample rates. Its not just a one to one system, you can have many different devices connected to the same network and route audio data between them.

    John S.

  8. #18
    Member
    Join Date
    Mar 2006
    Location
    Seattle
    Posts
    78

    Maybe

    Quote Originally Posted by radish View Post
    You mean a Transporter, right?
    I believe the Transporter's clock input is wordclock? Frequencies are based on the sampling rate, in the kilohertz range? The Lessloss solution operates in the superclock/megahertz range, which they believe is better. I'm a technical idiot, but I believe the Transporter's system is different.

    JohnSwenson - that sounds very neat!

  9. #19
    Senior Member Rodney_Gold's Avatar
    Join Date
    Dec 2007
    Location
    South Africa
    Posts
    188
    Yes of course , my SPDIF and AES outputs on all my gear sound like doo doo and adopting one of these newfangled not a hope in hell chance of being adopted universally schemes is going to transform my system , hell even my wife would be able to hear the difference...from the next room..
    SBT/Z-sys RDP1 digital Pre/Meridian DSP5500's Or TP/DSP5500's
    SBT/MDAC/Various amps/Osborn Epitomes
    SB3/Meridian DSP5000's
    SBT/ MDAC/Various HP amps/Senheisser Hd800's/650/600/Denon

    "The nicest thing about smacking your head against the wall is...the feeling you get when you stop"

  10. #20
    Member
    Join Date
    Mar 2006
    Location
    Seattle
    Posts
    78
    There's the audiophile spirit - Why try, you probably can't hear the difference anyway!

    Better response - As with everything, it's system dependent. For instance, one of my setups begs tube rolling in my new Peachtree Nova. It's an engaging system and begs to be played with. In the living room system, I couldn't care less because the speakers just don't encourage serious listening. This is the same idea - in my big rig the spendy DAC/transport are divine. In the other setup I probably couldn't hear the difference and almost certainly won't endeavor to find out.
    Last edited by miklorsmith; 2009-11-16 at 15:35.

Posting Permissions

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