Rob Watts on the DAC64

Discussion in 'Hi-Fi and General Audio' started by michaelab, Oct 2, 2003.

  1. michaelab

    michaelab desafinado

    Joined:
    Jun 19, 2003
    Messages:
    6,403
    Likes Received:
    1
    Location:
    Lisbon, Portugal
    Further to this thread I was involved in some heated debate over on the Portuguese hifi forum I'm a member of on the same subject. As it happens, one of the guys there who also runs the website http://www.hificlube.net/ (Portuguese) knows Rob Watts and asked him to reply to some questions about the DAC64 set by forum members. Here are some of the questions and answers.

    The responses on how the buffer works in particular are very interesting... :)

    In relation to transport compatability and jitter:

    Q: The DAC64 is not compatible with many transports. It does not lock to the incoming signal. However with a Toslink connection it works fine with all transports I've used so far. With coaxial after a long pause I have to switch it off and on to get it to lock again. Not so with Toslink, which somehow seems to sound better. I understand you changed the jitter window from 20nS to 45nS in later versions of the DAC64. I have compared both models and the tighter the window the better the sound (provided you have a compatible transport). Why?

    A: The jitter window was improved by improving the design of the receiver FPGA, making it capable of extracting data with very large jitter. The jitter window was more than doubled, and I am now close to tolerating 50% min bit cycle time - the theoretical limit. In performing this update, the DAC and filter FPGA programs are identical - only the extraction circuits. So it is difficult to see how this could significantly change the sound. I have not had a lot of experience with different transports, but I know people have had different experiences. The problem is the RF noise injected into the DAC from the transport, this can effect the sound, making it brighter and more up-front with noisy sources.The solution is (as you state) to use the optical which does not suffer from this effect. In normal situations optical generates so much jitter, that this benefit is masked by the degradation of the jitter. But with the RAM buffer, source jitter is eliminated, so you only have benefits in using optical. Trouble is, most audiophiles have such a prejudice against optical, they won't even try it.


    In relation to the clock and RAM buffer:

    Q: Does DAC64 have its own master clock? What type of clock do you use?

    A: It's a high quality crystal clock with its own buffer and PSU. The DAC 64 works by having a RAM FIFO buffer with two pointers - a write pointer, for writing new data into the RAM and a read pointer for reading the data. Now the two pointers are clocked totally independently, they are completely asynchronous. So how do we get the required delay? This is done by detecting 8 seconds of digital silence, when the pointers are both reset. The read pointer is reset to 1 sec or 4 sec delay, depending on the switch setting. As soon as data appears, the pointers are set on their merry independent paths, and remain in this state until 8 seconds of digital silence. The read pointer is clocked by a local master clock crystal - this provides all the clocks for the filter and DAC. Hence source jitter is removed.

    Q: If so why does it stop playing whenever you pull the transport plug from the mains, although it sometimes keeps playing when you just open the drawer? Does it rely on the transport clock or not?

    A: When you switch off the transport, digital lock is lost, the outputs are immediately muted via a shorting relay. The transport clock is only used to increment the write pointer into the RAM. All circuitry post RAM is clocked by the DAC clock.

    Q: Is the crystal oscillator at the output of the FIFO powered by the same power supply that powers the input of the FIFO?

    A: It uses the PSU for the receiver section. The filter, DAC and analogue all have separate regulated supplies. The input to the filter section is re-synched to remove receiver PSU induced jitter. The PSU is vital, I have always used separate regulators for the clocks. It makes a tremendous difference.

    On digital cables:

    Q: What is the preferred connection of the DAC64, AES/EBU, S/PDIF or Toslink?

    A: Optical. The DAC is still sensitive to the amount of RF noise the transport generates. But when using optical, it is very consistent. Moreover, analogue electronics are also sensitive to the noise the transport generates...


    On the different buffer lenghts:

    Q: What is the difference between the 1s and 4s buffer settings and should there be any audible difference? If not, why the 2 settings?

    A: There is no difference between 1 sec and 4 sec delay once data is running - its exactly the same circuitry doing exactly the same. Using an optical connection, I can't hear a difference between the two settings. The reason I give 1 and 4 seconds is the only time you get that is guaranteed 8 seconds of silence is when you are changing discs. So you can have 72 minutes of playing a CD where the pointers are completely separate. If the clock on the transport is low quality and not accurate, the 1 second delay can drift to so that there is no delay - then suddenly 8 seconds! So we give the option for poor accuracy transport clocks. The vast majority of hi-end transports are accurate enough for 1 second.


    My comments:
    I find what he said about how the clock/buffer works particularly interesting. The fact that the replay (buffer read) clock is totally independent of the transport does in fact mean that the DAC64 IS completely immune to jitter coming from the transport. It's impossible for it not to be. However, there are differences between transports caused largely by RF interference. On Saturday I'll have my newly modded Teac T1 (thanks Tone ;) ) hooked up and will play around with optical and other connections to see if this bears out in reality...

    Another interesting point that comes from this particular method of using the RAM buffer is that if you were using the DAC64 with a DAB radio and left it on all the time there would come a point where you'd get a buffer underrun or overflow and the DAC would have to be reset (by switching it off and on I suspect).

    So, interesting stuff and my thanks to José Henriques from http://www.hificlube.net/ for doing the "interview".

    Michael.
     
    michaelab, Oct 2, 2003
    #1
  2. michaelab

    rewster

    Joined:
    Jun 22, 2003
    Messages:
    96
    Likes Received:
    0
    Location:
    The near east
    ;) Nice work michealab,

    Rob Watts eh? He was becoming a bit of a mythical beast at one point, something of a hi fidelity unicorn.

    Anyone who has owned kit with his handiwork built in must hold him still in high regard.

    I still miss my little bit 3!
     
    rewster, Oct 2, 2003
    #2
  3. michaelab

    Robbo

    Joined:
    Jun 19, 2003
    Messages:
    2,371
    Likes Received:
    0
    Location:
    Berkshire, UK
    Thanks for posting this Micheal. very interesting. Rob always used to say when he was at dpa that optical connections are better as the rf is reduced. however I have found in my system that the coaxial connection is definately better.

    My dpa dac is still doing sterling service and sounding better than ever. one day I would like to a back to back with a DAC64 in my system to see which I prefer.

    Cheers, Robbo
     
    Robbo, Oct 2, 2003
    #3
  4. michaelab

    Paul Ranson

    Joined:
    Sep 4, 2003
    Messages:
    1,602
    Likes Received:
    0
    Location:
    An octopus's garden.
    Interesting stuff. Coincidentally, earlier I was reading a 1991 review of a DPA amp. There are common threads, I think Rob knows what he's talking about.

    The one thing that isn't explained is why an integrated player isn't intrinsically better, most of the effort in the DAC seems to be dealing with the connection between transport and DAC.

    It's a shame DPA didn't last longer, they had striking design and an individual electronic approach.

    Paul
     
    Paul Ranson, Oct 3, 2003
    #4
  5. michaelab

    The Devil IHTFP

    Joined:
    Aug 13, 2003
    Messages:
    4,613
    Likes Received:
    0
    Location:
    Disco Towers
    It's only a chuffin' DAC. Get over it.
     
    The Devil, Oct 3, 2003
    #5
  6. michaelab

    Paul Ranson

    Joined:
    Sep 4, 2003
    Messages:
    1,602
    Likes Received:
    0
    Location:
    An octopus's garden.
    I don't think it chuffs, Bub.....

    Paul
     
    Paul Ranson, Oct 3, 2003
    #6
  7. michaelab

    Robbo

    Joined:
    Jun 19, 2003
    Messages:
    2,371
    Likes Received:
    0
    Location:
    Berkshire, UK
    I think you may have hit the nail on the head here Paul. If dpa had concentrated their efforts on making a top notch one box player instead of going against the grain and persisting with transport/dac conbinations, they may well have been around today.
     
    Robbo, Oct 3, 2003
    #7
  8. michaelab

    wadia-miester Mighty Rearranger

    Joined:
    Jun 19, 2003
    Messages:
    6,026
    Likes Received:
    0
    Location:
    Beyond the 4th Dimension
    High praise indeed from Mr Pitbull, thing is Mr Watts does actually does the practice as well as theroy, resulting in some good products, interestingly Chord (unsubstanicated rumour here) only engauged his expertiese for 2 or 3 days design consaultancy on the Dac 64, at it's inception, and he felt things didn't quite go in the direct he first thought they might. Still you can always pay £4200 for a German Stein mug lookalike, that'll do a good job of decapitating your discs, should you be less than careful.
    But I'm sure the lifestyle mid management bmw owning guru's will own one to show their friends on a Fung sui night, Fondu anyone ? :) Cynical humm maybe, but John Franks answer was cutting edge when asked about the price of the 'Blue' transport :D WM
     
    Last edited by a moderator: Oct 3, 2003
    wadia-miester, Oct 3, 2003
    #8
  9. michaelab

    lowrider Live music is surround

    Joined:
    Jun 19, 2003
    Messages:
    1,309
    Likes Received:
    0
    I think he says that RF is his DAC's worst enemy, as it takes care of jitter better than usual...

    With other DACs, jitter causes more problems than RF, so we are better off with coax cables... :rolleyes:
     
    lowrider, Oct 3, 2003
    #9
  10. michaelab

    wadia-miester Mighty Rearranger

    Joined:
    Jun 19, 2003
    Messages:
    6,026
    Likes Received:
    0
    Location:
    Beyond the 4th Dimension
    Antonio, I would agree, but I feel Mr Watts has suggested this for more than a good reason, we have improved the Dac 64, also, but by other methods, the SWPS (great generator of RF) close proximity to associated electrionics, inadequate shielding, average anaolgue stages all add up, but It will be interesting to read Mike's view of this method :) we made yet be all surprised, open minds are the key Here WM
     
    Last edited by a moderator: Oct 3, 2003
    wadia-miester, Oct 3, 2003
    #10
  11. michaelab

    merlin

    Joined:
    Jun 23, 2003
    Messages:
    3,262
    Likes Received:
    0
    Most of today's better dacs offer buffering and PLL circuitry to be fair Antonio. Jitter is less of an issue than it was unless using a very poor transport.

    RF from the transport is still usually seriously high from stock. It's the old story of compromise. TosLink is sadly a compromise too, but keep the lengths down to less than a meter and it can be pretty good using a low jitter source.
     
    merlin, Oct 3, 2003
    #11
  12. michaelab

    michaelab desafinado

    Joined:
    Jun 19, 2003
    Messages:
    6,403
    Likes Received:
    1
    Location:
    Lisbon, Portugal
    Yes, but as far as I'm aware offer total immunity from transport induced jitter like the DAC64. It's the unique way the buffer is used in tha DAC64 that makes all the difference.

    So, having got rid of the issue that makes TOSlink a compromise (jitter) it's now the perfect connection :)

    Michael
     
    michaelab, Oct 3, 2003
    #12
  13. michaelab

    merlin

    Joined:
    Jun 23, 2003
    Messages:
    3,262
    Likes Received:
    0
    So in theory, the £1.99 optical lead (5m) that I got for me X Box, should sound better than WM's Liberator:confused:

    The Tact gear is linked together using a Sync cable and PLL circuitry (although nothing as extreme as Chord) and some people have recommended using the Optical connection for the very reason that you describe, namely reducing RF transmission. For me, it didn't work, but I will be interested in your experiments Michael (as I'm sure will WM;) )
     
    merlin, Oct 3, 2003
    #13
  14. michaelab

    wadia-miester Mighty Rearranger

    Joined:
    Jun 19, 2003
    Messages:
    6,026
    Likes Received:
    0
    Location:
    Beyond the 4th Dimension
    Mike, I this instance, I'm sort of in agreement with Mr Watts on this, however only the Dac 64 most ardent follower will tell us, plus doesn't the toslink have limited bandwidth???, 48KHz, as the 64 does have upsampling software/Hardware, one prosumes that 44.1khz is getting a serious polishing :) Mike will fill us in detailed on sunday I feel. Wm
     
    wadia-miester, Oct 3, 2003
    #14
  15. michaelab

    GrahamN

    Joined:
    Jun 20, 2003
    Messages:
    572
    Likes Received:
    0
    Location:
    Suwway
    So the purchase of the Teac, and the mods done to it, were a complete waste of time and money? ;) Would be interesting to compare that with say a Walkman as transport.

    Interesting that about the free running read-pointer and minimal difference between 1 and 4 second buffer. So what about all those reviews talking about hearing more depth, body etc. with the 4sec over the 1sec?

    1 sec in 80 mins = 0.02% = 200ppm
    4 sec in 80 mins = 0.08% = 800ppm

    A typical crystal oscillator (e.g. costing £5) would have something like a 50ppm temperature stability spec - so sounds OK. Not sure what goes into an el cheapo Walkman etc. though.

    Anyone actually tried the DAC64 on one of those Freeview/Pace boxes with the digital outputs?

    As for why he didn't go down the integrated route, he sort of implied the answer - RF transmission is probably more of a problem within a single enclosed box. The designer also then has more control over internal shielding etc. so more swings and roundabouts.

    Toslink bandwidth is completely irrelevant - unless trying to put in upsamplers etc - it does 44.1 and 48kHz, and that's quite enough for CD and DAT.
     
    GrahamN, Oct 3, 2003
    #15
  16. michaelab

    chris.gally

    Joined:
    Jul 24, 2003
    Messages:
    161
    Likes Received:
    0
    Location:
    Manchester, England
    I have a bit of experience of what your talking about here.

    I used the Tag transport, which was connected to the DAC64 with a Chord Anthem AES/EBU cable.

    Now i've sold that and hoping to replace it with a Big Old Sony SACD player. :D

    In the meantime i've gone back to using the Xbox as a transport with a noname, yet fairly substantial Toslink cable. Guess what...
    The Xbox with Toslink in my opinion sounds better than the Tag. :eek:

    The soundstage increased, better dynamics all the stuff you expect from a box UPGRADE

    Could this be anything to to do with the PC type archetecture of the XBox or the low Error reading from the DVD Drive, or the use of the Toslink, i dont know. :confused:

    The only problem is the noise that comes fron the xbox (fans, hard-drive).

    I only connected it up on Wednesday and was quite supprised when i read this thread.

    If anyone fancies a comparison PM me to arrange.

    Cheers

    Chris
     
    chris.gally, Oct 3, 2003
    #16
  17. michaelab

    wadia-miester Mighty Rearranger

    Joined:
    Jun 19, 2003
    Messages:
    6,026
    Likes Received:
    0
    Location:
    Beyond the 4th Dimension
    My point Graham was the 64 has the ability to accept 96 (and now the 192Kz) so why put it in on there ? it can accecpt the high sample rates why not use them especialy 88.2, virtualy Ideal sir, you be surprised how good you wadia sounds with 88.2Khz Graham :cool:
     
    wadia-miester, Oct 3, 2003
    #17
  18. michaelab

    Robbo

    Joined:
    Jun 19, 2003
    Messages:
    2,371
    Likes Received:
    0
    Location:
    Berkshire, UK
    My point was more from a marketing perspective. Why persist with 2 box solutions which were seriously out of vogue at that time? After all a business is all about maximising sales. only providing 2 box solutions kind of limited them a bit!
     
    Robbo, Oct 3, 2003
    #18
  19. michaelab

    michaelab desafinado

    Joined:
    Jun 19, 2003
    Messages:
    6,403
    Likes Received:
    1
    Location:
    Lisbon, Portugal
    Hmm - quite possibly was a waste of money :cry: However, I'd have no street cred with a little Sony Discman "transport" sitting on the top shelf of my rack would I? :D

    Seriously though, I'm going to give it a try with my Discman.

    Got the T1 up and running this morning. Was a bit tempramental at first and wouldn't recognize any discs but a quick call to the Tone support line ;) soon sorted it. Seems like these VRDS transports sometimes need the occasional "reboot", esp. after manhandling by DHL :)

    Got it running into the DAC64 on TOSlink at the moment. Holst's "Planets" (Montreal Symphony, Dutoit, Decca) in there to warm things up ;) When the wife has gone down the gym later I'll have time to do some proper tests at, ahem, GrahamN volume levels :JOEL:

    Still - sounding pretty impressive so far I have to say.

    Michael.
     
    michaelab, Oct 4, 2003
    #19
  20. michaelab

    michaelab desafinado

    Joined:
    Jun 19, 2003
    Messages:
    6,403
    Likes Received:
    1
    Location:
    Lisbon, Portugal
    Robbo - one thing seems to be clear from talking to the two people I know here (in Lisbon) who know Rob Watts personally: he doesn't have much business sense. He always did what he thought was technically superior regardless of other concerns. It's a shame that his "brilliance" has only now been generally recognized whilst under the wing of Chord and not on any of his own ventures (DPA, Deltec).

    Michael.
     
    michaelab, Oct 4, 2003
    #20
Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Similar Threads
Loading...