Naim DAC - tech. q.
Posted by: bhaagensen on 21 January 2010
Hi,
this is perhaps to the hairy side and may not even be particularly relevant (or allowed)? Nonetheless, its keeping me up at nights, so I'll try my luck
Having read the whitepaper I am very impressed by the nice solution to the jitter-issue. Its IMO the (only) right solution. But now that it seems so close to perfect, I can't help but think about two things that I would think could completely avoid the use of ASRC.
- Use larger buffer. Memory is cheap these days. Why not let the DAC buffer, say 100MB. That should give a lot of room for one of the 10 clocks to match sufficiently close.
- I can see that the above does not e.g. scale well. But then one could instead/also monitor the buffer level. If it falls below some threshold, resync on a sample and choose a slightly slower clock. If it's saturating, resync on a sample and choose a slightly faster clock.
Anyway, there it is. As I said its most of a technical curiosity on my side - answers highly appreciated though.
Bjørn
this is perhaps to the hairy side and may not even be particularly relevant (or allowed)? Nonetheless, its keeping me up at nights, so I'll try my luck
Having read the whitepaper I am very impressed by the nice solution to the jitter-issue. Its IMO the (only) right solution. But now that it seems so close to perfect, I can't help but think about two things that I would think could completely avoid the use of ASRC.
- Use larger buffer. Memory is cheap these days. Why not let the DAC buffer, say 100MB. That should give a lot of room for one of the 10 clocks to match sufficiently close.
- I can see that the above does not e.g. scale well. But then one could instead/also monitor the buffer level. If it falls below some threshold, resync on a sample and choose a slightly slower clock. If it's saturating, resync on a sample and choose a slightly faster clock.
Anyway, there it is. As I said its most of a technical curiosity on my side - answers highly appreciated though.
Bjørn
Thanks Bjorn, please leave your contact number here, and they'll call you when they'll do DAC2! 
AFAIK, the ASRC act as a digital filter. Since the Naim DAC is able to read files from many types of medias, it's seems important to have a strong entry point that will insure an ultra stable and time-aligned flow of data for the DAC. It's not like if you have a single and ponctual source of data (i.e. a Red Book CD player), where you could probably avoid the use of the ASRC... and simply fit the DAC inside the CD transport.
Well, that's my humble view about it.
Sleep well!
Bye.
AFAIK, the ASRC act as a digital filter. Since the Naim DAC is able to read files from many types of medias, it's seems important to have a strong entry point that will insure an ultra stable and time-aligned flow of data for the DAC. It's not like if you have a single and ponctual source of data (i.e. a Red Book CD player), where you could probably avoid the use of the ASRC... and simply fit the DAC inside the CD transport.
Well, that's my humble view about it.
Sleep well!
Bye.
@ bhaagensen: Take a look at the new PS Audio DAC - it uses a large buffer. The Naim DAC seems to use a smaller buffer, but this is good since this way the latency is very short and this enables you to use it for Blu Ray, pp. And from telling how good and precise the Naim DAC sounds in my living room I would say that there is no jitter problem at all
quote:Originally posted by John R.:
@ bhaagensen: Take a look at the new PS Audio DAC - it uses a large buffer. The Naim DAC seems to use a smaller buffer, but this is good since this way the latency is very short and this enables you to use it for Blu Ray, pp. And from telling how good and precise the Naim DAC sounds in my living room I would say that there is no jitter problem at all
John ... Isn't the buffer on the PS Audio system located in the transport NOT the DAC. Granted they are connected with an i2s conenction not SPDIF which may make a difference...
The DAC I know of with similar (but larger) buffer is the Chord QBD76 (and previous DAC64).
Eloise
Hi John R, and thanks for your reply.
I am not in the marked for a DAC, but if I was the Naim DAC would be a strong candidate. My question is more of a technical curiouslity on my side.
I don't see why latency need be dependent on the buffer size though. It could be, but ought still to be possible to overcome?
quote:Originally posted by John R.:
@ bhaagensen: Take a look at the new PS Audio DAC - it uses a large buffer. The Naim DAC seems to use a smaller buffer, but this is good since this way the latency is very short and this enables you to use it for Blu Ray, pp. And from telling how good and precise the Naim DAC sounds in my living room I would say that there is no jitter problem at all
I am not in the marked for a DAC, but if I was the Naim DAC would be a strong candidate. My question is more of a technical curiouslity on my side.
I don't see why latency need be dependent on the buffer size though. It could be, but ought still to be possible to overcome?