Delivery of bits to player

Far more important than the delivery method is the shelf you store them on, as this is closer to the transport. Cryogenically treated zebrano shelves are the best ime, ideally with a prime number of cds per shelf in order to avoid inter-cd resonance build up. And of course, only one shelf per wall bracket, and only take a cd off at night when subterranean rumble effects are at a minimum ... and if you ever drop a cd on the floor, you might as well throw it away as all the zeros will be dented and the ones will be wonky ..

quote:

Originally posted by likesmusic:
and if you ever drop a cd on the floor, you might as well throw it away as all the zeros will be dented and the ones will be wonky ..

I just threw away 314 cd's as i'm afraid all have been dropped at least once and wouldn't know if it hadn't. Starting over.

And if they arrive fed ex overnight, they become equal to hi-res files.

Feng Shui?

Tog

Crystal oscillator healing?

Tog

ROTF, when bits are delivered into the nDAC buffer then the history is wiped out and since that moment you are getting deterministic output onwards. This subject has been discussed here on the forum for almost two years. I used to advocate that thinking that all transports do impeccable job in delivering bits inside Naim DAC is utopia and lack of knowledge in physics of data transfer. There are at least two physical phenomena which influence the Transport-DAC performance.

The first one is jitter which can be so big that DAC's S/PDIF receiver can mess up in resolving the next incoming bit despite the heroic efforts to assess the distorted bitstream waveform.
Many people believe that every bistream is presented by a perfect square-shape waveform. It will be very useful for them to run a simple bitstream from dvd player dig out to oscilloscope and see howe does it look like in reality. In high quality transports the output bitstream waveform is not seriously distorted and re-clocking DAC will resolve all bits perfectly and Transport-DAC pathway will be the same perfect as network streamers. Obviously, Hiface and cheapish DVDs are not as good.

The second point is the power contamination which may come (or may not) from the transport: if you plug a shitty dvd player in the same power block or the same spur as your amps you will contaminate the sound even though dvd player may transmit a perfect bistream waveform to the DAC.

quote:

ROTF, when bits are delivered into the nDAC buffer then the history is wiped out and since that moment you are getting deterministic output onwards.

as SPDIF does not offer a flow control the sender and receiver are not completely decoupled and the ndac must synchronize to the sender. It does this by changing the clock out of the buffer. So the dac may immune against SPDIF related jitter but not completely against the senders clock jitter. This may also audible.

quote:

Originally posted by jerryct:

quote:

ROTF, when bits are delivered into the nDAC buffer then the history is wiped out and since that moment you are getting deterministic output onwards.

as SPDIF does not offer a flow control the sender and receiver are not completely decoupled and the ndac must synchronize to the sender. It does this by changing the clock out of the buffer. So the dac may immune against SPDIF related jitter but not completely against the senders clock jitter. This may also audible.

Hi Jerryct -

Are you sure about this comment?

My understanding is that the "sender's clock jitter" you refer to is the same thing as S/PDIF jitter.

Are you perhaps thinking of the jitter that was injected at the time the music was recorded (and on playback becomes indistinguishable from the music itself)?

As AMA said, this topic has been discussed to death, and I can recall several lengthy threads where one particular audio engineer named AndyS was very clear on this specific point.

Hook

jerryct - it is true that the nDAC and the NDX change the clock out of the buffer, but according to the NDX white paper the clock frequency changes (a very small amount)typically once every ten or fifteen minutes. Are you saying that you can hear this incrdibly small micro-event? What does it sound like? As long as the buffer doesn't under- or over-flow the dac is surely immune to what you call the senders clock jitter. What on earth is the difference between 'senders clock jitter' and s/pdif jitter?

You don't keep your cds on really expensive shelves by any chance do you? Or have them delivered by courier?

with spdif jitter i mean the jitter which is introduced using a spdif interface. it is commonly known that spdif itself (only by using such a interface) introduces jitter. that is why some people say that this audio interface is "flawed".

if you look at the block diagram of a "typical CD player" in the naim dac white paper. There is a master clock and this clock also has jitter. This clock is encoded into the spdif stream as the word clock but the jitter is not caused by the spdif interface.

you may found the Stereophile review of the Genesis digital lens interesting.

At least i find it interesting because it reminds me very much of the dac. This product has 2 disadvantages. It is a reclocker and converts spdif to spdif. And as the reviewer states only by outputting the audio data to spdif again it introduces new jitter.

The second disadvantage is that it is build around a cd. you know the play time of a cd in advance and you can calculate a buffer which will never over- or underrun. thus the sender and receiver clock are totally asynchronous.

Today you have endless playlists and with the concept above you get into trouble because the buffer cannot be infinite. With a limit buffer and no flow control you must (!) synchronize the frequencies. But with this the sender and receiver is not decoupled anymore. Back to the master clock from the beginning - you may have eliminate the spdif introduced jitter but not the jitter of the master clock. in an ideal world the master clock is exactly for example 44.1 kHz with no jitter and the dac would only use one fixed frequency without changing it. in an non ideal world this clock jitters and that is why you the ndacs "switchable fixed frequencies" "are selected so as to keep the average clock frequency the same as the source’s". This is caused by jitter (or a misaligned clock).

To your question if it is audible: naim says so. (you have to keep in mind not only the micro-event of the switching itself but also the slightly higher or lower pitched clock)

quote:

You don't keep your cds on really expensive shelves by any chance do you? Or have them delivered by courier?

No

jerryct - what Stereophile say of other players is not relevant to the Naim DAC.

Naim say:

"The data entering the downstream digital filtering and DAC chips is then completely isolated from the incoming S/PDIF jitter."

So how does what you are calling "senders clock jitter" get into the nDAC? The only connection is an s/pdif cable? And Naim say the DAC is "completely isolated" from that?

All that's needed to keep the buffer loaded is a very occasional - ie every 10 or 15 minutes - adjustment to the clock frequency at the output end of the buffer. This is hardly jitter! The whole design of the nDAC is intended to isolate it from any jitter in the clock in the sending device - are you saying they have failed?

You have a clock in the sender and every clock has jitter.
SPDIF merges this clock (word clock), bit clock and audio data together. SPDIF itself adds his own jitter.

So you have:
sender clock jitter + spdif induced jitter = jitter entering the dac.

Naim always talks about eliminating SPDF-induced jitter but not jitter in general.

So you get:
jitter entering the dac - spdif induced jitter = sender clock jitter
(this jitter is transmitted over the spdif interface but it is not induced by the spdif interface)

The nDAC needs to cope with this sender clock jitter and/or drift.
And this is the reason why naim needs to "keep the average clock frequency the same as the source’s"
This is cause and effect.

So as i said if the sender clock is ideal than you need no frequency switching - like usb stick playback.

Thats all.
And back to the OPs original question i only want to mention that this phenomena needs to be considered too!

So i do not think they failed because i never read a statement from naim that they are "isolating the nDAC from any jitter in the clock in the sending device"?

But regarding your second question wether this phenomena is audible and how does it sound there are 2 aspects:

1.) Is this generally audible?
Yes!

2.) In this special case: which maximum level of frequency switching per second, steps of frequency changes … is audible.
I can not tell you and i believe no one really can because this is a endless discussion which level of jitter is audible.

quote:

Originally posted by jerryct:
.. i never read a statement from naim that they are "isolating the nDAC from any jitter in the clock in the sending device"?

The statement you haven't read is here on the Naim DAC page:

"perhaps most importantly, it delivers jitter-free processing through its S/PDIF inputs."

Jitter-free. The clock isn't recovered from the sender, so can't be influenced by jitter from any source in the sender.

quote:

2.) In this special case: which maximum level of frequency switching per second, steps of frequency changes … is audible

Again, according to the NDX white paper which goes into a bit more detail about how the buffering works than the DAC white paper, the frequency switching that the DAC does happens every 10 or 15 minutes, not per second.

quote:

The clock isn't recovered from the sender, so can't be influenced by jitter from any source in the sender.

If there is a clock in the sender you #must# synchronize to it. This means you must recover the clock in some form. And with switching the frequencies the ndac is synchronizing. There is no magic behind it. They only implement a very elegant mechanism to minimize such effects.

This is why so many prefer other interfaces like ethernet, async usb, firewire or usb stick playback for a completely free running clock in the dac.

quote:

Originally posted by jerryct:

If there is a clock in the sender you #must# synchronize to it. This means you must recover the clock in some form. And with switching the frequencies the ndac is synchronizing. There is no magic behind it. They only implement a very elegant mechanism to minimize such effects.

This is why so many prefer other interfaces like ethernet, async usb, firewire or usb stick playback for a completely free running clock in the dac.

This is why I have suggested elsewhere why not then utilize the USB playback method and extend it by buffering a memory with data that is played back. The buffering is done from a NAS drive or even UPnP source. But it means developing a whole different interface. If this will eliminate completely all the different types of jitter then it will meet the objective.

ASenna04

quote:

Originally posted by jerryct:

If there is a clock in the sender you #must# synchronize to it. This means you must recover the clock in some form. And with switching the frequencies the ndac is synchronizing. There is no magic behind it. They only implement a very elegant mechanism to minimize such effects.

Sure the Naim DAC has to know which of its own clock frequencies to pick, but this (re-) synchronisation is only happening at a gross level, every 10 or 15 minutes according to Naim. The clock is not derived from the clock in the source, so jitter in the source cannot affect it as you claimed. ASRC DACs achieve a similar result a different way, as the White Paper points out.

i cannot describe it better than i did.

But what do you think is the intention of m2tech, wavelink, halide usb-spdif converters?
Aren´t they pulling the audio data out of the computer with an asynchronous method to get rid of the computers clock and generate a spdif signal with its own free running clock? Isn´t this the whole idea behind these products? I know that these products are designed with PLL-DACs in mind that can now better recover the clock but why do some people prefer them with the nDAC?

I am not claiming that this is directly audible and i cannot tell which level is needed that this is not audible anymore but this is one out of other variables that should be considered and is different between different transports.

For example AMA claims different transports sound different because of random bit flips? How often should they occur? Is always the most significant bit affected? if they occur every minute do you think this is audible if one sample is affected?

quote:

Originally posted by jerryct:

1.) Is this generally audible?
Yes!

quote:

Originally posted by jerryct:

I am not claiming that this is directly audible

!

you have teared apart my answer. it is not meant as a contradiction because as i said there are 2 aspects: 1.) in general (in a global sense) it is audible 2.) in this special case i cannot tell because i do not know which level of jitter is audible

quote:

Originally posted by jerryct:

1.) Is this generally audible?
Yes!

2.) In this special case...
I can not tell you ...

Doesn't matter - unless it is spoiling your enjoyment of the music. It is very easy to get obsessed with jitter - there are some threads on Computer Audiophile that have lasted for decades - well feel like they have anyway.

Tog

Warning - Grumpy comment follows

Frankly I don't care about theory, all I know is my system sounds clearly better from my Weiss int202 than it does from my $1000 NAD DVD (which I am using right now because the Weiss developed a fault and is being fixed).

Some might blither on about it being somehow impossible that this could be the case. No more diplomacy. You are wrong. The wow factor is clearly missing, detail is clearly being lost. There is a level of detachment from vocals and instruments. Not so much a veil, more like a dumbing down and a more "digital" sound. Its simply not as good.

The sooner I get my interface and computer source back the happier I will be.

End of story.

No arguments here - there are quite clearly differences in sound quality between different bits of kit or connections - the reason why?

Could be jitter - most probably something else.

Tog

I agree with you that there are far more aspects to consider than this jitter thing and one should not buy a dac solely because of his interface.

Tog, you raised the question "Which sounds better streaming or dig out?". There were also arguments like "Streaming has no inherent jitter", differences between pulling and pushing data, "network is not part of the signal path–all sample clock timing occurs within the player itself".

This is exactly the same as i described above - maybe in more technical way. Thus i do not understand why there are now so many doubts?!? I only repeat what others said before and seems to be a common understanding in this forum.

(Or it does not matter than it is a superfluous discussion if streaming gives better sound quality.)

quote:

Originally posted by jerryct:
I agree with you that there are far more aspects to consider than this jitter thing and one should not buy a dac solely because of his interface.

Tog, you raised the question "Which sounds better streaming or dig out?". There were also arguments like "Streaming has no inherent jitter", differences between pulling and pushing data, "network is not part of the signal path–all sample clock timing occurs within the player itself".

This is exactly the same as i described above - maybe in more technical way. Thus i do not understand why there are now so many doubts?!? I only repeat what others said before and seems to be a common understanding in this forum.

(Or it does not matter than it is a superfluous discussion if streaming gives better sound quality.)

In the real world - if you like the sound who cares?

I do think there is a growing polarization in the hifi industry between those who believe streaming is the best route forward and those that prefer dig out solutions where they can tweak individual components.

Linn and Naim seem to be moving towards a streaming future. Linn certainly feel that integrating the dac and streamer leads to a better sound.

Interestingly the next B&W Zeppelin will have built in Airplay and a new dac. If I was a betting man ... I would say the future is streamed.

Tog

we do not disagree that streaming is the future but that was not the initial question of this thread.

but who cares