Jump to content

Naim R2R Multibit Network player, keeping up the iconic CD555 legend.


Recommended Posts



On 07/12/2018 at 4:54 PM, georgehifi said:

The PCM1704K ("K" stood for special selected) is "said" to be the last and best R2R made for audio, my Linn CD12 MkIII uses 4 of these K's in it. There are old R2R chips still being made like the Schiit Yaggy uses (Analog devices AD5791) but it wasn't designed for audio but said to be for a US missile guidance system or as it says on the data sheet high end aerospace instrumentation.

 https://www.analog.com/media/en/technical-documentation/data-sheets/ad5791.pdf

 

Now it up to the discrete R2R designers Holo, Denafrips, Sokris, Total,  ect, ect  to get up to and maybe beat the PCM1704K, and if the Holo was any indication that I used, there already up there.

 

Cheers George 

The DAC I uses 8 chip pcm1704uk (audio GD nos 7) good to know the newer discreet are up there in performance. Cos I tried yggy and personally is not as good as the audiogd nos7

  • Like 1
Link to comment
Share on other sites

On 06/12/2018 at 5:47 PM, georgehifi said:

....go talk to the Deqx designers Allan Langford and Kim Ryrie round the corner from me. I'd believe them before you any day of the week.

So I did this, and Kim from DEQX just sent me an email, explaining that what I've been saying in this thread is correct.

 

As I explained earlier.    This was never about being a choice between DEQX or I being right  (ie. "I'd believe them before you any day of the week").... and as I explained,  I knew both DEQX and I were correct (cos this is basic stuff, and I know they know it backwards) ... so as I explained my assumption was that George simply misunderstood or misinterpreted their advice.

 

As I explained.... The DEQX guys may have indeed given advice which boiled down to  "you will always have a delay problem" ..... however this is because they were specifically about their own product .... and so only considering the FIR filter case  (DEQX primarily uses FIR filters)  .... even though for the majority of DSP users the IIR case is more common.

 

Quote

If you connected outputs of a preamp to two speakers (eg. main speaker and sub, was the example used before)  .... and one of those outputs goes via a DSP.

  • If you use an IIR filter (which is more common) in your DSP, you will have no delay between each speaker.
  • If you use a FIR filter (which is what DEQX use, but isn't as common) in your DSP, then there will be a large delay between each speaker.

 

 

Edited by davewantsmoore
grammar/clarity
  • Like 1
Link to comment
Share on other sites

Happy to post up excerpts from said email.....  Although rather than "believing someone" (like me, or Kim), I'd recommend anyone who is that interested to just have a quick google (and you will find endless other sources) as this is basic/public knowledge.   We're not breaking any new ground here.

Link to comment
Share on other sites

  • Volunteer
20 hours ago, davewantsmoore said:

So I did this, and Kim from DEQX just sent me an email, explaining that what I've been saying in this thread is correct.

 

As I explained earlier.    This was never about being a choice between DEQX or I being right  (ie. "I'd believe them before you any day of the week").... and as I explained,  I knew both DEQX and I were correct (cos this is basic stuff, and I know they know it backwards) ... so as I explained my assumption was that George simply misunderstood or misinterpreted their advice.

 

As I explained.... The DEQX guys may have indeed given advice which boiled down to  "you will always have a delay problem" ..... however this is because they were specifically about their own product .... and so only considering the FIR filter case  (DEQX primarily uses FIR filters)  .... even though for the majority of DSP users the IIR case is more common.

 

 

 

Interesting. I’m learning stuff 

thanks Dave ?

  • Thanks 1
Link to comment
Share on other sites



13 hours ago, frankn said:

Could you post in full the question you posed to DEQX and their reply?  

Sure.    I've removed part of the Kim's answer (and my subsequent reply) where is gets "DEQX specific"   (the patents I mentioned previously on FIR filters which have lower than normal delay, are discussed)

 

Quote

DAVE

I currently run two outputs from a preamplifier .... one straight into my main speakers (which have an internal passive high-pass filter) ... and the second preamplifier output through a DSP and then to my woofers. I use the DSP to do a low pass filter, and fix some bumps in the woofer.   The little DSP I use (you probably know the one) offers me a choice between IIR filters and FIR filters.

  • When I use FIR more for the filters - The woofers are delayed vs the main speakers - and it sounds bad.
  • When I use IIR mode for the filters - It works as expected (there is no delay problem)

Can you clarify why this is? I've been told it is just how those filters work.... and that I should expect it to work just like that - ie. no delay issue with IIR ... but to expect the delay issue with FIR filters.    Can you confirm if my understanding is correct?

 

 

 

Quote

KIM

Yes you are correct that normally FIR filters add in the order of 100 – 200 milliseconds for high resolution filtering.   <SNIP>

 

Quote

 

DAVE

So, in summary....

  • FIR.   Long delay.    Phase correction.
  • IIR.    Zero delay.    No phase correction. 

If the above is true, it explains why in my little DSP (that I might replace with a DEQX) that when I use a FIR filter on my woofers (but run my mains without DSP) that I get an huge delay between the woofer and mains ...... but when I do the same with only an IIR filter on woofer, there is no delay problem.

 

 

Quote

KIM

IIR filters would not have 'exactly zero’ delay, but yes their delay is similar in effect to traditional passive speaker crossovers. So any phase-response error/delay is usually pretty negligible in ‘millisecond’ terms at least. That is why you notice the long delay using traditional DSP/FIR vs IIR (DSP or analogue). We definitely recommend using DEQX calibration for main speakers if that’s an option. 

 

Quote

DAVE

Yes!  :)   Thanks Kim, that's what I mean.    An “extremely short” delay, just like in a typical passive XO in a speaker.

 

Note that this "extremely short delay" .... is in many cases (ie. when you have  matching high pass and low pass response .... which is what you would have in the example that we discussed in this thread)  the short delay that you get.... is exactly the delay that you want.   (ie. just like what happens in passive XO filter)

 

Link to comment
Share on other sites

13 hours ago, frankn said:

Could you post in full the question you posed to DEQX and their reply?  

As I said way back I found with my MiniDSP 2 x 4 HD, with no fix for the latency, short of pulling the woofer out of the box and re-positioning it well out into the room. 

 

"If he's doing the bass room correction, FIR and bass digital xover management with it, and not going through it for the mids and highs staying passive direct, there is a latency problem, even Allan and Kim from Deqx will tell you the same."

 

"KIM:  Yes you are correct that normally FIR filters add in the order of 100 – 200 milliseconds for high resolution filtering."

 

Sound travels at 1130 feet per second. Thus it takes about 9/10 of a millisecond for sound to travel a one foot distance. Recording engineers conveniently round this off to the timing formula of 1 ms = 1 foot.

 

Cheers George

 

  • Haha 1
Link to comment
Share on other sites

36 minutes ago, georgehifi said:

As I said way back I found with my MiniDSP 2 x 4 HD, with no fix for the latency, short of pulling the woofer out of the box and re-positioning it well out into the room. 

 

"If he's doing the bass room correction, FIR and bass digital xover management with it, and not going through it for the mids and highs staying passive direct, there is a latency problem

Yes, of course there is .... BECAUSE YOU ARE USING A FIR FILTER.

 

If you use an IIR filter (which is more common), there will be no delay.

 

This is what I've been saying all along.

 

36 minutes ago, georgehifi said:

 even Allan and Kim from Deqx will tell you the same."

 

Yes. They did... which is the same thing that *I* said.

 

I said....     "there's not ALWAYS delay problem .... only a delay problem when you use certain types of filters".

 

F-me.   This is just mean and stupid at it's finest.   Bye.

 

 

  • Like 1
Link to comment
Share on other sites

  • Volunteer
31 minutes ago, davewantsmoore said:

Yes, of course there is .... BECAUSE YOU ARE USING A FIR FILTER.

 

If you use an IIR filter (which is more common), there will be no delay.

 

This is what I've been saying all along.

 

 

Yes. They did... which is the same thing that *I* said.

 

I said....     "there's not ALWAYS delay problem .... only a delay problem when you use certain types of filters".

 

F-me.   This is just mean and stupid at it's finest.   Bye.

 

 

?

Link to comment
Share on other sites



On 09/12/2018 at 10:48 AM, zenelectro said:

At 44.1k  :)

 

Not a 1704 however not 44K either.

 

George put me onto this model a number of years ago and it been exceptional for me.  It has seen off a number of newer high priced dacs and cd players.

 

CAL Cl-15

CL-15 is based on the PowerBoss system—the combination of a Pacific Microsonics PMD-100 HDCD decoder/filter, dual colinear 20-bit Burr-Brown PCM1702P DACs, and a series of regulated power supplies. Matsushita supplies both transports, the CL-15 has a thick, single-disc drawer with silky-smooth action. Both players use discrete circuitry in their analog output stages, but the CL-15 adds a number of enhanced performance features. The first is a multilayer glass-epoxy circuit board that results in improved separation of the various circuits and a 4–5dB reduction in the noise floor. Other improvements include the CL-15's much more extensive power supply and regulation, and its true dual-mono construction, which starts with separate power transformers.

 

 

Link to comment
Share on other sites

35 minutes ago, DAMO 1147 said:

the combination of a Pacific Microsonics PMD-100 HDCD decoder/filter, dual colinear 20-bit Burr-Brown PCM1702P DACs

Lovely sounding unit, the PCM1702 is the older one to the 1704 that was 20bit but still a very nice sounding R2R dac, as is the PMD100 hdcd, the Cal CL15 for memory had a discrete output stage and I think discrete I/V stage also.  You should get a spare laser, before they become extinct. Matsushita RAE0113Z a (division of Technics)

https://www.ebay.com.au/sch/i.html?_from=R40&amp;_trksid=m570.l1313&amp;_nkw=RAE0113Z&amp;_sacat=0

 

Cheers George

Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...
To Top