Why is the nDAC so cheap?

quote:

Originally posted by gav111n:
Andy,

I can’t say I am totally convinced by my own response here. Let’s unpack it anyway:

The DSP is a naim controlled item. They know the DSPs frequency characteristic so can design around it. I am thinking here about things like PCB layout, component selection and positioning etc. The connected front-end is an unknown, in terms of both (1) (electrically) linking to a ‘thing’ with complex and unkown electromagnetic properties (as I think Richard is saying) and (2) an additional RF source in the form of an S/PDIF signal with unkown frequency fingerprint.

Let's unpack your answer . If you're talking about RF interfering with the DAC, then the way it can affect it is:

• Through the power supply
  
• Through the circuit boards
  
• Through radiated RF
  

The power supply is separate for the digital section and this would only be further separated by plugging an XPS in. The circuit boards are electrically isolated anyway, so that can't affect it, so the only option is through radiated RF. Having said that, I really can't see how a 1 bit signal being clocked in at the same sort of frequency as the DSP will have a significant impact on the RF characteristics. If Naim really wanted to remove RF, they could put a metal box around the DSP and associated circuitry (see: http://www.domin-connect.de/Atari/Lynx/l2_pcbtop.jpg for a quickly googled example).

quote:

I am in the ‘you can hear a difference with front-end’ camp but get the feeling that there is a complex interaction between the front-end and the nDAC. I suspect a more expensive CDP may not necessarily sound better through it compared with a cheaper one. It seems likely that a naim front-end will give the best results as naim will surely have optimised the nDAC to work with naim. Having said that, I use a MacBook Pro and think its great!

Gavin.

LOL.. How are you connected to the nDAC?

Andy,

I connect to the nDAC with a HiFace(BNC) and DC1(BNC-BNC).

I started off connecting with an Optichord minijack-toslink. I ran this for a couple of weeks and thought it was great. Then I swapped to the HiFace with DC1 and this made a real improvement. Seriously, not something you had to think about. Mrs. Gavin (who loves music but could not care less about the technical details of hifi) spotted it straight away. In a bet on spotting the difference in a side by side swap over, I would probably go for a £500 bet and think 'oh yes I can get a free powerline here'.

I also heard a comparison between HDX and CDX2.2 into the nDAC/555ps at the Heathrow show. I could hear a difference but Mrs. Gavin couldn't. I believe there was a sonic difference, however, in a bet on spotting the difference in a side by side swap over I would probably go for a £25 bet and think 'I should be able to get a free meal out'.

Gavin.

quote:

Originally posted by gav111n:
Andy,

I connect to the nDAC with a HiFace(BNC) and DC1(BNC-BNC).

I started off connecting with an Optichord minijack-toslink. I ran this for a couple of weeks and thought it was great. Then I swapped to the HiFace with DC1 and this made a real improvement. Seriously, not something you had to think about. Mrs. Gavin (who loves music but could not care less about the technical details of hifi) spotted it straight away. In a bet on spotting the difference in a side by side swap over, I would probably go for a £500 bet and think 'oh yes I can get a free powerline here'.

Are you sure you had 44.1 coming out of the mini-jack? Unfortunately, you can't tell what you do have going into the nDAC - and I've just been playing around with my second system and whilst xbmc will switch to 44.1 on playing back CDs, foobar won't without an awful lot of persuasion through the same interface! The resampling will make an awful lot of difference to sound quality.

I have to say, I'm toying with the idea of getting an M-audio to check if it makes a difference and to be able to capture the bitstream off the toslink to check it is bit perfect

quote:

I also heard a comparison between HDX and CDX2.2 into the nDAC/555ps at the Heathrow show. I could hear a difference but Mrs. Gavin couldn't. I believe there was a sonic difference, however, in a bet on spotting the difference in a side by side swap over I would probably go for a £25 bet and think 'I should be able to get a free meal out'.

Gavin.

LOL.. like the comparisons.

I'm still not fully convinced that the nDAC can be fully divorced from the characteristics of the transport - i think we all agree that Jitter is not an issue...

You still have the issue of the clock from the original SPDIF data which the DAC needs to recover to set its buffer readout clock (the 1 of 10 set frequencies) and also to clock the data into the buffer in the first place. This clock is close but not related to the selected ouput clock. Whilst i can see Naim have gone to great lengths to minimise interactions between different parts of the circuit (layout, PCB design, supply isolation etc) there is still potential for some parastic coupling between these clocks. As i put in a previous post, if the RAM buffer could be loaded with the complete track and the the clock stopped on the buffer input then you'd be playing straight from memory with one single clock and complete isolation from the transport.

Reports from users seem to indicate that playing files from the USB stick sounds better - here the DAC is generating the readout clocks and i would take it that this internal clock has a fixed relationship with the nDAC internal clocking structure compared to the averaged clock recovered from the SPDIF link.

James

Very true James. Perhaps I'm just lucky not being able to hear a difference between USB and optical.

Having said that, I can tell the difference between different interconnects between nDAC and amp and also I can also hear the complete loss of rhythm and timing when I route the sound out of the back of the 52, into the TV and back to the 52 (I do this to make it easy for the family to watch films off the htpc as we route the TV sound through the system...).

quote:

Originally posted by Richard Dane:
Just to throw something into the discussion here. In an email conversation with one Naim's digital engineers, he confirmed that the DAC does remove "all" (his parenthesis) spdif related jitter, and that this can easily be measured. However, he went on to say that nobody has yet figured out a way to avoid jitter completely...

Also, what may be of more blame for performance differences is the different levels of RFI that different sources emit. His comments; "You can actually easily hear if you just put a terminated (resistor in the end of the cable) spdif cable in the box, without electrically connecting it at all to the DAC".

Hi Richard -

With all due respect, and with the understanding that Naim engineering owes nothing to this forum, I would ask (again) for them to comment fully on this issue.

Until there is qualified technical response on the effects of jitter, electrical noise and RF interference, the evidence (i.e., Andy's argument) seems to weigh against there being a real difference in DAC transports, and for a mass placebo effect.

As I said before, I do hear a difference among my transports, but am totally open to explanations of why that is so. All in my head? Would be more willing to accept that if my two blind listeners didn't also hear the same differences...

Just don't see this issue going quietly away.

Hi Andy -

Just curious - how far are you willing to take this argument? Let's assume, for the moment, that you are correct, and all transports that output *bit perfect* over digital and/or optical S/PDIF to the DAC sound the same.

Do all digital cables then sound the same?

Do all software players (again, assuming the same source file delivered *bit perfect* to the computer's S/PDIF out) sound the same? Is, for example, Amarra software then built on a house of cards?

Thanks.

Hook

the major variable to all this,in most cases the dac will be connected in some way to pc of some kind that will give the chance of rfi being introduced to the audio chain.

quote:

Originally posted by ghook2020:
Just curious - how far are you willing to take this argument? Let's assume, for the moment, that you are correct, and all transports that output *bit perfect* over digital and/or optical S/PDIF to the DAC sound the same.

Do all digital cables then sound the same?

Assuming that there is no interference between source and DAC through some electrical connection, I don't see how you can come to any other conclusion.

quote:

Do all software players (again, assuming the same source file delivered *bit perfect* to the computer's S/PDIF out) sound the same?

This is one area where there are big areas for problems. Bits are bits if they are delivered without changing them. I can write a set of bits to disc and read them back a month later and they will be identical. How can different players sound different if they deliver the same bits to the output?

Unfortunately, bit perfect may not be easy to achieve (I'm not a Mac person, so don't know the details). Let me give you my experience (as a confirmed PC geek - I manage a network of 8 here in the house). I have a htpc running Linux and I bought a Cambridge Audio DACmagic a couple of weeks ago. It has LEDs on the front telling me the sample rate, so I know what the rate being received is. I spent some hours investigating how to get audio output to both the HDMI output and SPDIF on the HTPC. Finally, I found out how to configure it, but since my HDMI interface only supports 48kHz or 88.2kHz I wasn't getting 44.1kHz out of the SPDIF - only 48kHz and the data was being resampled . I've now configured things so that audio is only fed through the SPDIF interface and it outputs 44.1 and 48 (and 96!) depending on the source material. However, that isn't the end of the story - bit perfect means bit perfect. You have to make sure that you have no digital volume controls in place - in my case, that's with the volume full up (0dB attenuation). Take it anywhere from there and things do change as you're no longer bit perfect.

Also, on my Windows (XP) system that the DACmagic now inhabits, I can get xbmc to output bit perfect through the USB port easily, but I have to install ASIO for foobar to output bit perfect. Thanks Microsoft for specifying everything should be output at 48kHz!!

At present, I think you need to understand every point in the digital playback chain to ensure you get bit perfect output.

quote:

Is, for example, Amarra software then built on a house of cards?

Thanks.

Hook

No. It's aimed at giving the best possible reproduction, but that really should be a case of "read the bits and output them unchanged" which is pretty simple really - assuming the OS underneath you allows you to do this properly. What your chosen software does is implement the library functions - enabling you to browse your music collections. As I say, I use xbmc which provides an interface on the TV (although I can drive it from the 2 line display on my HTPC so the TV doesn't need to be on). All the rest (cover art/visualisations/sexy bars and graphs etc) are just window dressing and help attract you to the product in the first place.

I don't believe there is any magic in digital audio (although I have heard wild claims to the contrary) - especially with a well designed DAC like the nDAC.

quote:

Originally posted by rich46:
the major variable to all this,in most cases the dac will be connected in some way to pc of some kind that will give the chance of rfi being introduced to the audio chain.

Mine is connected by 5m of optical cable and is sited 3m from the DAC (with a hulking speaker in the way too!) which may go someway to explaining why I don't hear a difference. I have compared it to the same system connected (at the same time) with coax and couldn't tell a difference either...

quote:

Originally posted by JYOW:
We are all obsessed with all things Naim, but does this forum have to be a love fest?

I am missing on where this love fest is happening.

I totally agree with you JYOW, although Andy will argue with you for thinking there could be a difference.

He is unwilling to concede that.

In any case we ALL agree the Naim DAC is stunning, great VFM, and a top notch DAC.

The only argument happening now centers around Andy's complete unwillingness to accept that sources could sound different. Which MANY who own the DAC are finding for themselves with their ears (not a white paper).

-patrick

Hook,

Yes, that would be nice, but I don't think R&D will be in any rush to engage here, so I'll just have to keep plugging away...

As to evidence, I have all the evidence I need just from my subjective tests to show me that source really does make a difference with the nDAC (see my transport thread in the Hifi Corner). Like you, I have enlisted some blind test volunteers and the results were totally conclusive. Indeed, the differences are so obvious, even to non-enthusiast ears, that any suggestion of a placebo effect is faintly ridiculous in my opinion.

That's not to denigrate Andy's findings. He hears what he hears on his system in his environment, on his mains etc.. I hear what I hear on my system, in my environment, on my mains etc.. On the latter point at least, I'm out in the styx, with a 3 phase mains supply and newly refreshed mains wiring and plugs. Maybe that has an effect here? I don't know.

Either way, I'm open minded here. It would be nice to get to the bottom of it. I mean, if the theory goes that USB should be "perfect", why is it that I found the Meridian transport via DC1 into the nDAC/XPS to be the more enjoyable and engaging listen? The USB had the same disc ripped via EAC to WAV files. Perhaps there is some "loss" or effect of interference here that actually makes things sound nicer... And why is it that the same file on the iPod in ALAC (lossless, let's not forget) when fed digitally through the DAC sounds lacklustre compared to the others?

But OK, let's get some perspective here. None of the sources through the DAC could be considered poor, or even anything less than very good indeed. It's just that a couple in my tests managed something very special indeed. Ultimately, I'm not even sure I need to know why (although it would be nice), they just are and I'm having a ball...

quote:

Originally posted by pcstockton:
The only argument happening now centers around Andy's complete unwillingness to accept that sources could sound different. Which MANY who own the DAC are finding for themselves with their ears (not a white paper).

Really?

I accept that people hear sources differently (although I don't) and I'm trying to get an understanding of what is causing these differences since theory and the white paper says they should sound the same. The laws of physics are the same inside the nDAC as outside so either there is something that is causing the difference or it is a case of a mass placebo effect.

What no one has yet done is provide a credible explanation as to why different sources could sound different. "It just does" doesn't cut it for me given the extremely high cost of things at this level of hi-fi.

quote:

Originally posted by Richard Dane:
On the latter point at least, I'm out in the styx, with a 3 phase mains supply and newly refreshed mains wiring and plugs. Maybe that has an effect here? I don't know.

Just for completeness: I'm on a new suburban development. New house (and consequently wiring - all MK by the way) in Dec 2007....

quote:

Originally posted by Andy S:

quote:

Originally posted by ghook2020:
...

[QUOTE] Is, for example, Amarra software then built on a house of cards?

Thanks.

Hook

No. It's aimed at giving the best possible reproduction, but that really should be a case of "read the bits and output them unchanged" which is pretty simple really - assuming the OS underneath you allows you to do this properly. What your chosen software does is implement the library functions - enabling you to browse your music collections. As I say, I use xbmc which provides an interface on the TV (although I can drive it from the 2 line display on my HTPC so the TV doesn't need to be on). All the rest (cover art/visualisations/sexy bars and graphs etc) are just window dressing and help attract you to the product in the first place.

I don't believe there is any magic in digital audio (although I have heard wild claims to the contrary) - especially with a well designed DAC like the nDAC.

Hi Andy -

Am using a DIY PC (fanless, SSD) running Windows 7, J River Media Center 14 with WASAPI output plug-in, and an RME 9632 PCI card connected to the DAC with a custom DSub9-to-BNC cable. This is a combo that the folks at computeraudiophile swear up and down delivers bit perfect.

Also have a laptop using Mediamonkey with ASIO plug-in sending output through USB to an M-Audio Transit. It is connected to the DAC with a 5m Toslink. I have the Transit configured at 16/44.1, and have to manually change the bit depth/sampling rate when playing something of a higher resolution (and yes, the DAC's high rez light does go on). Again, no reason to think that MM/ASIO is not delivering bit perfect.

So, am convinced that both of these sources are delivering bit perfect.

And yet, the DIY box sounds better than the Laptop to my (and my blind tester's ears). Am willing to accept that it is placebo. But it is still the reality in my room. In fact, for me, this thread is primarily about intellectual curiosity, and determining whether or not there are holes in your argument. Even if there are no holes, I am still going to listening primarily to my DIY box....'cause I think it sounds better.

But there are a couple of things that bug me. First, there are people paying $1000 and more for Amarra (JRMC costs me $50). Even worse, and this pains me to say it, there are folks paying $5600 to only use the S/PDIF out of a CDX2. Until we hear Naim engineering fully comment on the potential difference among transports, and on the specific advantages of using the CDX2 (and/or CD5 XS) as a transport to the DAC, it will be difficult to feel very good about their value proposition.

Ditto for the HDX/DAC combo -- again, if there really is nothing that distinguishes transports other than their ability to provide bit perfect.

Hook

quote:

What no one has yet done is provide a credible explanation as to why different sources could sound different. "It just does" doesn't cut it for me given the extremely high cost of things at this level of hi-fi.

Quite true - apart from the limited info in the white paper, we are all second guessing Naims approach. The Ram buffer approach isn't a new concept in a DAC(Lavry Crystal lock on the DA924 for example) and i somehow doubt that Naim have managed to engineer a transport independent solution that had aluded some of the best digtial engineers in the business.

Have you managed to try a different transport yet ?

James

[QUOTE]Originally posted by Richard Dane:
Hook,

Yes, that would be nice, but I don't think R&D will be in any rush to engage here, so I'll just have to keep plugging away...

.......

Understood. Would rather they focus on building great new products! Still, my gut tells me this issue will not go quietly into the night until they do.

Please understand that I am 100% thrilled with the sound I am getting from the DAC/XPS2 in my room. Naim customer for life, etc.

Thing is I am feeling pretty good right now that I did not buy a Naim CD player and/or HDX to front-end the DAC. Would simply love to be told by the men in white that I am missing something beyond what I have now (EAC confirmed rips and bit perfect output).

Thanks.

Hook

quote:

Originally posted by james n:
Have you managed to try a different transport yet ?

James

No. Still waiting for my cable to come back from Naim...

quote:

Originally posted by ghook2020:
Laptop to my (and my blind tester's ears). Am willing to accept that it is placebo. But it is still the reality in my room. In fact, for me, this thread is primarily about intellectual curiosity, and determining whether or not there are holes in your argument. Even if there are no holes, I am still going to listening primarily to my DIY box....'cause I think it sounds better.

I'm at a loss to explain why there would be a difference - and like you I'm trying to understand what could cause it (btw, my htpc is a silent SSD atom/ion based PC). How big a difference would you say it is?

quote:

But there are a couple of things that bug me. First, there are people paying $1000 and more for Amarra (JRMC costs me $50). Even worse, and this pains me to say it, there are folks paying $5600 to only use the S/PDIF out of a CDX2. Until we hear Naim engineering fully comment on the potential difference among transports, and on the specific advantages of using the CDX2 (and/or CD5 XS) as a transport to the DAC, it will be difficult to feel very good about their value proposition.

Ditto for the HDX/DAC combo -- again, if there really is nothing that distinguishes transports other than their ability to provide bit perfect.

Hook

Yup. Agree with you 100% there My htpc cost around £300 and I'd need to see a huge difference to justify spending CDX2.2 territory money when I can't tell the difference between USB sticks and optical in from my htpc... A friend of mine is selling his CDS3 head end and replacing it with streamed music from his PC - so close are the two (through 252/250/DBLs) that he'd rather have the money back from the CDS3 (second hand, the head unit is worth more than the DAC new!)

quote:

Originally posted by ghook2020:
Thing is I am feeling pretty good right now that I did not buy a Naim CD player and/or HDX to front-end the DAC. Would simply love to be told by the men in white that I am missing something beyond what I have now (EAC confirmed rips and bit perfect output).

I fit in here too

Me too, to some extent. I put off buying a CD player a couple of years back, and don't think I ever will, now. However, I don't know that my iTunes/Mac Mini front end is the last word in sources. The UnitServe looks tempting. It seems the perfect partner for my SuperNait and HiCap2. I want to see a great iPad remote app for it first, though. That would be the duck's nuts.

quote:

Originally posted by ghook2020:
First, there are people paying $1000 and more for Amarra (JRMC costs me $50).

Sorry, I missed that first time around - some people have more sense than money....

quote:

Originally posted by ghook2020:
Even worse, and this pains me to say it, there are folks paying $5600 to only use the S/PDIF out of a CDX2.

Thankfully you dont need a CDX2 to enjoy crazy good music from the Naim DAC.

Those that can hear improvements from the CDX2 are welcome to spend their hard earned in that fashion.

No one is telling you that you MUST use the best possible transport/source. No one is even saying you need an expensive transport to get great music out of the DAC. All of us that have one know this.

But on the other hand, just because some of us (myself included) cannot, and will not spend $5000 on a transport does NOT diminish the enjoyment and improvement other find with more expensive sources.

Everyone needs to back away from the cost factor and supply their DACs with the most sensible source given SQ, budget, and logistics.

It shouldn't so mystifying that $5000 transport optimized to work within the confines of a Naim system sounds better than the coax output of a $15 sound card in a stock Dell. Is it?

My lord, if stacking arrangement of components on a Fraim has big effects on presentation, why would different sources sited to within a few feet of the DAC not sound different?

Bit perfect or not, if I place a $300 PC that vibrates and buzzes like a jackhammer when powered up on the top shelf of my Fraim, wont people think I am degrading the sound?

There are hundreds of ways, i imagine, to degrade what is going on inside the DAC beyond the simple digital inputs.

If you look at the issue from pure science and theory you may miss out on the fact that the "shitty" source is probably plugged into the same mains, sitting on the same rack, and probably connected with coax. Sure, if we could connect the DAC with a perfect connection from a bit-perfect source that is FULLY isolated from other kit in every possible way, we might be able to negate the effects of sources.

But as we all know, bumble bees fly.

If flat earth has taught me anything it is regardless of what the specs say, in the end all that matters is your own ears. And the ears are the single most relative and variable part of the hifi chain.

This argument will never end. Luckily everyone wins. Those who cannot hear a difference will benefit from spending their hard earned elsewhere. Those who can hear a difference have an option to upgrade their experience. Either way everyone is enjoying great music.

-Patrick

quote:

Originally posted by Andy S:
I accept that people hear sources differently

What no one has yet done is provide a credible explanation as to why different sources could sound different. "It just does" doesn't cut it for me given the extremely high cost of things at this level of hi-fi.

Its fine if people cannot successfully explain how it works. Especially when you consider that no one here knows as much as you do about it.

Then again no one owes you anything in the form of an explanation.

Maybe you could do what I do when I have discussions with Christians.... I accept their first principles and go from there. There is no point in me stopping the discussion at the fact that I don't believe in a god. I will instead accept their notion that the Christian god and its attributes actually exists. Then we can have a discussion from the same framework.

For example, we can argue all day long about god's omnipotence, but without me at least accepting the existence of god, how far would we get? Instead I accept God existence then discuss how certain attributes could not be possible.

Maybe you could "pretend"/accept that there is in fact a difference between say a CDX2 and a $30 Best Buy DVD player (assuming both are bit perfect). Now using your impressive wealth of knowledge on the Naim DAC and D-A conversion in general to help figure out some reasons how and why?

In this type of exercise, as with my god example, belief in first principles is not necessary in order to have a very fruitful, logical discussion.

-Patrick

quote:

Originally posted by pcstockton:
It shouldn't so mystifying that $5000 transport optimized to work within the confines of a Naim system sounds better than the coax output of a $15 sound card in a stock Dell. Is it?

Well, yes, that's the point I'm making. Bit perfect should mean same sound given the technology involved.

quote:

My lord, if stacking arrangement of components on a Fraim has big effects on presentation, why would different sources sited to within a few feet of the DAC not sound different?

Because the stacking has an effect due to analogue interactions such as vibration(at least that's my understanding...). Digital (until it's converted to analogue) either works or it doesn't.

quote:

Bit perfect or not, if I place a $300 PC that vibrates and buzzes like a jackhammer when powered up on the top shelf of my Fraim, wont people think I am degrading the sound?

If you're putting a $300 vibrating and buzzing PC on a rack, you've built the wrong PC

[/QUOTE]I'm at a loss to explain why there would be a difference - and like you I'm trying to understand what could cause it (btw, my htpc is a silent SSD atom/ion based PC). How big a difference would you say it is?

If I listen casually, not much. If I listen critically, it is more significant. Here's one example, maybe not the best, but I recall it clearly.

Krauss & Plant's "Raising Sand" is a CD that has some eccentricities. The mix is extremely bass heavy, and I think they used some old fashioned mics (ball & biscuit?) to give the recordings a unique, slightly hiss laden sound. Strange, but interesting, the CD does hold my attention.

Anyway, there is a song called "Fortune Teller". When played through my laptop, the bass sounds more one-note-ish, and the timing seems less crisp (harder to tell the begin/end of bass notes). When switching to my DIY PC, the bass became clearer and I noticed greater soundstage depth. Plant's voice and a drumstick tapping the drum's rim are both pretty much on the same plane when played through my laptop. On my DIY PC, Plant's voice comes forward and the rim tapping goes backward. Lastly, the rim tapping sounds more woody (less like electronic noise bursts) on the DIY PC.

My wife and her girlfriend also picked the DIY PC as sounding better, but couldn't articulate why. At that point, I explained what I thought I heard. We repeated the test (not blind at that stage), and they both agreed with my observation. So, not claiming any scientific controls were used, but I do recall clearly how they both got wide-eyed when they too could hear the change in soundstage depth.

Please understand that even though I am describing what we heard in some detail, these effects were very subtle. You had to really listen for them. But once we did, we all agreed they were real.

Hook

PS - If you are interested, you can see the specs for the PC I built by google'ing computeraudiophile and CAPS server. I described it in detail in a thread I started a while back -- was really easy to do.