Denafrips DACs might not actually be NOS?

[PeterSt]

12 hours ago, GoldenOne said:

It'll be interesting to try nos on the phasure NOS1a if I can actually get it working. At the moment it's an H shaped paperweight

 

An H shaped paperweight which will byte you with its legs if you dare use it without reconstruction in advance of it.

You may like to play test signals at 16/44.1 though to prove its NOS-being. Note that XXHE will not be able to do that natively because it will always upsample 2x, never mind it's 16/44.1 still (it will just be 2x of the same).

 

12 hours ago, GoldenOne said:

I think technically though if upsampling to 705.6/768mhz then the dac filter itself makes very little difference regardless of design (so long as its not doing something completely crazy).

 

I would not count on that too much. :-) However, do notice your own context, which responds to other posts. Without context you are completely correct. In-context, upsampling to 705.6 without filtering will give very poor THD on to an NOS DAC. With proper filtering (and the same 16x upsampling) will be way way better. You have a real NOS/Filterless (R2R) at hand, so go ahead ...

 

 

[PeterSt]

1 minute ago, numlog said:

The advantage of a real NOS DAC is it will remain internally NOS at any sample rate, 88.2k up to 1.5M, so you have you full external control.

 

Exactly. And THE reason why the NOS1 was created. It came along with playback software ...

So Yes, I created the filtering first. The genuine NOS allowed me to judge it best.

Another thing - just saying - the NOS1 "chain" was made such, that at each sample rate it behaves / should behave electrically 100% the same. See the "2x of the same" I just mentioned for 16/44.1. That's a similar thing.

[PeterSt]

18 hours ago, idiot_savant said:

My take on this isn’t actually about a “gotcha”, denafrips are evil or rubbish - it’s just that there is so much speculation about how something may be better even though you can’t measure it, that I’d hope people would pick up on something that is clearly being measured. 

 

Hey Idiot, it is good to have you back ! I hope you stay for a while. Us morons seriously could use some idiot, once in a while.

[PeterSt]

From the Ares II manual :

 

 

At first I thought this was Chinese odd language. But this kept on bothering me, from where ever I read it :

 

 

This THD+N can not be mentioned without telling whether this is in OS Mode or in NOS mode. Plus, in NOS mode the figure would be way too good because it should be 0.04%.

 

Back to the first screenshot and the Chinese language :

 

NOS, as the name suggested, does not over-sampling to digital input data.

 

For me this is just too much emphasis on (the superfluous) "Input". But next up is:

- An FPGA which could do anything - mind you, this is way beyond the "input" as such (in our NOS1 situation it is right in front of the D/A);

- Nothing tells anywhere that this is about filterless, when it is about NOS (mentioned).

- It even is so that the filter responses are mentioned without telling that this can't work in NOS mode. However, it would be so that the filters can not be selected in NOS mode (dunno where I read that, but I guess @GoldenOne can confirm easily.

 

But now about this one:

 

 

It may again be about how to decompose this Chinese language, but what I (like to !) read is that there's analogue filtering in there. And otherwise I may not be knowing what an analogue "filter" as such is doing there for decoding DSD.

Anyway ...

 

A NOS DAC is only a genuine NOS DAC when /filterless is mentioned as well. I mean, nobody wants a NOS DAC with filtering behind it. Also not in advance of it. Thus, NOS/Filterless.

So as How I read it, the Ares II could be a NOS DAC all right, but it is not filterless ?

 

Which of course is different from people seeing linear interpolation in it. But this should be easy to check for literally that. IIRC Linear interpolation just does not want to work well above 11.05 ... (or a bit above that - I forgot).

Otherwise the FPGA could be doing things (wrongly). ... My own design of the NOS2 (which never saw real light) contained very complicated stuff around PCM and DSD at the same time, as the FPGA turned PCM into DSD-like behavior. I have no idea how it would have measured, but surely far from normal.

 

 

[PeterSt]

One last thing which could also be important:

 

Someone hinted at some calibration which could be going on. Well, I don't know how much the Chinese steal from each other (or from me, for example), but the manual explicitly tells about the audio data being read "into the memory of the FPGA".

Why ?!?

Well, if you provided the FPGA with some pre-calibrated "data" for the way (waayyyy) too rough resistors in there (0.01%) then calibration (a mapping) would sure help. And I think one of the Holo versions really does this. So some Chinese will know about this.

 

And from their website:

 

The cost of high precision resistors is expensive, each resistor must be stringently measured and matched by machine with human intervention to cross-check.

 

I say: BS. Nice story, but BS. Matching them, by machine or otherwise, is useless because at one degree temperature change, all is a mess again. I can show you my attempt with uncountable bins, all in vain when proceeding with it a next day. Of they think they calibrated them. This is how the real-time calibration (of the Holo ?) does work. 

I must add that only when you know what you are doing, you can see the mess. So, measuring a 0.01% resistor at two digits of precision (make 3 digits of it, logically), is useless when you know that for a 24 bit DAC you'd need 0.000022% of precision (I do this from the top of my head and there could be a 0 too many).

[PeterSt]

34 minutes ago, idiot_savant said:

Point of pedantry - OS/NOS/filtering won't strictly speaking affect THD - The H stands for Harmonics - so e.g. a 6kHz tone will generate HD products @ 12kHz, 18kHz, 24kHz.... NOS artifacts will be at Nyquist +/- signal - so a 6kHz tone will generate "image" products @ 38.1kHz, 50.1kHz, 82.2kHz, 94.2kHz and so on for 44.1k samples, which will typically be much, much bigger than THD in amplitude terms.

 

16/44.1

 

With Filtering (16x) :

 

 

 

 

Without filtering (still 16x):

 

 

And exactly the same as native 16/44.1 which should be 0.04 (I told about that in the topic from 60 minutes or so ago).

 

So I am confident we talk passed each other ?

 

Edit: this is all @1KHz.

[PeterSt]

Here's one I did not even ever look at myself ...

 

This is 24 bits 44.1 filtered (16x):

 

 

And this is 24 bits, not filtered (16 x) :

 

 

I don't know what I had to expect from this, but not that.

[PeterSt]

4 minutes ago, idiot_savant said:

you might be clipping slightly, so if you reduce the level by say 3dB and see what's happening?

 

It's already at -3dBFS.

 

4 minutes ago, idiot_savant said:

is this NOS doing a zero-order hold ( so sample-hold ) or a zero stuff?

 

Sample-hold. Well, it's PCM1704.

[PeterSt]

For further reference, this is what this DAC makes of 24/96 filtered 16x:

 

 

(we have this unfinished one coincidentally on the test bench, and later this will show as 0.00059% or so).

This is still better than "untreated" 24/96. With 24/192 there is no difference to be shown (plus it shows the same as the above plot).

[PeterSt]

5 minutes ago, idiot_savant said:

And you're *sure* that the 1kHz tone is -3dBFS going *in* to the DAC?

 

Yes, I am sure. I could show you clipping from a file not being -3dBFS. And do notice that my filtering is interpolating which does not allow for a higher output signal than its input signal (not for average and also not for "tops" (or valleys - haha)).

 

Btw, I could drop the signal another 3dBFS if you want, but currently someone is working on that DAC so I can't show you now.

 

Why would this matter with filtering improving ? (signal gets lower ? :-)

Anyway, we also test with lower output and this really does not matter (except for the S/N doing things, which makes THD+N also worse.

 

The 0.04% is a dead-normal figure as all NOS DACs show that.

On-topic and repeat: for the Denafrips the 0.004% looks strange to me. It should either be better (with filtering) or worse Genuine NOS).

[PeterSt]

17 minutes ago, idiot_savant said:

the DAC at 705.6kHz -so does the DAC input see sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample0,sample1,sample1,sample1,sample1,sample1,sample1,sample1 etc.

 

I suppose this was a question, right ?

Earlier on I wanted to say "I myself determine the sample and hold" but I was not sure about the decentness of the answer. Anyway, Yes. Each sample has an explicit value, determined by the in-PC filter.

[PeterSt]

22 minutes ago, idiot_savant said:

as the measurements show that even at 96kHz it is droopy in the audio band.

 

Thinking about this twice, I would say that this can only because by some error (bug).

OR

some erroneous analogue filter. This may not necessarily be a bug but it may mask something else purposely (take out some (whining) peak)

[PeterSt]

40 minutes ago, ecwl said:

But I see lots of people getting upset over NOS R2R not being truly R2R.

 

OK, part of the understanding could be about what's not truly which. 😏

 

An R2R DAC should be truly that, because it is built around R2R ladders (if the "division" - as in R/2R is different than I wouldn't care about that). I'd say that there should be no doubt, already because the ladders can be seen with discrete ladder DAC's (as with the Denafrips) or else the chips could be observed. And easy for 24 bits:

This can only be the PCM1704.

 

So I'd say your "statement" should be different:

"But I see lots of people getting upset over NOS R2R not being truly NOS."

 

Agreed ?

So if I Upsample in-PC (this is not Oversampling, BTW) is my DAC than NOS ? It definitely is. But if I "demand" people to use upsampling/filtering because that is the whole idea about it ? then the chain will net not be NOS at all. It still would be filterless as in "no analog filter". So mind you, this is quite crucial for NOS die-hards who now suddenly get some best of both worlds thing. You saw my THD+N figures and you can take it from me that genuine NOS is THD+N 0.04%. This is super bad. Add an analogue filter behind it, and I doubt whether THD+N (which is in-(audio-)band) will improve. However, it will prevent aliasing to some degree and that in itself would imply distortion (THD). Same with a decent (sufficiently steep) digital filter.

The difference with the analogue filter alone, however, would be the staircasing being way higher.

 

@idiot_savant, Please consider that as the reason of digital filtering really helping vs upsampling only not helping much. This should be in the aliasing department and not the staircasing thing (OK, 20KHz is a huge staircase, but ...).

 

@ecwl, I am afraid I drifted off. It is not all that easy. Maybe you should change some of your "NOS" or "R2R" parts into "PCM" parts. That compares better with (versus !) DSD.

[PeterSt]

19 minutes ago, idiot_savant said:

and the difficulty in making comparisons between oranges and apples. You're the one stating that there is no current hardware currently capable, without stating capable of what?

 

I hope you guys don't end up in a fight and that we must miss your insights for 10 years again. But ... maybe it is good to know that the world really has changed:

 

I would say that there's not a single soul any more who uses a DAC without the filtering in advance of it. Thus, everybody uses some kind of in-PC filtering. That by far most use HQP is something I can't help (they should be using XXHighEnd obviously :-) but hey, this is a fact. 

Fact is also that Miska from HQPlayer will tell you that no hardware can do the job he is capable of in-PC. This, of course, is BS. But in the end not so much BS at all, because it is about the WHAT is done, and like me in XXHighEnd, I can change it on a daily basis for you the customer, and it WILL determine most of the sound as such (I am not talking SQ per se, but just the nature of the sound). To each their own, but as long as we can set in-layer how the sound is to be, just by means of selecting a filter(type) which overrules what happens in-DAC and which almost per definition is worse-sounding (oops) ...

 

... then you should give in somewhat ?

I know of no-one who uses XXHighEnd and who finds HQP to be sounding better. The importance of this is ... it is the filter which determines the sound. That people (me included) coincidentally find the one player sounding better than the other, alas.

So let's turn this story upside down a little:

 

30 minutes ago, idiot_savant said:

Oh, and no analogue filter? So you have large images above 768k?

 

Yeah, isn't that great. Well, what if HQP could undo that by means of a lot of horse power.

I am not saying it can, but similar things it surely can do.

 

Or another twist: my filter virtually does not use any CPU at all (you will see 0% during playback only). And now I claim that sounds way better (hey, it does). It is still filter related because it is power consumption related.

Now who is right about what ?

And how to do something about that when all happens in-DAC ? ... oh wait, this thread is about that, right ?

 

haha

[PeterSt]

11 hours ago, semente said:

11 hours ago, PeterSt said:

I would say that there's not a single soul any more who uses a DAC without the filtering in advance of it. Thus, everybody uses some kind of in-PC filtering.

 

Peter, how long have you been in self isolation for? I am a member of other forums and most people there do not use a computer, nor a streamer.

 

OK, I am out of date then (in isolation for 14 months now, with my huge task of watering the office plants).

 

Do they celebrate high teas in those other forums ? 🤪

 

 

[PeterSt]

1 hour ago, idiot_savant said:

If you’re a red book NOS user it isn’t really

 

No, really not.

 

??

[PeterSt]

3 hours ago, semente said:

Are there different kinds of NoOverSampling?

 

With discrete ladder DACs, I am afraid the answer can be Yes, although this will probably not be what you mean (to ask).

 

... I am pretty sure that this boils down to what Oversampling actually is, and this is not the same as Upsampling. Examples:

 

Suppose I have that discrete DAC and it is 17 bits at least, then I would be able to UPsample the data in advance of the D/A, so 16/44.1 would become 17/88.2. The DAC would be NOS because, well, it is not OVERsampling.

Technically there would be no difference with the in-DAC solution vs the outboard (PC playback software) solution, and the sound will depend of the kind of filtering used. Also notice that leaving out reconstruction filtering will make no sense.

 

When I have an SDM (chip) based DAC and I feed that 16/44.1, it may OVERsample to whatever idiot rate, so the modulator works with decent feedback. The number of bits is not importance any more because that is how the modulator works (with 1 bit, or a few bits, whatever, but not 16 or 17 or 24 or 32). Although it may be hard to find proper definitions, I myself regard oversampling as a kind of sampling the samples (again). This should be done at a (way) higher rate than the original is.

Also, from an SDM based DAC it will generally not be known what its sampling rate is, because it is unimportant.

Notice that this is different from the INPUT rate of a DSD DAC. And as far as I can tell, that too can still oversample.

 

Do I make sense ?

Completely incorrect ?

 

PCM-me.

[PeterSt]

12 minutes ago, idiot_savant said:

Type A: This user wants to leverage the processing of a PC to do filtering and noise shaping to present very high rate processed data to the DAC, with as little DSP inside the DAC as possible to interfere with the PC processing

 

I am sorry, but it this is the way NOS is explained, than about everything is OS, if only "as little as possible DSP" is performed.

So No, no DSP at all.

 

One thing could be excluded, and this is volume control. That is, if we regard that DSP too. But on top of it all, the volume control, thus proposed as digital, should be lossless.

 

I myself would not even incorporate dither.

[PeterSt]

1 minute ago, numlog said:

Oversampling wouldn't be something you see ever in PCM dacs?

 

This is something I am not good at, but I would say that this can exist, BUT then it is not a NOS DAC. Thus, PCM is not equal to NOS.

The other day I mentioned the PCM1792 ad a hybrid D/A chip. IIRC it can upsample (???) and use its filters from there, or it may not upsample, but still the filter behind it can't be shut off. It would be a PCM DAC, that upsamples.

I hope someone corrects me here.

 

Possibly all DACs using PCMxxxxx chips should be called PCM DAC's. But this is then only because it is not an SDM based DAC.

I should be looking in the datasheet for this ...

The difference would be "no modulator" involved with the PCM based DAC.

 

There is (was) only one chip which was able to "be" NOS at 24 bits, and this is the PCM1704. Officially that one was to be used with the filter chip behind it (DF1704 or something like that). Here the "trick" was that the PCM1704 was UPsampled towards that filtering chip, BUT that we could leave out that filtering chip and apply the upsampling IN ADVANCE. This is how the NOS1 emerged and how it can do 24/768. And the by now known specialty, do the upsampling/filtering in-PC.

 

To be not too much confusing, I should emphasize that both Miska and me were working on the very same principle, say from of 15 years ago : overrule what the DAC does with PC software. And the nice division: I do it with PCM and he does it with DSD (or makes it that). This makes the target DACs inherently different (PCM vs SDM based, respectively).

[PeterSt]

You can dive endlessly in the specs of this chip and still end up with not what you want:

 

 

 

But look at the 8x oversampling digital filter.

Nobody (datasheets) calls that upsampling.

 

Now look at my or anyone's playback software. Nobody calls that oversampling.

 

The PCM1792 is natively 768 (see above last line). The 8x oversampling exists outside of that. Does this make the 1792 a 6144K sampling chip ? no. Instead the 768 is "re-sampled" (hey !) and the result is for the filter (or it *is* that right away).

 

Often the oversampling is related to the native sampling rate of the chip. For example, it can take 768, and o.s. 8x but if it is fed with 384 it may o.s. 16x.

 

PS: While this is from 2006 edit 2003, the 1704 is from 1996 or whatever exactly. Still this 1792 is worthless for NOS. It is good for DSD, however (no 1704 can do that).

 

Blurp.

 

[PeterSt]

3 minutes ago, PeterSt said:

The PCM1792 is natively 768 (see above last line).

 

This is probably not correct and officially it is 192. Apologies.

Similar (but other way around) with the 1704. This is officially 96 but can do 768.

[PeterSt]

1 minute ago, The Computer Audiophile said:

Most of the time it isn’t ok to boast about your products, like you’ve done in this thread. 

 

I understand. If I would have one other example, I would have used that definitely. But I know of no other example. I will also 100% definitely know about my own product.

 

I'll be extra careful. Thank you for the warning.

[PeterSt]

Quoting from a very good post:

 

11 minutes ago, idiot_savant said:

Now I'm guessing that it *is* genuinely NOS at 1.5MHz, but we *know* it isn't at say 192k, so when the linear interpolate is active the PC provided noise shaping is effectively ruined.

 

I wonder whether you/someone noticed this from the  Stereophile measurements:

 

The early roll of (L/R) is from the 192KHz sampling rate. The later (but still strangely early) roll of is from 96KHz.

So again I just see a bug in the (FPGA) software somewhere.

 

But otherwise, I.S., maybe this gives you a clue somewhere ?

[PeterSt]

16 minutes ago, The Computer Audiophile said:

Here are the settings I use in HQP with the Terminator.

 

Chris, DAC Bits 20 would be insufficient for upsampling 32x.

Possibly I misinterpret things (I don't know HQP well).

32x in my own book requires 21 bits at least.