• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Why do NOS dacs sound different to oversampling designs?

THW

Active Member
Patreon Donor
Joined
Nov 21, 2018
Messages
296
Likes
402
I don't mind derogatory comments. :p It shows your sucking attidude and complete lack understanding of physics.

The most fatiguing is a complex music. A single vocal, leading instruments like piano, guitar is not (but is losing details over the time). Calvyn Harris is not - there is no much to process from square waves. But complex harmonics like those created by choir or symphonic orchestra is.
im not sure how i was being derogatory, but ok.

the way i see it is this, your signal chain isn't made up of just the DAC, it is also made up of other components. if the DAC is measurably transparent then to me it doesn't really make sense to try to pin the problem that you have with your listening experience to the DAC rather than other components in the chain, to me it just seems like trying to fix the problem by not actually fixing the problem.
 

graz_lag

Major Contributor
Joined
Oct 13, 2018
Messages
1,230
Likes
1,255
Location
Le Mans, France
OK, so no evidence. Sighted listening and anecdote are completely unreliable for evaluation, a point that's been beaten to death.
It looks like the chicken-and-egg dilemma ... which of the two parties - manufacturers or customers, should be considered the cause and which should be considered the effect: (simply maybe both with equal responsibility :oops:)

From NAIM UK, but there are millions of similar statements from dozens of other manufacturers:

Modern low-bit delta-sigma DACs, which operate on an entirely different principle, are less costly to make and provide superior measured performance – but R2R conversion is preferred by many audiophiles for its superior sound quality.
The benefits of noise-shaping delta-sigma converters are such that they have come to dominate the design of high performance audio DACs in the past two decades. But the sound quality achieved by the best R2R ladder DACs is still considered by many critical listeners to be superior.


And so as a result, NAIM sells it's ND 555 Streamer/DAC @ $16,500 and TotalDAC it's TotalCRAP of a DAC @ $13,000!
 

sajunky

Active Member
Joined
Jul 14, 2019
Messages
186
Likes
55
Location
South Africa
Just made a photo:
Creative_AWE64_CT4380.jpg
Is there proof for this other than anecdotal and sighted listening tests ?
My life is a proof. For 20 years I was happy, but it is stolen. Now looking for something similar that gives me long-term listening pleasure.
It's the tolerance of the MSB that is problematic, not the LSB.
It's the MSB that is monitored and adjusted.
It is more complicated (current source), but yes I tend to test respondent's capacity...
The reasons are higher output current so less noise and errors getting smaller due to averaging.
Avergaing could be beneficial because errors get smaller.
The 'hype' however is because it is supposed to sound better. Which of course is found 'sighted'.
Is there compelling evidence showing a noticeable improvement in THD ?
Now I see, you knew the answer, but you was testing me.
BTW, Where I did say TDH? It is more than THD. It is about reproduction of tonal coherence, so I can recognise small variations of frequency changes I think. I will come with details this when talking to the other respondent.
Specs are given at 192kHz as mentioned in the sata sheet. Both TDA1543 and TDA1387 are intended to be used oversampled.
They specified it this way because 44.1 would not look great nor was it designed for that.
I have responded already in #E. My PCM63 DAC was also oversampling in the spec sheet. Also in implementation if I remember correctly.

I think, discussion over oversampling or not, filter/filterless will not give any clue. It is something different and more important.
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
3,333
Likes
5,835
Location
The Neverlands
So you think someone bought up thousands of unsold boards and salvaged parts ? Presumably only the DAC chips and then sell (untested ?) desoldered chips ?
because they are so incredibly good and innovative ?

My life is a proof. For 20 years I was happy, but it is stolen. Now looking for something similar that gives me long-term listening pleasure.
I can understand your frustration.
Your life is proof of your existence.

It is more complicated (current source), but yes I tend to test respondent's capacity...
Still MSB, 2SB, 3SB etc. just not one current source but a lot less accurate ones (cheaper to make) which are combined.
Is that technique also used in the next generation of DAC chips that followed ?

Now I see, you knew the answer, but you was testing me.
BTW, Where I did say TDH? It is more than THD. It is about reproduction of tonal coherence, so I can recognise small variations of frequency changes I think. I will come with details this when talking to the other respondent.
Why would 'tonal coherence' be better in filterless NOS R2R ?
Small frequency changes can only occur with poor X'tals. Why would frequencies change or not change ?
Above 7kHz frequency reproduction of filterless R2R is arguably and provable MUCH MUCH worse than any DS DAC or filtered and upsampled R2R DAC. So... that can't be it. I guess you need to look for another reason.

I think, discussion over oversampling or not, filter/filterless will not give any clue. It is something different and more important.
What exactly is different and more important. There is absoultely NOTHING more accurate about filterless NOS DACs.
 

sajunky

Active Member
Joined
Jul 14, 2019
Messages
186
Likes
55
Location
South Africa
im not sure how i was being derogatory, but ok.
As long was unintended, I am not taking attention anymore.
the way i see it is this, your signal chain isn't made up of just the DAC, it is also made up of other components. if the DAC is measurably transparent then to me it doesn't really make sense to try to pin the problem that you have with your listening experience to the DAC rather than other components in the chain, to me it just seems like trying to fix the problem by not actually fixing the problem.
It is definitely in the chain. I got Topping D30 first in the attempt to restore my stolen equipment, driving some crappy amplifier, so I decided to upgrade this part. FX502SPro brought great improvement in clarity over the old NAD319 (not the best model, never mind, I was happy). However the fatigue element didn't disappear. I blamed some distortion or colouration in the 5-7kHz range, so I got IcePower 125ASX2. This amp is absolutely neutral, ice cold, but there is still the same problem. It might be combination with D30, as both share the same characteristic. Individulal solo vocals and instruments are reproduced well, but distortions come up with complex harmonics. But I want to say more:

When 20 years ago I decided for PCM63 DAC it was the same problem, even worse, as all one-bit delta-sigma players were distorting much more. However distortions were not on my list of interest. I had one CD, a superb quality recording of Beethoven 9 Symphony that I brought to many showrooms. There was a moment when a complex choir passage turned to a silence and a hudge gong hit. You know how it sounds in nature. It creates a basic tone, then sound is split in overtones, combined with continuous reverbation changes that lasts for couple seconds. PCM63 (and some other DACs I couldn't afford) reproduced it properly, but all Delta-Sigma players completely failed, giving a single BOOM for a half second of lower amplitude and it disappeared quickly in the background giving no perception of reverbations.

Lets give you another example. My favourite contemporary composer is Arvo Pärt. There is a strange CD from ECM ARVO PÄRT: ALINA . You must listen this on a good equipment. Alexander Malter plays all the time with overtones keeping piano pedal engaged most of the time, pressing the next key only in the right moment, tuning to the existing reverbations. It creates complex resonance between individual strings of the same tone, wood case and other tone registers. It sounded superb when my HiFi was properly tweaked. Now I downloaded the same CD, even the HiRes version, it sounds with superb clarity I never had before, but flat. It is why I think it is a DAC. Yes, I conclude, it is the same story of delta-sigma DACs.
 

sajunky

Active Member
Joined
Jul 14, 2019
Messages
186
Likes
55
Location
South Africa
Do the comparisons level-matched and ears-only (no peeking, double-blind) and those claims can be taken more seriously.
What???

The flatness, lack of reverbation of a gong I mentioned above is so severe that there is no reason for blind tests. Listen to the Alexander Malter performance on the Topping D30 and you have no idea why he pressed key a second later. What a hell the guy is doing? Why? You won't ask these question when listening the same track on the PCM63 DAC.
 

THW

Active Member
Patreon Donor
Joined
Nov 21, 2018
Messages
296
Likes
402
What???

The flatness, lack of reverbation of a gong I mentioned above is so severe that there is no reason for blind tests. Listen to the Alexander Malter performance on the Topping D30 and you have no idea why he pressed key a second later. What a hell the guy is doing? Why? You won't ask these question when listening the same track on the PCM63 DAC.
Because psychological bias will skew your perception of the sound, many times in ways you actually don’t even realise?

http://seanolive.blogspot.com/2009/04/dishonesty-of-sighted-audio-product.html?m=1
 

SIY

Technical Expert
Technical Expert
Joined
Apr 6, 2018
Messages
3,208
Likes
5,955
Location
Phoenix, AZ
Basic controls. I mean really basic. Especially the level matching, which very few faith-based audiophiles will bother doing but is absolutely essential.

It is amazing how some of those night and day differences disappear when you have to just rely on your ears. The first time I tried this with a night and day difference (it was a tube preamp vs a solid state preamp), I was stunned, and re-examined my own belief system. But that's the scientist in me- if what I know is true doesn't stand up to a real test, I discard what I "know."
 

Blumlein 88

Major Contributor
Joined
Feb 23, 2016
Messages
7,234
Likes
7,943
What???

The flatness, lack of reverbation of a gong I mentioned above is so severe that there is no reason for blind tests. Listen to the Alexander Malter performance on the Topping D30 and you have no idea why he pressed key a second later. What a hell the guy is doing? Why? You won't ask these question when listening the same track on the PCM63 DAC.
Such claims are made often. So obvious no need to test blind. Yet when sighted knowledge is removed, all those differences disappear as well. What you are saying about reverb is highly improbable considering how the gear works.
 
Last edited:

sajunky

Active Member
Joined
Jul 14, 2019
Messages
186
Likes
55
Location
South Africa
Such claims are made often. So obvious no need to test blind. Yet when sighted knowledge is removed, all those differences disappear as well. What you are saying about reverb is highly improbably considering how the gear works.
I do appreciate your effort, but blind tests are done when differences are small. The effects I described are severe. Even if on delta-sigma DAC you pop volume much higher the gong will sound flat and will decay below the background much faster. If you don't have a good R2R converter for comparison, go to the show room.
 

SIY

Technical Expert
Technical Expert
Joined
Apr 6, 2018
Messages
3,208
Likes
5,955
Location
Phoenix, AZ
I do appreciate your effort, but blind tests are done when differences are small. The effects I described are severe. Even if on delta-sigma DAC you pop volume much higher the gong will sound flat and will decay below the background much faster. If you don't have a good R2R converter for comparison, go to the show room.
OK, so your claims can't be taken seriously.
 

j_j

Senior Member
Audio Luminary
Technical Expert
Joined
Oct 10, 2017
Messages
379
Likes
514
Location
My dining room.
May I stir up the discussion a bit? NOS DACs need no filter because there always exist two filters: number one is the speakers/phones that shut off around 20 kHz. Number 2 are our ears.

Of course that won't prevent me from doing it right in the first place, but....you get it.
Do you expect that amp, speaker, and ears are all linear?

If you expect that, you will be profoundly disappointed, and in that, your hope of using them as an anti-imaging filter fails miserably. Do it right.
 

j_j

Senior Member
Audio Luminary
Technical Expert
Joined
Oct 10, 2017
Messages
379
Likes
514
Location
My dining room.
Stability of the amplifier is an output load thing.
I would be more afraid of non linear behavior of amplifiers which creates aliasing and speakers that behave non linear in ultrasonics and create aliasing.
I don't think you mean "aliasing" but rather distortion products resulting from the aliasing, and yes, you're quite right, it's a problem.
Furthermore digital amps that do not have very effective brickwall filtering on the inputs can also create audible aliasing in the audible band.
Best to sufficiently filter out all non-essential crap in any case. There is no excuse or justification for not filtering out 'crap'.
I think "AMEN" covers it.

Most speaker tube amps roll-off just within or without the audible band (not steep enough) and may well be less problematic (depends on the actual design)
Except for the performance of the output transformers, core losses, core linearity, etc, which can be kind of startling in some cases, as well. The spectrum will be of a different sort, and probably lesser, than, say, a Class D responding to signals near it's switching frequency, for sure.
 

j_j

Senior Member
Audio Luminary
Technical Expert
Joined
Oct 10, 2017
Messages
379
Likes
514
Location
My dining room.
I do appreciate your effort, but blind tests are done when differences are small. The effects I described are severe. Even if on delta-sigma DAC you pop volume much higher the gong will sound flat and will decay below the background much faster. If you don't have a good R2R converter for comparison, go to the show room.
No, that's not correct. Blind tests are done in order to avoid inadvertent self-influence.

I have done many blind tests, way too many, sometimes it seems, but it's the only way to accurately establish an audible difference, even of differences described as "large" or "obvious".

In addition you have blown right past the issue of level match, which can completely confound any test, even a blind test. I would suggest that you work on testing your assertions in a double-blind setting, be it ABX, ABC/hr, or 3/4 signal descent, or something else if you can accurately describe the test and its evaluation.

So, no, your "severe" claims are more suggestive of a problem somewhere than any sort of relationship between r2r and delta-sigma DAC's.

Another thing to consider, if you are using signals that peak above digital zero and therefore clip, are the possibility of intersample overs, which is an issue I've had to examine recently, and I must say that some DAC systems do a much better job than others. Some otherwise very good DAC's do a particularly awful job on intersample overs, I'm sorry to say.

None of that would read on your claims about reverberation at all, of course, but it could relate to initial attacks if a recording was unwisely conditioned, and many, many recordings in the modern day are far, far beyond merely "unwisely conditioned". By how much, I'm appalled.
 
Last edited:

j_j

Senior Member
Audio Luminary
Technical Expert
Joined
Oct 10, 2017
Messages
379
Likes
514
Location
My dining room.
Have a closer look to an 'impulse' in music or a crescendo or sudden loud noise in any editing software.
You will see that such an 'impulse' spans multiple samples.
Well, since all real PCM systems must have an anti-alaising filter that starts to cut off below half the sampling frequency, you can not, ever, get a single "impulse", only a train of them. No real input into a real anti-aliasing filter should be able to create such a signal, because it contains frequencies that were filtered out.

Most "attacks" are much wider than two samples, in fact, because of the impulse response of the anti-aliasing filter. The closer the filter cutoff gets to half the sampling frequency, the longer that impulse response *MUST* be. As they say, it's not just the law, it's mathematics.

With a NOS DAC that impulse becomes 'jagged' and has extremely poor timing characteristics because it rises and drops at the wrong moment and stepped and too steep.
BUT at the output the proper anti-imaging filter will remove those 'edges' completely. So this is not actually a real issue if filtering is properly done. That word "properly" has some connotations, of course. Except, well, continue below, please.

So while some people like it and feel impulse response has improved (because it is said to be so) in reality the actual signal is 'f'ed up, has lots of ultrasonic crap that might become problematic. In most cases it isn't problematic and the hearing is so crappy and bandwidth limited that it isn't even that audible when average roll-off is 'compensated'. Only when 44.1 and 48kHz files are used. For higher sampling frequencies the EQ must be 'off'
Now you're talking about the (sin x)/x rolloff for a step function. Now this gets really complicated. But let's assume that the sinc compensation is literally perfect (HAH!) just for grins, ok?

In an r2r dac, let's assume that the MSB is perfectly accurate (HAH! fat chance), so are all other bits (***HAH***) and that there are no low-level nonlinearities as a result. It's a PERFECT r2r DAC in terms of level out vs. digital in.

That's going to give you a perfect output spectrum, right?

Well NOT necessarily. The dynamic behavior of each of all of the switches in the ladder must ALSO be absolutely identical (if you think level is hard, this is even worse). Furthermore, there must be zero capacative coupling from switch control voltage to the output r2r current. Yeah. Zero. Physics. That again. :) So we put in a sample/hold after the DAC, so we can capture the "perfect" level with the SH system. I've spent many nights working with SH systems long ago. Aside from dielectric absorption, switch pedestals, droop, and input-level dependent jitter, all of which are bloody hard problems, yeah, let's assume you whupped all that, and it's PERFECT.

So then you're good, right?

Oops. No. IN ADDITION the waveshape of the "steps" must be a linear function, i.e. exponential approach to each level, NOT slew-rate limited approach.

Yes, I turned this into a discussion of why r2r DAC's are a (*&(*& to build. It wasn't really an accident. :D

The point is simple, 2x oversampling or 4x oversampling is a no-brainer for a bazillion reasons, and going to delta-sigma with good signal processing likewise, that way there's just no nonsense about low-level linearity, etc.

Such is life. It is possible to get an 18 bit Burr Brown DAC to 95.5dB SNR, including ground isolation, output buffering, etc. Been there, done that. Never EVER want to do that again. Yes, that's 16 bits, but that was the actual spec on the device, so I don't feel too bad about that.
 
Last edited:

sajunky

Active Member
Joined
Jul 14, 2019
Messages
186
Likes
55
Location
South Africa
No, that's not correct. Blind tests are done in order to avoid inadvertent self-influence.

I have done many blind tests, way too many, sometimes it seems, but it's the only way to accurately establish an audible difference, even of differences described as "large" or "obvious".
You are talking about small differences that can change your perception based on auto suggestion, so you need blind tests. This inferior reproduction of tonal variations, overtones or reverbations in delta-sigma modulator is perceived exactly the same by many people. When I go to the showroom I don't explain details, but I present a sample. When the response is repeatedly the same (sometimes expressed in different words - a common one is a 'full body' or lack of), it is conclusive. So I did in fact blind tests maybe not sufficient to strict requirements, but repeatedly accurate.
In addition you have blown right past the issue of level match, which can completely confound any test, even a blind test. I would suggest that you work on testing your assertions in a double-blind setting, be it ABX, ABC/hr, or 3/4 signal descent, or something else if you can accurately describe the test and its evaluation.
Level match was set equal, this is what we normally do. I did even set delta-sigma player lever higher at times to see whether it helps, but it didn't. Don't put in my mouth things I didn't do.
Another thing to consider, if you are using signals that peak above digital zero and therefore clip, are the possibility of intersample overs, which is an issue I've had to examine recently, and I must say that some DAC systems do a much better job than others. Some otherwise very good DAC's do a particularly awful job on intersample overs, I'm sorry to say.
I explained it was a very good quality recording (Deutche Grammophone, Telarc or Sony, I don't have it anymore). Before the gong was engaged there was a culmination of choir performance, very intensive, then a second of silence. The amplitude of the gong was about -40dB, VU meter merely jumped. There is no case of clipping. A similar amplitude are reverbation of piano recording, you should know about.
None of that would read on your claims about reverberation at all, of course, but it could relate to initial attacks if a recording was unwisely conditioned, and many, many recordings in the modern day are far, far beyond merely "unwisely conditioned". By how much, I'm appalled.
Man, I am talking about a quality recording. Want to test? Get a sample of ARVO PÄRT: ALINA and will see. I am sure result will be the same with all delta-sigma DACs, sometimes cleaner, but the same averaging of the lower volume frequency tones.
 

SIY

Technical Expert
Technical Expert
Joined
Apr 6, 2018
Messages
3,208
Likes
5,955
Location
Phoenix, AZ
You are talking about small differences that can change your perception based on auto suggestion, so you need blind tests. This inferior reproduction of tonal variations, overtones or reverbations in delta-sigma modulator is perceived exactly the same by many people. When I go to the showroom I don't explain details, but I present a sample. When the response is repeatedly the same (sometimes expressed in different words - a common one is a 'full body' or lack of), it is conclusive. So I did in fact blind tests maybe not sufficient to strict requirements, but repeatedly accurate.

Level match was set equal, this is what we normally do. I did even set delta-sigma player lever higher at times to see whether it helps, but it didn't. Don't put in my mouth things I didn't do.

I explained it was a very good quality recording (Deutche Grammophone, Telarc or Sony, I don't have it anymore). Before the gong was engaged there was a culmination of choir performance, very intensive, then a second of silence. The amplitude of the gong was about -40dB, VU meter merely jumped. There is no case of clipping. A similar amplitude are reverbation of piano recording, you should know about.

Man, I am talking about a quality recording. Want to test? Get a sample of ARVO PÄRT: ALINA and will see. I am sure result will be the same with all delta-sigma DACs, sometimes cleaner, but the same averaging of the lower volume frequency tones.
Do a real listening test with real controls, otherwise this has all the evidentiary value of people claiming to be kidnapped and anally probed by space aliens. Stamping your feet and endlessly repeating your unsubstantiated claims does not make them any more true.
 
Top Bottom