• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Why do NOS dacs sound different to oversampling designs?

daftcombo

Major Contributor
Forum Donor
Joined
Feb 5, 2019
Messages
3,687
Likes
4,068
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.

Hi,
What about IMD that could be triggered down the chain because of the transmission of ultrasonic content by the NOS DAC?
 

abm0

Active Member
Forum Donor
Joined
May 12, 2019
Messages
126
Likes
58
Kunchur made some fundamental signal processing mistakes, and moved on from there.
So far I see that he made 2 important mistakes:
1. in his first paper (and reiterated in his second) he didn't correctly account for the possibility of loudness-based discrimination of his square wave with vs. without partial ultrasonic filtering (used an outdated threshold of detection at 7 kHz)
2. in his second paper, where he introduced the source position difference experiment, he seems to have confused time alignment discrimination with frequency discrimination (but I still have to read that one more carefully to be fully satisfied I understand what he was purporting to do there).

I've also noticed that thread on Stereophile which you quickly turned into a flame war, failing to explain and quote which paper of his made which mistake. That one's not particularly helpful for readers coming to this late. :)
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,498
Location
Monument, CO
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.

Sending undesired ultrasonic images through the signal chain can have undesirable consequences ranging from extra energy to tweeters that turn into heat to intermodulation distortion that creates in-band distortion to causing an amplifier to go unstable.
 
Last edited:

Soniclife

Major Contributor
Forum Donor
Joined
Apr 13, 2017
Messages
4,500
Likes
5,417
Location
UK
Sending undesired ultrasonic images through the signal chain can have undesirable consequences ranging from extra energy to tweeters that turn it into heat to intermodulation distortion that creates in-band distortion to causing an amplifier to go unstable.
The made up FUD around class D seems much more relevant to filteless DACs. In most cases it probably resolves itself as stated, but without testing each system how would you know for sure.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,498
Location
Monument, CO
The made up FUD around class D seems much more relevant to filteless DACs. In most cases it probably resolves itself as stated, but without testing each system how would you know for sure.

Not sure what you mean about making up FUD about class D... I did not mention class D or any other amplifier class of operation. Confused...
 

Soniclife

Major Contributor
Forum Donor
Joined
Apr 13, 2017
Messages
4,500
Likes
5,417
Location
UK
Not sure what you mean about making up FUD about class D... I did not mention class D or any other amplifier class of operation. Confused...
I wasn't implying you were, but that some people think the switching frequency in class D is a problem, when in good class D it isn't, but nos has way more out of band noise, and much lower down.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
Hi,
What about IMD that could be triggered down the chain because of the transmission of ultrasonic content by the NOS DAC?
He was talking about the difference of NOS FILTERLESS and NOS filtered. Actually no matter with or without the filter it's already shit. The filter is not really important, unless sampling rate is lower than 32k. For example let's say sampling rate of 16khz, there will be imaging after 8khz which can be easily audible. But when sampling rate is 44.1 there is no way we can hear it. So it's irrelevant. The issue was NOS in the first place.
 

JohnYang1997

Master Contributor
Technical Expert
Audio Company
Joined
Dec 28, 2018
Messages
7,175
Likes
18,292
Location
China
Sending undesired ultrasonic images through the signal chain can have undesirable consequences ranging from extra energy to tweeters that turn into heat to intermodulation distortion that creates in-band distortion to causing an amplifier to go unstable.
Burn tweeter yes. Causing amplifier to be unstable that's the problem with the amplifier. Amplifier should be unconditionally stable.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,498
Location
Monument, CO
Burn tweeter yes. Causing amplifier to be unstable that's the problem with the amplifier. Amplifier should be unconditionally stable.

Should be, yes. But I, and I suspect @restorer-john, @solderdude, and many other current and former techs can provide a lengthy list of amps and times that was not the case.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Amplifier should be unconditionally stable.

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.
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'.

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)
 
Last edited:

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,498
Location
Monument, CO
One thing to beware with tube amps is that many do not like being unloaded...
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,588
Likes
38,288
Location
Gold Coast, Queensland, Australia
Burn tweeter yes. Causing amplifier to be unstable that's the problem with the amplifier. Amplifier should be unconditionally stable.

You aren't considering the era (1982) the compact disc format was released. It was also the era of DC-daylight amplification and ultra high slew rate, high powered, low/no NFB, power amplifiers. That is a recipe for disaster without well designed filtering.

The LPFs in the first compact disc players were extremely carefully designed, very expensive and manually aligned for each pole, there was absolutely no thought of filterless or otherwise faulty engineering. In fact, several brands of amplifiers offered twin sets of inputs (one HF filtered) in order to roll-off the HF when using CD players as they knew (inaudible) traces of the sampling frequency may be otherwise amplified/intermodulated and cook tweeters or cause HF instability.

The designers and incredibly skilled engineers of the day knew what they were doing, unlike the clowns pulling out LPFs and mucking about with NOS filterless D/A converters.
 
Last edited:

rwortman

Addicted to Fun and Learning
Forum Donor
Joined
Jan 29, 2019
Messages
740
Likes
683
Everything that's old is new again. I remember when CD players first arrived. The audio press hated them and complained about the sound. Then there was an arms race to see how fast you could oversample paralleled with development of D/S DAC's. Gradually technology advanced and most audio writers and audiophiles accepted 16/44 as at least good enough. Now audiophiles are falling in love with the same sound they hated 30+ years ago. High end audio seems to be all about competing fetishes. Romanticizing ancient audio history or searching for ever higher SINAD, everyone seems to have one.
 
Last edited:

orangejello

Active Member
Joined
Apr 18, 2019
Messages
232
Likes
354
Don't know about NOS dacs sounding "better", but the RME has a NOS filter which sounds the best to me out of the five that AKM 4490 provides. According to the RME manual this is the only filter that has essentially perfect impulse response.

A side effect of the NOS filter is that it slowly rolls off from 5k to 20k where it is down 3db when sampling at 44.1k. So it would be easy to attribute liking the sound of CDs through this filter to the roll off. However @MC_RME provided the parameters to use with the parametric equalizer to flatten the frequency response again. Using this combination of NOS and PEQ you get perfect impulse response and flat frequency response. My CD play back now sounds extremely nice to me.
 

Willem

Major Contributor
Joined
Jan 8, 2019
Messages
3,659
Likes
5,277
Yet RME did not make it the default... On the contrary.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Don't know about NOS dacs sounding "better", but the RME has a NOS filter which sounds the best to me out of the five that AKM 4490 provides. According to the RME manual this is the only filter that has essentially perfect impulse response.

A side effect of the NOS filter is that it slowly rolls off from 5k to 20k where it is down 3db when sampling at 44.1k. So it would be easy to attribute liking the sound of CDs through this filter to the roll off. However @MC_RME provided the parameters to use with the parametric equalizer to flatten the frequency response again. Using this combination of NOS and PEQ you get perfect impulse response and flat frequency response. My CD play back now sounds extremely nice to me.

Actually impulse response with music is worse in NOS mode. On top of that EQ only increases the average amplitude to 'flat' but the treble above 10kHz is very poor in quality with extremely poor timing characteristics.

Is RME wrong ? No... indeed squarewave response and needle pulses are produced (Note that I did not say reproduced) quite admirably.
Unfortunately these well looking test signals do not exist in music nor in any recording. These are just test signals that exist so an engineer can evaluate filter behavior. The guys thinking that because these almost unlimited bandwidth (they are limited to about 100kHz or so) the DAC performs admirably are mistaken.
True, one can use squarewaves to test amplifiers and a nice squarewave response says a lot about the performance.
But here too it is merely a test signal and NOT something found in music.
In DACs these squarewaves and needle pulses are not supposed to look perfect. They are supposed to be 'rounded' yet reach the intended output level. Squarewaves are supposed to 'ring' as well as that is a side effect of the bandwidth being properly limited (so it complies to Shannon-Nyquist).

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.
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.

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'
 
Last edited:

orangejello

Active Member
Joined
Apr 18, 2019
Messages
232
Likes
354
Yet RME did not make it the default... On the contrary.
They used a filter with flat frequency response for the default. Also, the data that they provide in the manual is for 44.1k sample rate. A lot of people stream at much higher rates, and this filter many not be the best at higher rates. Don't know, just guessing.
 

BDWoody

Chief Cat Herder
Moderator
Forum Donor
Joined
Jan 9, 2019
Messages
6,949
Likes
22,627
Location
Mid-Atlantic, USA. (Maryland)

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,386
Likes
24,752
Location
Alfred, NY
Everything that's old is new again. I remember when CD players first arrived. The audio press hated them and complained about the sound.

With the exception of Gordon Holt, who even titled his article, "Run Right Out." He liked the same aspects I did- pitch stability, neutral tonality, and low noise.
 
Top Bottom