The person who mentioned it is well known and pretty accomplished industry consultant. Maybe he was trying to see if he could help a client mitigate a problem, don't know.
According to the survey, the largest group of respondents think that there can be an audible difference between DACs that measure similarly. Reading between the lines a little, there seems to be a view that different chipsets exacerbate this difference.
There are reports that changing certain components can also influence the sound.
It is also abundantly clear the everything that influences the sound we hear is contained in the analogue signal emitted by the DAC.
So when we measure DACs in many ways we can get results that suggest that any differences should be well below the threshold of audibility. Psychoacoustics can of course play an insidious role here, but may not be the only factor.
Digital devices operate in such a way that the digital signals have a very high RF content. Designers are at great pains to keep these signals out of the audio circuitry.
A few messages above this there was an observation that on occasion a changed power lead caused a difference in the sound. It seems very far fetched that on ohm or two in the cable would do this especially when regulators are used. It seems perhaps a little more likely that other factors are involved such as EMI and that some cables emit or receive less EMI.
Similarly opamp rolling in DACs may produce a difference due to the susceptibility to the fast rise times on the input to the filter stage.
So why don’t these show up in measurements? Possibly because the measuring gear has very good RF filtering and removes the offending signals.
If good DACs really do sound different there must be a reason, and it may be that the system as a whole has a part to play.
I just don’t know and until/unless I get another fairly good DAC, I probably won’t find out if there is an audible difference in my system.
Hopefully some food for thought, and not too contentious I hope.
Anyway, very late for me here, so goodnight for now!
There are reports that changing certain components can also influence the sound.
It is also abundantly clear the everything that influences the sound we hear is contained in the analogue signal emitted by the DAC.
So when we measure DACs in many ways we can get results that suggest that any differences should be well below the threshold of audibility. Psychoacoustics can of course play an insidious role here, but may not be the only factor.
Digital devices operate in such a way that the digital signals have a very high RF content. Designers are at great pains to keep these signals out of the audio circuitry.
A few messages above this there was an observation that on occasion a changed power lead caused a difference in the sound. It seems very far fetched that on ohm or two in the cable would do this especially when regulators are used. It seems perhaps a little more likely that other factors are involved such as EMI and that some cables emit or receive less EMI.
Similarly opamp rolling in DACs may produce a difference due to the susceptibility to the fast rise times on the input to the filter stage.
So why don’t these show up in measurements? Possibly because the measuring gear has very good RF filtering and removes the offending signals.
If good DACs really do sound different there must be a reason, and it may be that the system as a whole has a part to play.
I just don’t know and until/unless I get another fairly good DAC, I probably won’t find out if there is an audible difference in my system.
Hopefully some food for thought, and not too contentious I hope.
Anyway, very late for me here, so goodnight for now!
@Jeffh01
Please tell us how much do you know about sigma-delta modulators and the somewhat bizarre signal correlated noise/distortion they can produce?
Do you know what constitutes a PSS measurement (Periodic Steady State)?
Also, do you understand how an FFT is just like a DFT in that it measures averaged correlation of signal and noise artifacts with bin frequencies?
If you do understand all those things well enough, then you should be able to understand that PSS distortion and noise (such as are measured at ASR) are not the only potentially audible effects sigma delta dac can produce.
Regarding what designers go to great pains to do, tell us about it. We have some dac designers here.
Please tell us how much do you know about sigma-delta modulators and the somewhat bizarre signal correlated noise/distortion they can produce?
Do you know what constitutes a PSS measurement (Periodic Steady State)?
Also, do you understand how an FFT is just like a DFT in that it measures averaged correlation of signal and noise artifacts with bin frequencies?
If you do understand all those things well enough, then you should be able to understand that PSS distortion and noise (such as are measured at ASR) are not the only potentially audible effects sigma delta dac can produce.
Regarding what designers go to great pains to do, tell us about it. We have some dac designers here.
Last edited:
Strictly speaking that isn't correct. The electrical signal tells you about the vibrations which then get interpreted by the ear/brain to give sound. The McGurk effect demonstrates this - the sound changes according to whether or not you watch the video but the vibrations remain the same.
Wondering if we are getting into an area of semantics? A dac outputs an electrical signal (or two of them if a stereo dac), as we all know. The ultimate sound depends on the rest of the system, especially the speakers and room, and perhaps even more so on the ear and brain.
In the context of a discussion on dacs, it seems reasonable to consider a dac's electrical output as all it can be held responsible for in relation to the final perception of sound.
Yet it seems like in some post or posts the dac is being held to an electrical output standard, along with a final perceived sound standard? Perhaps that would be because the electrical output standard was based only on standard measurements?
In the context of a discussion on dacs, it seems reasonable to consider a dac's electrical output as all it can be held responsible for in relation to the final perception of sound.
Yet it seems like in some post or posts the dac is being held to an electrical output standard, along with a final perceived sound standard? Perhaps that would be because the electrical output standard was based only on standard measurements?
Last edited:
If that's the intended meaning of 'the electrical signal tells you everything about the sound' then I misunderstood.
Hi @Markw4,
Only what I have read here on on the links that folks here have provided. I do often read the references provided.
I am still trying to understand more of this.
If anyone is looking to me as an authority I’m afraid I’ll have to disappoint 😟.
My contribution was aimed at trying to elicit further information in the light of some recent comments.
I see in the discussion here that there are some aspects that there is not agreement on, and I don’t see how the differing views can be reconciled as a considerable effort in time and equipment would be needed to do so.
One area I find interesting with sigma delta DACs is the noise shaping. The addition of dither to push the audio noise floor down and push the noise output to much higher frequencies that are then filtered out is clever. I don’t know if this is an area where there is scope to vary exactly how this is done and that it is this aspect that causes the different noise sidebands that you describe.
I will admit to being a bit rusty on FFT vs DFT.
On PCB design, from many posts here from the very smart folks who do design DACs and related boards, I gather that it isn’t exactly a cakewalk to get it right. I have only begun to dabble with Kicad and have not had a board manufactured at this time.
When I do I will reach out to folks here to help fix my inevitable blunders!
I continue to find all this endlessly fascinating, whether speaker design, amplifier design or the inner working of DACs and look forward to learning more from the members here.
Thank you for responding to my post, you have given me some food for thought and some areas I can look into further to increase my understanding of this complex subject.
I think we can all agree on that the sinad value is at least a figure to weed out the bad seeds. Who would like to buy a DAC that measure crap at this?
Low distortion is a sane indication on that engineering was proper. Low noise is not so trivial to achieve so thats also a good indicator that basics are in place.
Then if you are a thinking person you investigate some more before your purchase decision. Audition could be one such investigation - and you are allowed to listen for whatever you like ;-)
//
Low distortion is a sane indication on that engineering was proper. Low noise is not so trivial to achieve so thats also a good indicator that basics are in place.
Then if you are a thinking person you investigate some more before your purchase decision. Audition could be one such investigation - and you are allowed to listen for whatever you like ;-)
//
Maybe the thinking instead of listening does not deliver.
Used to read, read and read up on stuff but after letting that loose I found gambling also works OK with current generation of DACs. Apparently most (but not all) are over the threshold of what I can hear to be mediocre/bad/unpleasant. That sound is a ticked box but now annoyance comes from a missing pixel in the OLED screen. What has me thinking is a top measuring device that I switch off after 15 minutes. I sure had a few of those.
There seems a clear line between autodidact people that have no predefined limits and the people that have been educated in the field (or another field, a special group) and want to explain/solve everything their instruments see and dismiss stuff the previous group simply just hears/accepts without too much thinking.
A simple example is breaking in of electronics. Many of the educated group say it is nonsense. The other group says it is like that. I also think it is a phenomenon but what if I am fooling myself? So I had 2 new identical devices and had 1 of them under power for 24 hours and the other one stayed in the box. A sceptic was also there. We all heard a difference that was gone after a few hours. Explanation? No, let's have a drink.
Used to read, read and read up on stuff but after letting that loose I found gambling also works OK with current generation of DACs. Apparently most (but not all) are over the threshold of what I can hear to be mediocre/bad/unpleasant. That sound is a ticked box but now annoyance comes from a missing pixel in the OLED screen. What has me thinking is a top measuring device that I switch off after 15 minutes. I sure had a few of those.
There seems a clear line between autodidact people that have no predefined limits and the people that have been educated in the field (or another field, a special group) and want to explain/solve everything their instruments see and dismiss stuff the previous group simply just hears/accepts without too much thinking.
A simple example is breaking in of electronics. Many of the educated group say it is nonsense. The other group says it is like that. I also think it is a phenomenon but what if I am fooling myself? So I had 2 new identical devices and had 1 of them under power for 24 hours and the other one stayed in the box. A sceptic was also there. We all heard a difference that was gone after a few hours. Explanation? No, let's have a drink.
Last edited:
That’s the way I have been thinking also and then check out of reviews like this to investigate further:
https://headfonics.com/smsl-d-6s-review/
https://headfonics.com/topping-d90-iii-sabre-review/
Attachments
I don't consider ear/brain part of the sound (i.e. audible acoustic wave). And if you listen "blinded" to sounds produced from 2 identical electrical signals in the same enviroment (amplifier, transducers, room, background noise, ...) your ear/brain should process them quite similarily.
Why would anyone on a DIY forum want to buy a DAC when you can also build one yourself?
Regarding SINAD, my valve DAC has a much higher noise floor than most modern DACs have, but so far, everyone who has listened to it seems to like it. As it was in sighted listening, that could simply be due to its impressive size, but it could also be because I took measures against intersample overshoots, pre-echoes and the low-level artefacts sigma-delta DACs often have.
Regarding SINAD, my valve DAC has a much higher noise floor than most modern DACs have, but so far, everyone who has listened to it seems to like it. As it was in sighted listening, that could simply be due to its impressive size, but it could also be because I took measures against intersample overshoots, pre-echoes and the low-level artefacts sigma-delta DACs often have.
The topic of this thread mentions "measure well" which is not defined. But since some seem to think this thread was inspired by ASR then the definition of "measure well" could also be taken from ASR. AFAIK the SINAD limit of "excellent" DACs at ASR is 110dB. So if somebody is going to organize a test the devices to be tested should also be measured to fit the "measure well" criteria.
SMD, difficulty for inexperienced builders, difficulty in fault finding in digital circuits and the fear of ending up with a $300 useless brick are my concerns here! 😀
But I am very slowly starting to come round to the idea.
Does anyone have a recommendation for a complete kit?
In Astronomy where is one of my interests, they use the science of Physics. There, a theory, in order to be accepted (until the next theory) has
to be describe the world according to our PERCEPTION and the CORRECT UNDERSTANDING of data from our current technology instrunments.
Then, these results, are made as a paper from a team of experts and they are subjected to other teams for verification. After years of evaluation
it is accepted or rejected.
Forgive my ignorance, but when people claim "science" they have to follow the protocols established around the world in each relative field.
So my questions are:
1) has this theory been transformed to a paper?
2) has it been verified and accepted by other scientific groups like AES etc?
3) When someone claims something, usually they have al the relevant data and they can answer all questions easily.
for example i will repeat my question
"what is the difference in % of distortion for a human being (which human being really) to undestand 1 perceived level (the smallest) of difference?"
0,5%? 0,1% 0,01% 0,001% 0,0001%?
because if i recall correctly there are programs that can add distortion to a signal. this way we would answer this question very fast correct?
we can easily draw a curve of results if 100 (or 1000) people listen to such files. This alone should be sufficient for the THD question. But if we make the unexpected discovery that 80% of people can't distinguish for example 0,1% and 100% can't distinguish 0,02% then we do have a serious problem right?
We do know that a device is A SYSTEM OF COMPONENTS and these components have properties.
4) what happens with all the systems around the worrld that have distortion from 5-20%? shoudlnt they be completely unlistenable to?
what is the SINAD of lets say 7% or 15%?
Is this level of SINAD easily distinguishable (ACCORDING TO QUESTION NO3) from another system (dac, pre-power) SINAD of lets say 60 or 80?
Because if for example we accept a level of 0,01% as something that can be perceived as "worse" (with less perceived transparent) then a system with 10% thd vs a system with 0,01% total thd should have 1000 levels "worse" sound. And this can't be argued.
If we cant say its "worse" and we say its "different" i'm sorry this is not science, because science must be exact. Of course we DONT know all,
we are still advancing, but we cant also draw such a conclusion, nor make a living out of it (i read something about money from companies in a post?) or persuade others about it as the one and absolute truth.
So how does this theory explains this stuff? because so far i have heard only one sentence and it's my first time of listening to a theory with a single sentence.
For example this system below should be totally unlistenable to right? (5-10% from a tonearm + an OTL, lets pretend there is no pre -🙂
5) Is the test system transparent enough to show differences? It is another thing to say "i cant hear any difference in my system" and "there is no difference between these two devices". science needs to be exact.
6) Has the author of the theory answered all arguments against this theory with a dialogue with other scientists? because in other scientific fields this is very important! If he can't answer all, then the theory isnt complete but stays under examination until everything is clear.
7) I haven't seen any serious arguments against post no61 or another post that someone said he was changing op amps, all with very low THD and he could hear differences. i think this is also a reality isnt? Can we have some scientific (definite) answers to these posts also please?
Please forgive my ignorance, but i need specific answers that can be repeated to all devices and the majority of humans and this is what i perceive as science.
People today say a lot of things to please their ego, or to follow someone because they are insecure or to make money.
but these behaviours have nothing to do with science nor they can be called this way.
where do we stand?
to be describe the world according to our PERCEPTION and the CORRECT UNDERSTANDING of data from our current technology instrunments.
Then, these results, are made as a paper from a team of experts and they are subjected to other teams for verification. After years of evaluation
it is accepted or rejected.
Forgive my ignorance, but when people claim "science" they have to follow the protocols established around the world in each relative field.
So my questions are:
1) has this theory been transformed to a paper?
2) has it been verified and accepted by other scientific groups like AES etc?
3) When someone claims something, usually they have al the relevant data and they can answer all questions easily.
for example i will repeat my question
"what is the difference in % of distortion for a human being (which human being really) to undestand 1 perceived level (the smallest) of difference?"
0,5%? 0,1% 0,01% 0,001% 0,0001%?
because if i recall correctly there are programs that can add distortion to a signal. this way we would answer this question very fast correct?
we can easily draw a curve of results if 100 (or 1000) people listen to such files. This alone should be sufficient for the THD question. But if we make the unexpected discovery that 80% of people can't distinguish for example 0,1% and 100% can't distinguish 0,02% then we do have a serious problem right?
We do know that a device is A SYSTEM OF COMPONENTS and these components have properties.
4) what happens with all the systems around the worrld that have distortion from 5-20%? shoudlnt they be completely unlistenable to?
what is the SINAD of lets say 7% or 15%?
Is this level of SINAD easily distinguishable (ACCORDING TO QUESTION NO3) from another system (dac, pre-power) SINAD of lets say 60 or 80?
Because if for example we accept a level of 0,01% as something that can be perceived as "worse" (with less perceived transparent) then a system with 10% thd vs a system with 0,01% total thd should have 1000 levels "worse" sound. And this can't be argued.
If we cant say its "worse" and we say its "different" i'm sorry this is not science, because science must be exact. Of course we DONT know all,
we are still advancing, but we cant also draw such a conclusion, nor make a living out of it (i read something about money from companies in a post?) or persuade others about it as the one and absolute truth.
So how does this theory explains this stuff? because so far i have heard only one sentence and it's my first time of listening to a theory with a single sentence.
For example this system below should be totally unlistenable to right? (5-10% from a tonearm + an OTL, lets pretend there is no pre -🙂
5) Is the test system transparent enough to show differences? It is another thing to say "i cant hear any difference in my system" and "there is no difference between these two devices". science needs to be exact.
6) Has the author of the theory answered all arguments against this theory with a dialogue with other scientists? because in other scientific fields this is very important! If he can't answer all, then the theory isnt complete but stays under examination until everything is clear.
7) I haven't seen any serious arguments against post no61 or another post that someone said he was changing op amps, all with very low THD and he could hear differences. i think this is also a reality isnt? Can we have some scientific (definite) answers to these posts also please?
Please forgive my ignorance, but i need specific answers that can be repeated to all devices and the majority of humans and this is what i perceive as science.
People today say a lot of things to please their ego, or to follow someone because they are insecure or to make money.
but these behaviours have nothing to do with science nor they can be called this way.
where do we stand?
My point is the following: if everything is in the electrical signal, then there must also be the component that tells us how the device sounds.
However, at present we don't know how to recognize that component and how to extract it from that signal.
So those (certainly not you) who act arrogant would do better to adopt a more intellectually humble attitude, in my view.
Maybe because the price/quality ratio of DIY DACs simply has become very unfavorable today? Also that ChiFi developments and also DAC technology have matured so that DIY makes less and less sense as the average DIY project will not perform likewise? So one will need 10 x the effort and 4 times the price to even have something usable.
I am sure I could have built a better house than the well insulated but old one I live in but with todays material prices I would not gain much (and loose a lot). No one will think I am crazy when I say that. When I needed a better double glazing wall (very large window used as wall) in 2018 I needed to fork out 6500 which I needed to think about. Installation was DIY. Todays price is 14000.
When I started building DACs they were always cheaper than ready made ones. Today the casing already costs more than some affordable but excellent DACs. I paid 100 Euro of the one I currently use. It is very good, I can not build anything equal for even double that price. Plus the risk of not being a successful project, needed revisions, difficulties in even finding the parts, unsolderability of some parts, what ifs, could be's etc. Despite loving electronics it just makes not much sense certainly when the heard difference is non existent. That probably explains ASRs popularity partly too. Pick one of the list that measures well, try out another one and repeat and still the sum is lower than a DIY DAC costs.
That is why I repeatedly suggested to have a group effort in developing a diyaudio.com DAC that does perform better than ChiFi otherwise this will stay a sea of words and opinions. I know it can be done as we already did such.
Last edited:
We don’t have the funding to explore this matter in a more scientific way or statistical approach. So I am afraid we are at a standstill 😕
- Status
- Not open for further replies.