Possible reasons why some people claim to hear a difference between similar DAC's

[righthookmike]

There was a debate recently with some saying they could here a difference in DAC's and others claiming it would not be possible if level matched. I believe in my system( I am only saying my system) I can hear a difference. I was genuinely curious how much we really understand about sound and started reading up. I came across some sound engineering references and wondered this. Could minute differences in the harmonics or overtones of any given frequency produced by said DAC's, well within the manufacturers tolerance range, have an effect on the timbre of said frequency in any individual speaker because of differences in materials etc. used, and enable some people to genuinely hear a difference in dac's in their system? Thus simplifying the debate to " those two DAC's are solid units. Try them in your system to see which one you prefer"

 

[RickSanchez]

There was a debate recently with some saying they could here a difference in DAC's and others claiming it would not be possible if level matched.

Well, more than a recent debate ...

Here's one place to start regarding DAC measurements, and at what point measurement differences are no longer audible:

• https://www.audiosciencereview.com/...-thresholds-of-amp-and-dac-measurements.5734/

So it is possible to hear differences in DACs, but typically what you're talking about in that case is the comparison between a solid-performing DAC and one that is broken or really poorly engineered. At a point all solid-performing DACs are audibly transparent and you wouldn't be able to distinguish between them.

If you do want to compare DACs via listening tests keep in mind there are two ingredients not one:

1. Level-matching between the devices. (Which you mention.)
2. Double-blind testing.

A sighted test on level-matched devices will still introduce listener bias that could lead to someone perceiving sound differences that don't exist.

As to your question regarding "minute differences in the harmonics or overtones of any given frequency" I'll let the other experts weigh in on that.

 

[righthookmike]

Than

Well, more than a recent debate ...

Here's one place to start regarding DAC measurements, and at what point measurement differences are no longer audible:

• https://www.audiosciencereview.com/...-thresholds-of-amp-and-dac-measurements.5734/

So it is possible to hear differences in DACs, but typically what you're talking about in that case is the comparison between a solid-performing DAC and one that is broken or really poorly engineered. At a point all solid-performing DACs are audibly transparent and you wouldn't be able to distinguish between them.

If you do want to compare DACs via listening tests keep in mind there are two ingredients not one:

1. Level-matching between the devices. (Which you mention.)
2. Double-blind testing.

A sighted test on level-matched devices will still introduce listener bias that could lead to someone perceiving sound differences that don't exist.

As to your question regarding "minute differences in the harmonics or overtones of any given frequency" I'll let the other experts weigh in on that.

.
thank you for the response. your last sentence is my real question,

 

[fieldcar]

Most people have no clue how sensitive they are to distortion. Take the klippel test, and you'll know. Then take your results and compare it to the SINAD/THD of your DAC. You'd be surprised how often it's just a volume mismatch when comparing two DAC's. Louder sounds better.

http://www.klippel.de/listeningtest/

 

[righthookmike]

Most people have no clue how sensitive they are to distortion. Take the klippel test, and you'll know. Then take your results and compare it to the SINAD/THD of your DAC. You'd be surprised how often it's just a volume mismatch when comparing two DAC's. Louder sounds better.

http://www.klippel.de/listeningtest/

I listen mostly at extremely loud levels. I can take a spl reading with an iphone of my normal listening levels. not sure how accurate it will be. Alright. I purposely didn't turn it up all the way. A slow blues song without a lot of heavy drums or lots of gain in the guitar was 70db-a . that's purposely below any song that I like volume level

 

[fieldcar]

I listen mostly at extremely loud levels. I can take a spl reading with an iphone of my normal listening levels. not sure how accurate it will be

I don't have anything to back it up, but I believe that you can discern distortion more easily at higher levels. Maybe someone else has something to add to this.

Also, in regards to harmonics, try playing with an additive synthesizer to get a grasp of what a 1st or 2nd harmonic sounds like. 1.000 = 100% and 0.005 = 0.5%. It kind of illustrates how crazy-clean a SINAD of 0.0005% is.

https://meettechniek.info/additional/additive-synthesis.html

 

[righthookmike]

I don't have anything to back it up, but I believe that you can discern distortion more easily at higher levels. Maybe someone else has something to add to this.

Also, in regards to harmonics, try playing with an additive synthesizer to get a grasp of what a 1st or 2nd harmonic sounds like. 1.000 = 100% and 0.005 = 0.5%. It kind of illustrates how crazy-clean a SINAD of 0.0005% is.

https://meettechniek.info/additional/additive-synthesis.html

cool, that's why I'm asking. I'm just beginning to learn about this stuff and I find it fascinating. Music is a part of my life , trying to understand More about sound and equipment takes it to a different level for me

 

[Blumlein 88]

I don't have anything to back it up, but I believe that you can discern distortion more easily at higher levels. Maybe someone else has something to add to this.

Also, in regards to harmonics, try playing with an additive synthesizer to get a grasp of what a 1st or 2nd harmonic sounds like. 1.000 = 100% and 0.005 = 0.5%. It kind of illustrates how crazy-clean a SINAD of 0.0005% is.

https://meettechniek.info/additional/additive-synthesis.html

Actually no. Your ear itself begins to distort a bit as you pass 75 db or so. While increased levels would raise harmonics making them more audible, your ear might distort enough that internal distortion masks any low level difference between two devices. All depending upon the particulars.

 

[fieldcar]

Actually no. Your ear itself begins to distort a bit as you pass 75 db or so. While increased levels would raise harmonics making them more audible, your ear might distort enough that internal distortion masks any low level difference between two devices. All depending upon the particulars.

Thanks! I should have trusted my gut and not posted it. I've noticed that when taking the klippel test, a bit more volume helps, but it's usually on the higher end of moderate volumes to where I can hear the most difference in distortion. I don't listen to anything very loud anymore, as I don't want to damage my hearing.

 

[solderdude]

Could minute differences in the harmonics or overtones of any given frequency produced by said DAC's, well within the manufacturers tolerance range, have an effect on the timbre of said frequency

Consider that DAC's are 'flat' within the audible range. Even for 44.1kHz. Depending on the used filter there can be a small ripple.
That ripple is much smaller than the 'adjustments' made in a studio for each individual track (instrument or voice) if they think it sounds better that way. We are not talking of adjustments of 0.1dB or so.

A DAC does not treat 'overtones' differently than fundamentals.
The purpose of a DAC is to 'translate' a digital 'value' to an analog voltage of the same (or extremely close) value as the 'described' value.

There was a debate recently with some saying they could here a difference in DAC's

When you are a young person then yes one can hear differences (also in blind tests) between certain DACs and even between the same DAC when specific filters are used. When DAC A has a slow filter and there is a roll-off in the upper treble of a few dB then this can be quite audible.
For older folks this may well not be the case.
This is not because DACs differ but because filters differ within the audible band.

Possible reasons why some people claim to hear a difference between similar DAC's

In most cases where the filters adhere to the sampling theorem and there are no high distortions (tubes for instance) the perceived differences are generally caused by incorrect level matching, too much time has passed between swapping out DACs or knowing what DAC is playing (sighted test).

Then there is the often mentioned 'well music is not the same as test tones' argument. And its not. When looking at transducers this certainly is the case. Not so with DAC's. They just put out a voltage equivalent to the value they should be producing and seamlessly 'slide' to the next sample value they must produce. There are ways to test with music called 'nulling'. @pkane wrote lovely software for this. The method has its challenges but it can show the actual differences between 2 DACs using music. Not surprisingly the effects (when well executed using a good ADC) are not unexpectedly different than those that can be shown to exist when using test tones.

 

[righthookmike]

Consider that DAC's are 'flat' within the audible range. Even for 44.1kHz. Depending on the used filter there can be a small ripple.
That ripple is much smaller than the 'adjustments' made in a studio for each individual track (instrument or voice) if they think it sounds better that way. We are not talking of adjustments of 0.1dB or so.

A DAC does not treat 'overtones' differently than fundamentals.
The purpose of a DAC is to 'translate' a digital 'value' to an analog voltage of the same (or extremely close) value as the 'described' value.


When you are a young person then yes one can hear differences (also in blind tests) between certain DACs and even between the same DAC when specific filters are used. When DAC A has a slow filter and there is a roll-off in the upper treble of a few dB then this can be quite audible.
For older folks this may well not be the case.
This is not because DACs differ but because filters differ within the audible band.



In most cases where the filters adhere to the sampling theorem and there are no high distortions (tubes for instance) the perceived differences are generally caused by incorrect level matching, too much time has passed between swapping out DACs or knowing what DAC is playing (sighted test).

Then there is the often mentioned 'well music is not the same as test tones' argument. And its not. When looking at transducers this certainly is the case. Not so with DAC's. They just put out a voltage equivalent to the value they should be producing and seamlessly 'slide' to the next sample value they must produce. There are ways to test with music called 'nulling'. @pkane wrote lovely software for this. The method has its challenges but it can show the actual differences between 2 DACs using music. Not surprisingly the effects (when well executed using a good ADC) are not unexpectedly different than those that can be shown to exist when using test tones.

I like your quote about listening to music. That's the real point of this for me. Then sometimes I get trapped by my ego into trying to argue my belief that there is so much more going on with music that we can't measure but by definition it's not provable.

 

[JoachimStrobel]

Well, given the fact that two speaker will not reproduce a soundfield and that hence any perception of the stereo field changes by very tiny differences and clues, I do believe that DACs with their separate treatment of the left and right signal path cause differences. Those will not be genuine but are artifacts of the stereo reproduction technique. It might be interesting to check if that happens to mono recordings too.