Investigation Into Effects Of PC load On DAC Analogue Output

[botrytis]

12 hours ago, idiot_savant said:

@PeterSt - you say you’ve spent many years trying to optimise DACs and PCs/ software, which I believe. So, I’ve never seen you share these improvements/optimisations, apart from that you sell some kind of bespoke system? 
How annoying would you find it if a competitor came on here and just kept say “haha - I know this” repeatedly? 
I understand if you believe you’ve discovered some secret sauce, but just saying you have a secret sauce doesn’t make it fingerlicking good?

 

your friendly neighbourhood idiot 

 

He is sounding like fas42, believing his own ideas w/o ever being able to prove it to others, with objective measurements.

[botrytis]

10 hours ago, manueljenkin said:

Just because it's not measurable with present state of apparatus doesn't mean it doesn't exist. Has happened in science many times, that a top level abstraction is shattered with a new discovery bringing life to older approaches that were earlier ridiculed.

 

I don't think anyone here has demonstrated lack of knowledge. You seem to be adamant that there is a magical fix down the line that can remove "all" aberrations introduced by noise from pc ground planes while the only measurements you have are mostly uncorrelated to real world audio signal performance, and some of the issues beyond that has already been demonstrated by John Swenson and others. 

 

Measurements are great, but conclusion is not. Objectivity is different from science. If required to be objective one can rank cows by their aerodynamic structure, but would it make any sense?

 

 

Two different arguments, but in the case of hearing, measurements are usually more sensitive than hearing, especially while using a proper DBT. Any other type of listening suffers from bias, period. That is how are brains were designed.

[botrytis]

3 hours ago, manueljenkin said:

Oh yeah, trying to derive conclusions from a paper about jitter effects from 1974 when the whole domain of phase noise analysis methods was described to fair accuracy only after 2000s.

 

I do this for a living. Basically, much new research is really based on older research, if you look hard enough. Saying that a 1974 is not valid, is a red herring and utterly arrogant. Researchers had to design experiments better because it was HARDER to prove their points. Many of them also describe the the results better because people now a days like to use technobabble as it makes them seem smarter. A great example is MQA with 'deblurring' etc.

[jabbr]

17 hours ago, March Audio said:

Please ask questions, it's a complex subject.  I'm sure between myself, @pkane2001 and @idiot_savant we can make it more understandable.

 

Indeed, controlled subjective listening tests are something I intend to move on to.  Will take a bit of time and organising however.

To be clearer, I don’t have any issue with your measurements. You’ve clearly stated that they apply to your specific tested equipment.
 

The arguments that I find incomprehensible are those which argue that you are wrong or haven’t tested appropriately. This argument is old enough that anyone who wants the “prove” that x, y or z causes and audible difference, need themselves to present a clearly understandable graph/measurements. The measurement techniques are all very well described. The physics of electronics is all very well known. That said there aren’t comprehensive measurements so who knows what is real. It’s trivial to show that two cables measure differently but there are few demonstrations that this causes a difference at the DAC output. Obviously speaker cables cause electrical differences at the speaker inputs. 

 

That said taking a condescending tone doesn’t do anything to help whatever cause you may have. The archives here contain any objections I have to measurement assumptions, namely I don’t want to bandwidth limit the output signal in any fashion to allow all non linearities to present themselves but I’m not going to repeat those arguments here in detail, the point stands that if anyone objects to your findings they are free to repeat your measurements and use different equipment and/or better techniques as they see fit. This isn’t a dead horse that I care to make into glue.

[botrytis]

Well, if I can use USB to collect all data from a MS, that will collect AMU (Atomic Mass Units) from 10 to 1000 AMU 60 times a second for 2 hours w/o any issues and also control the MS and the GC feeding it, I think USB can handle music. 

[kumakuma]

4 minutes ago, botrytis said:

Well, if I can use USB to collect all data from a MS, that will collect AMU (Atomic Mass Units) from 10 to 1000 AMU 60 times a second for 2 hours w/o any issues and also control the MS and the GC feeding it, I think USB can handle music. 

 

I wonder if there's a market for jitter-free GC columns...

[botrytis]

1 minute ago, kumakuma said:

 

I wonder if there's a market for jitter-free GC columns...

 

Well, just saying that is a large amount of data and very small changes in the signal matter.

 

Well, PLOT columns are made of fused silica - while they will flex some, the are also prone to cracking. They are expensive too.

[idiot_savant]

@botrytis- you fool, you’ve forgotten the golden rule - audio has ethereal properties that can only be captured with random noise and distortion. I’m pretty sure if you chanted “ommm” while capturing, your results would be, er, better, or faster, or more “close to the original”. 
 

Your friendly neighbourhood idiot 

[idiot_savant]

Probably make you cry as well

 

 

 

 

 

your friendly neighbourhood idiot 

[jabbr]

3 hours ago, botrytis said:

Well, if I can use USB to collect all data from a MS, that will collect AMU (Atomic Mass Units) from 10 to 1000 AMU 60 times a second for 2 hours w/o any issues and also control the MS and the GC feeding it, I think USB can handle music. 

I know you guys are joking but this is the same issue as using logic to decide what you should hear.

 

The simple objective fact is that the USB signal presenting at the receiver need not be close to a true square wave, and the digital systems tolerate quite a bit of input ringing. Its certainly conceivable that certain DACs do not adequately filter out the ringing and so small changes such as rise time etc might have an effect that makes it to the DAC output -- the DAC really needs to be treated as an analog device at some point. **might** make a change. It would need to be measured. All the rest is speculation and assumptions until the DAC output is measured by someone who cares ... and not just 20-20kHz ... full band, I mean a 1 MHz or GHz oscillation might make a difference ... I had that happen once with an unstable I-V converter that was missing a gate stopper... sounded like ass. The only problem was that the factory which made the part changed ... little stuff can make a difference unless thorough measurements.

[idiot_savant]

I do apologise if I appear flippant, but I think the point remains - audio is incredibly important to many people, for very good reasons - but how is it harder than other things? A picoscope, for sake of argument captures data at precise moments in time. And transfers it’s results via USB but I don’t recall anybody having to do fancy things to have it work properly?

 

 

your friendly neighbourhood idiot 

[March Audio]

5 hours ago, jabbr said:

To be clearer, I don’t have any issue with your measurements. You’ve clearly stated that they apply to your specific tested equipment.
 

The arguments that I find incomprehensible are those which argue that you are wrong or haven’t tested appropriately. This argument is old enough that anyone who wants the “prove” that x, y or z causes and audible difference, need themselves to present a clearly understandable graph/measurements. The measurement techniques are all very well described. The physics of electronics is all very well known. That said there aren’t comprehensive measurements so who knows what is real. It’s trivial to show that two cables measure differently but there are few demonstrations that this causes a difference at the DAC output. Obviously speaker cables cause electrical differences at the speaker inputs. 

 

That said taking a condescending tone doesn’t do anything to help whatever cause you may have. The archives here contain any objections I have to measurement assumptions, namely I don’t want to bandwidth limit the output signal in any fashion to allow all non linearities to present themselves but I’m not going to repeat those arguments here in detail, the point stands that if anyone objects to your findings they are free to repeat your measurements and use different equipment and/or better techniques as they see fit. This isn’t a dead horse that I care to make into glue.

I don't havexany cause.  This is an investigation.  Will it test every combination of dac / pc on the planet?  No.  It will however start to lay some facts down about what really happens instead of the unsupported claims, hand waving and inaccuracies seen from some people.

[March Audio]

1 hour ago, jabbr said:

I know you guys are joking but this is the same issue as using logic to decide what you should hear.

 

The simple objective fact is that the USB signal presenting at the receiver need not be close to a true square wave, and the digital systems tolerate quite a bit of input ringing. Its certainly conceivable that certain DACs do not adequately filter out the ringing and so small changes such as rise time etc might have an effect that makes it to the DAC output -- the DAC really needs to be treated as an analog device at some point. **might** make a change. It would need to be measured. All the rest is speculation and assumptions until the DAC output is measured by someone who cares ... and not just 20-20kHz ... full band, I mean a 1 MHz or GHz oscillation might make a difference ... I had that happen once with an unstable I-V converter that was missing a gate stopper... sounded like ass. The only problem was that the factory which made the part changed ... little stuff can make a difference unless thorough measurements.

No they aren't joking.

 

Why do people think that audio is some how special?  That it evades the laws of physics and scientific investigation?

 

Actually this isn't based on logic about what you can hear.  It's based on probably 100 years of scientific research of Psychoacoustics.  I mentioned above some of the research into the audibility of jitter for example.

 

What we see from some people is just Han waving, "oh we might be missing something", when they haven't even conducted controlled listening tests to remove their own biases.

[March Audio]

1 hour ago, jabbr said:

. All the rest is speculation and assumptions until the DAC output is measured by someone who cares ... and not just 20-20kHz ... full band, I mean a 1 MHz or GHz oscillation might make a difference ... .

For there to be audible differences the end effect needs to be in band, ie below 20kHz.

 

To be clear about what I'm saying, that doesn't preclude out of band issues causing in band ones, but the point is the problem will be seen if it manifests in band.

[March Audio]

1 hour ago, jabbr said:

 

The simple objective fact is that the USB signal presenting at the receiver need not be close to a true square wave, and the digital systems tolerate quite a bit of input ringing. Its certainly conceivable that certain DACs do not adequately filter out the ringing and so small changes such as rise time etc might have an effect that makes it to the DAC output -- the DAC really needs to be treated as an analog device at some point. **might** make a change. It would need to be measured. All the rest is speculation and assumptions 

OK, so how might that manifest?

[March Audio]

Considering that the data from the usb receiver is subsequently buffered and clocked out by a separate low phase noise clock.

[jabbr]

1 hour ago, March Audio said:

No they aren't joking.

 

Why do people think that audio is some how special?  That it evades the laws of physics and scientific investigation?

 

I never suggested audio is different nor that it could evade the laws of physics and scientific investigation.

 

There is no law in physics that says that a USB cable can never alter the SQ of a particular DAC, nor is there a law of physics that prohibits server load from influencing the DAC.

 

Electrical physics is very well known and accepted. Do Maxwell's Equations say anything about cable or server load audibility? Unless they've changed since I was in school: zilch.

 

1 hour ago, March Audio said:

 

Actually this isn't based on logic about what you can hear.  It's based on probably 100 years of scientific research of Psychoacoustics.  I mentioned above some of the research into the audibility of jitter for example.

 

What we see from some people is just Han waving, "oh we might be missing something", when they haven't even conducted controlled listening tests to remove their own biases.

 

On the other hand you are completely misstating what the "Laws of Physics" actually say on the topic: zero.

 

Seriously the so called "scientific research" on psychoacoustics isn't really there either. Believe me.

 

That said there are a lot of implausible and fanciful arguments that have zero scientific support. As far as I am concerned I am sufficiently certain that my fiberoptic Ethernet electrically isolates my server from my DAC that I am not concerned with this topic. As far as I am concerned people should use good solid electrical design and the illusion that bits are just bits works for our everyday purposes.

[jabbr]

1 hour ago, March Audio said:

OK, so how might that manifest?

How might it? Common mode noise *might* sail right through the digital section of the DAC and affect the analog output. It really might. Im not saying that it does or does for every design but it might.

 

To be clear assuming that something **can't** happen just because you can't imagine how it would is not science. To disprove my above hypothesis for a particular DAC, you would need to apply common mode noise of varying intensities to the inputs and measure the outputs. I would want the CMN to extend into the Ghz range because, well, it does.

[March Audio]

1 minute ago, jabbr said:

 

I never suggested audio is different nor that it could evade the laws of physics and scientific investigation.

 

There is no law in physics that says that a USB cable can never alter the SQ of a particular DAC, nor is there a law of physics that prohibits server load from influencing the DAC.

 

Electrical physics is very well known and accepted. Do Maxwell's Equations say anything about cable or server load audibility? Unless they've changed since I was in school: zilch.

 

 

On the other hand you are completely misstating what the "Laws of Physics" actually say on the topic: zero.

 

Seriously the so called "scientific research" on psychoacoustics isn't really there either. Believe me.

 

That said there are a lot of implausible and fanciful arguments that have zero scientific support. As far as I am concerned I am sufficiently certain that my fiberoptic Ethernet electrically isolates my server from my DAC that I am not concerned with this topic. As far as I am concerned people should use good solid electrical design and the illusion that bits are just bits works for our everyday purposes.

I didnt suggest you did :)

 

Oh for sure, but we have suggestions from some that the manifestation of these alleged effects are beyond measurement.  Its somehow magic that cant been seen.

 

Sorry I dont mean to be rude but thats rubbish regarding psychoacoutic research.  How do you think compressed formats were developed?  Audibility of kitter has been researched, etc, etc etc.

[jabbr]

Just now, March Audio said:

I didnt suggest you did :)

 

Oh for sure, but we have suggestions from some that the manifestation of these alleged effects are beyond measurement.  Its somehow magic that cant been seen.

 

Everything is measurable with sufficient motivation even gravitons and mesons.

 

Just now, March Audio said:

 

Sorry I dont mean to be rude but thats rubbish regarding psychoacoutic research.  How do you think compressed formats were developed?  Audibility of kitter has been researched, etc, etc etc.

 

Let's just say that my faith in the conclusiveness of this research does not rise to my faith in Maxwell's Equations ;) Compressed formats do not take into the nonlinearity of the human brain i.e. cable color does actually affect SQ because the visual system affects the auditory system at the cortical level. I'll stick to electrical output of the DAC thank you.

[March Audio]

3 minutes ago, jabbr said:

How might it? Common mode noise *might* sail right through the digital section of the DAC and affect the analog output. It really might. Im not saying that it does or does for every design but it might.

 

To be clear assuming that something **can't** happen just because you can't imagine how it would is not science. To disprove my above hypothesis for a particular DAC, you would need to apply common mode noise of varying intensities to the inputs and measure the outputs.

well measure single ended ;)  Thats actually one of the tests I will be dong as the Gustard has XLR and single ended outputs.

 

Again its just more nebulous hand waving.  Lets be clear about this, these are effects that some audiophiles claim they hear.  Yet there is no controlled subjective testing that confirms these things are audible.  "Chasing Ghosts" springs to mind.

[jabbr]

2 minutes ago, March Audio said:

well measure single ended ;)  Thats actually one of the tests I will be dong as the Gustard has XLR and single ended outputs.

 

Again its just more nebulous hand waving.  Lets be clear about this, these are effects that some audiophiles claim they hear.  Yet there is no controlled subjective testing that confirms these things are audible.  "Chasing Ghosts" springs to mind.

Single ended ... errmmm common mode is common to the signal and ground ... seriously (it is very tricky to deal with)

 

If you think that common mode noise is just nebulous hand waving you need to go back to school -- of course this is an example, a plausible example nonetheless. But man its just freshman electronics...

[March Audio]

1 minute ago, jabbr said:

 

Everything is measurable with sufficient motivation even gravitons and mesons.

 

 

Let's just say that my faith in the conclusiveness of this research does not rise to my faith in Maxwell's Equations ;) Compressed formats do not take into the nonlinearity of the human brain i.e. cable color does actually affect SQ because the visual system affects the auditory system at the cortical level. I'll stick to electrical output of the DAC thank you.

Sorry you have lost me.  You may have a misunderstanding of what the science of psychoacoustics entails.

 

Well MP3 and AAC work.  Built on the study of how we hear things.

 

We are looking at the electrical output of the DAC.  

 

 

[March Audio]

1 minute ago, jabbr said:

Single ended ... errmmm common mode is common to the signal and ground ... seriously (it is very tricky to deal with)

you will lose the common mode rejection hence it becomes more visible.

[jabbr]

1 minute ago, March Audio said:

you will lose the common mode rejection hence it becomes more visible.

idk balanced does differential mode noise rejection. fiber rejects common mode noise. again its tricky.