Jump to content
IGNORED

Cleaning up USB signal for DAC


Recommended Posts

However, I haven't seen citations to academic work and testing directly relating levels of THD and IMD caused by jitter with audibility, so I can't tell you what the generally accepted view is regarding the audible impact of such jitter with current typical async USB DACs, nor even whether such a generally accepted view exists.

 

Jud, I agree with you. There is lack of clear evidence what makes the audible difference. Some possible scenarios were given with believe that they can explain the audible difference, but we have trouble to confirm them with measurements.

 

Look here what J. Atkinson wrote: AudioQuest JitterBug USB noise filter | Stereophile.com

You can see from the "Measurements" sidebar that I could find no significant effect that the JitterBug had on the analog signals output by three of the DACs I had to hand. Yet with those DACs and others, I heard an improvement in sound quality that I can attribute only to the JitterBug. I hate when that happens!

 

And how J. Swenson commented the measurement problem:

UpTone Audio USB Regen Manufacturers' Comment | Stereophile.com

Measuring the signal integrity of the USB signal is something fairly easy to do with a 3–4GHz scope and an eye-pattern test. Jitter measurements are a lot harder; at the exceedingly low level we are looking at, it takes some rather expensive test equipment. Unfortunately, where a lot of the final jitter happens is inside the DAC chip, and that is essentially impossible to measure.

...

None of the standard audio tests have been able to find any difference in the analog audio output. Part of the problem here is that in order to make any sense, the analog-to-digital converter used in the test equipment must have lower jitter than the jitter you are trying to measure the effects of. People doing these measurements are using ADCs that have known jitter levels significantly greater than what the best DACs have. This makes these tests pretty useless for measuring the supposed effects caused by changes in jitter.

...

The theory I am going on for these effects is that the operation of the USB receiver is generating noise in the DAC's power and ground system, and this should be measurable. I'm putting together a test setup to better measure this. Yet that is still an intermediate effect; I don't know how to test the analog out well enough to actually measure the analog changes. I think I would have to build the world's best ADC in order to do so. Don't hold your breath on that.

...

Improvements at this level do not cause changes to gross frequency response or distortion, yet they do make things sound 'more real.' The problem is that nobody knows what the actual changes are that cause people to perceive the sound as being more real. If we did, it would be a lot easier to set up tests to look for it. And on top of that, because it is so intertwined with the aural perception system, it is probably different for different people.

i7 11850H + RTX A2000 Win11 HQPlayer ► Topping HS02 ► 2x iFi iSilencer ► SMSL D300 ► DIY headamp DHA1 ► HiFiMan HE-500
Link to comment
Jud, I agree with you. There is lack of clear evidence what makes the audible difference. Some possible scenarios were given with believe that they can explain the audible difference, but we have trouble to confirm them with measurements.

 

There has not been a beginning-to-end suite of measurements showing analog effects, and further showing those analog effects are audible, for these devices. (I say this as a happy user of a Regen.)

 

I think what I was talking about regarding distortion resulting from jitter was a little different, however. Measurable types of analog distortion result from jitter. We know these distortions have audible effects (slightly different from being audible themselves - for example there is research on hearing aids where aids using Class D amplifiers have jitter and the concern is the impact to understandability of speech, entirely apart from whether the noise, THD or IMD is itself separately audible). What has not been academically tested is whether the levels of analog distortion characteristic of today's async USB DACs have any such audible effects (to understandability or appreciation of music, apart from audibility of the distortion itself).

One never knows, do one? - Fats Waller

The fairest thing we can experience is the mysterious. It is the fundamental emotion which stands at the cradle of true art and true science. - Einstein

Computer, Audirvana -> optical Ethernet to Fitlet3 -> Fibbr Alpha Optical USB -> iFi NEO iDSD DAC -> Apollon Audio 1ET400A Mini (Purifi based) -> Vandersteen 3A Signature.

Link to comment
I have most of times been quite satisfied with my current system until heard a better system. That has happened several times ;) Just want to make sure I have the best sound possible at my budget level.

 

Were the speakers the same in those systems or different? Was the room the same or different? Were the room treatments the same or different?

 

You may be focused on the wrong major contributors to SQ.

 

What is your total budget level?

Link to comment
Yes, I linked an Q&A article part 2 in my previous post

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/cleaning-universal-serial-bus-industry-standard-cables-connectors-and-communications-protocols-between-computers-and-electronic-devices-signal-digital-analogue-converter-31736/#post639093

Read also the 1st part of Q&A Q&A with John Swenson. Part 1: What is Digital? | AudioStream

 

A very simplified example scenario to make a picture (read rather the John Swenson Q&A series):

Noise originating from computer hardware (it depends on load = running software) on ground computer plane can cause time inaccuracy in generating USB packets at computer side. That so called packet jitter results to ground plane noise in DAC side USB receiver. That noise can affect DAC clock and can result to time inaccuracy of generated analog values. If that time inaccuracy has some pattern (it can depend on computer software and hardware), it can cause some form of sound coloration, detail loss or background noise. Depending on level of isolation between digital and analog part of DAC it can happen that noise from digital part can influence the analog part also directly.

 

This is testable. There are all sorts of tools for loading the computer with inane tasks to drive up CPU, Memory, Storage utilization.

 

I could setup a clean Windows 10 box and ADC a track with a minimal load and another with fairly high load.

 

I wonder how many could guess which is which.

Link to comment
This is testable. There are all sorts of tools for loading the computer with inane tasks to drive up CPU, Memory, Storage utilization.

 

I could setup a clean Windows 10 box and ADC a track with a minimal load and another with fairly high load.

 

I wonder how many could guess which is which.

 

I agree it's testable and share the sentiment that demonstration of objectively measurable differences are the only good way to answer the permaquestion.

 

I wouldn't test it the way you suggest -- fraught with many too many confounding issues.

 

What I would accept would be a phase error plot at the connected DAC with focus on low offset or "close in" noise e.g. 1-10hz.

 

I think there is an error in doing the measurement which is easy to do or convenient to do, rather than the correct measurement.

Custom room treatments for headphone users.

Link to comment
I agree it's testable and share the sentiment that demonstration of objectively measurable differences are the only good way to answer the permaquestion.

 

I wouldn't test it the way you suggest -- fraught with many too many confounding issues.

 

What I would accept would be a phase error plot at the connected DAC with focus on low offset or "close in" noise e.g. 1-10hz.

 

I think there is an error in doing the measurement which is easy to do or convenient to do, rather than the correct measurement.

 

We will have to agree to disagree. My approach is if no one can reliably tell a difference then what need of further investigation is warranted?

 

If they can then we move onto contributors. If someone can't reliably tell the difference between 320Kbps MP3 and 24/192 there's no sense in testing them for 16/44.1

Link to comment
We will have to agree to disagree. My approach is if no one can reliably tell a difference then what need of further investigation is warranted?
You are conflating two methods of testing. Do you mean:1) "Reliably tell a difference" as in subjective testing by hearing?2) Reliably measure a difference, as in objective measurement?But suppose I assert that your ADC wipes out any subjective or objective difference? Prove to me it doesn't, and no I won't accept "proof" as in "no one can hear it", I want objective proof as in "no one can measure it".

Custom room treatments for headphone users.

Link to comment
Okay so the picture is far more complicated than I though. Kind of wished it were simpler :) Thank you for all the explanations!

 

Now on comparison list Ifi Micro/Nano iUSB3.0, Intona and new UpTone Audio device.

 

the picture is far more complicated in mechanistic terms

 

you can still just do a series of listening tests, both extended passages, and A/B switching to listen for differences

 

just be sure you have a friend help you so you can do blind tests, and not be seeing the difference instead of hearing it

 

then the question becomes what to concentrate on listening to...

 

one thing I have seen printed somewhere (Stereopile??) is that jitter gives a "veiled" sound

 

they also have a test disc with jitter added at various levels to the tracks

Link to comment
You are conflating two methods of testing. Do you mean:1) "Reliably tell a difference" as in subjective testing by hearing?2) Reliably measure a difference, as in objective measurement?But suppose I assert that your ADC wipes out any subjective or objective difference? Prove to me it doesn't, and no I won't accept "proof" as in "no one can hear it", I want objective proof as in "no one can measure it".

 

If you have group of colorblind people debating what they think are multiple shades of red (when it's actually the same red) why move forward with instrumented testing when they show they can't actually discern it?

 

Blinded testing IS an OBJECTIVE measurement.

Link to comment
If you have group of colorblind people debating what they think are multiple shades of red (when it's actually the same red) why move forward with instrumented testing when they show they can't actually discern it?
You don't know the "red" is actually the "same red" without measuring the color. This should be obvious.

Custom room treatments for headphone users.

Link to comment
You don't know the "red" is actually the "same red" without measuring the color. This should be obvious.

 

I don't need to measure a 24/192 file to know it's 24/192. It's simply '24/192'.

 

Just like if I was to get people that believe they can actually hear the difference in Ethernet cables: One such pre-selection test I could do is 128Kbps MP3 vs 24/192. If you can't pick it out you have no value sitting in the test to generate useful data.

Link to comment
Were the speakers the same in those systems or different? Was the room the same or different? Were the room treatments the same or different?

 

You may be focused on the wrong major contributors to SQ.

 

What is your total budget level?

 

Using headphones (Senn HD600 for now) so not so many variables. Budget is 3000€ including DAC, hp amp, USB cleaner and cables. Hugo2 is strong contender since won't need hp amp and interconnect between. If the sound just is right...

Link to comment

Hi miksu8,

 

I took the liberty of doing something that people that are without agenda should be encouraging you to do: contact the manufacturer. After some delay and a re-prompting email sent here is what Chord says:

 

"I apologise for the delay on this, it would appear that your email didn’t make it through to us.

 

We reclock everything on the input of the Hugo 2 anyway, so it won’t be needed for your system.

 

I hope this helps!

 

Kind Regards

Tom Vaughan

Pro-Audio & Manufacturing"

 

Never mind the fact I used their web-form.... sigh.

Link to comment
Hi miksu8,

 

I took the liberty of doing something that people that are without agenda should be encouraging you to do: contact the manufacturer. After some delay and a re-prompting email sent here is what Chord says:

 

"I apologise for the delay on this, it would appear that your email didn’t make it through to us.

 

We reclock everything on the input of the Hugo 2 anyway, so it won’t be needed for your system.

 

I hope this helps!

 

Kind Regards

Tom Vaughan

Pro-Audio & Manufacturing"

 

Never mind the fact I used their web-form.... sigh.

 

 

...another way of saying they use async USB input.

One never knows, do one? - Fats Waller

The fairest thing we can experience is the mysterious. It is the fundamental emotion which stands at the cradle of true art and true science. - Einstein

Computer, Audirvana -> optical Ethernet to Fitlet3 -> Fibbr Alpha Optical USB -> iFi NEO iDSD DAC -> Apollon Audio 1ET400A Mini (Purifi based) -> Vandersteen 3A Signature.

Link to comment

Even though there's debate on whether USB filtering products are worthwhile, I think most people would agree that downstream components (e.g., speakers, amps, etc.) and processes (e.g., DSP, listening room optimization, etc.) have more impact on sound quality than the electronics. USB filtering products might make more sense for someone trying to squeeze out the last bit of improvement they can get.

Link to comment
Hi miksu8,

 

I took the liberty of doing something that people that are without agenda should be encouraging you to do: contact the manufacturer. After some delay and a re-prompting email sent here is what Chord says:

 

"I apologise for the delay on this, it would appear that your email didn’t make it through to us.

 

We reclock everything on the input of the Hugo 2 anyway, so it won’t be needed for your system.

 

I hope this helps!

 

Kind Regards

Tom Vaughan

Pro-Audio & Manufacturing"

 

Never mind the fact I used their web-form.... sigh.

 

Thanks! Also got confirmation that cleaning USB power for Hugo2 won't have impact on SQ so based on these all is left is galvanic isolation, which Intona should perform well.

Link to comment
Thanks! Also got confirmation that cleaning USB power for Hugo2 won't have impact on SQ so based on these all is left is galvanic isolation, which Intona should perform well.

 

You should ask Chord if they galvonically isolate their USB input. Correct me if I'm wrong but they don't directly bus power off the data input. They have a separate connection for that.

Link to comment
Even though there's debate on whether USB filtering products are worthwhile, I think most people would agree that downstream components (e.g., speakers, amps, etc.) and processes (e.g., DSP, listening room optimization, etc.) have more impact on sound quality than the electronics. USB filtering products might make more sense for someone trying to squeeze out the last bit of improvement they can get.

 

I have started to believe that out of single components DAC's output makes the biggest contribution to sound for me. If DAC is giving the sound you like with a "neutral system", its hard to ruin it with lesser speakers or amp. But if DAC is outputting sound you don't like (harsh, grainy, digital sounding etc) you just can't fix that with any amp / speakers. At most you can try to hide it with less revealing system.

 

A bad digital source (laptop) you can fix with a dedicated USB cleaner.

Link to comment

plissken, you cannot speak for all forum members with their different audio equipment if different USB tweaks can provide them sonic improvement or not.

 

Isolating USB input from influences of computer noise costs some money and DACs have solved it on different level - from none to excellent.

i7 11850H + RTX A2000 Win11 HQPlayer ► Topping HS02 ► 2x iFi iSilencer ► SMSL D300 ► DIY headamp DHA1 ► HiFiMan HE-500
Link to comment
plissken, you cannot speak for all forum members with their different audio equipment if different USB tweaks can provide them sonic improvement or not.

 

Isolating USB input from influences of computer noise costs some money and DACs have solved it on different level - from none to excellent.

 

You find out where I'm making that argument in this thread and you are a daisy if you do. Your bias is clouding your reading comprehension

Link to comment
Hi miksu8,

 

I took the liberty of doing something that people that are without agenda should be encouraging you to do: contact the manufacturer. After some delay and a re-prompting email sent here is what Chord says:

 

"I apologise for the delay on this, it would appear that your email didn’t make it through to us.

 

We reclock everything on the input of the Hugo 2 anyway, so it won’t be needed for your system.

 

I hope this helps!

 

Kind Regards

Tom Vaughan

Pro-Audio & Manufacturing"

 

Never mind the fact I used their web-form.... sigh.

 

Not convincing. When you find ANY manufacturer that will tell you their USB input "needs help", let us know. Basically, they will pretty much all tell you they've done a good job with the USB input. Even if they have no particular understanding of it or put any special thought into it.

Exactly the reason why these USB devices exist and help in many cases.

Main listening (small home office):

Main setup: Surge protector +>Isol-8 Mini sub Axis Power Strip/Isolation>QuietPC Low Noise Server>Roon (Audiolense DRC)>Stack Audio Link II>Kii Control>Kii Three (on their own electric circuit) >GIK Room Treatments.

Secondary Path: Server with Audiolense RC>RPi4 or analog>Cayin iDAC6 MKII (tube mode) (XLR)>Kii Three .

Bedroom: SBTouch to Cambridge Soundworks Desktop Setup.
Living Room/Kitchen: Ropieee (RPi3b+ with touchscreen) + Schiit Modi3E to a pair of Morel Hogtalare. 

All absolute statements about audio are false :)

Link to comment
Not convincing. When you find ANY manufacturer that will tell you their USB input "needs help", let us know. Basically, they will pretty much all tell you they've done a good job with the USB input. Even if they have no particular understanding of it or put any special thought into it.

Exactly the reason why these USB devices exist and help in many cases.

 

Using the same logic, you can't trust the claims of the USB filtering manufacturers either.

Link to comment

It's clear that the USB signal generated by most computers is just plain terrible. Otherwise we wouldn't be seeing report after report, here and on many other forums, saying that some type of decrapifying device or devices inserted between the computer and the DAC improves the systems SQ.

 

What's also interesting is that there are many reports suggesting that introducing a quality DDC (e.g., Singxer SU-1/F-1 and Mutec MC-3+USB/MC-1.2) in the chain between the decrapifiers and the DAC improves SQ in the system more than the decrapifiers alone. These DDCs output SPDIF coax, AES, or IS2. In some cases, it has been reported that the DDCs by themselves produce better SQ than the decrapifiers do by themselves.

 

What would explain that?

 

One possibility is that the USB interfaces on most DACs aren't all that good compared to the other digital inputs. So even cleaning up the USB signal dramatically doesn't "fix" the problem. That the D/A conversion process using USB as the source is problematic. That whatever anomalies, no matter how small, that exist in the USB signal "poison" the digital analog conversion done in the DAC negatively affecting the SQ.

 

Is it possible that the DDCs, being all digital, handle the USB anomalies much better and are able to pass on a relatively clean converted digital signal with low jitter via SPDIF, AES, or I2S that has far less of a chance of poisoning the digital analog conversion done in the DAC? Maybe the SPDIF, AES, or IS2 connections are just inherently better.

 

I am not an electrical engineer so I could be all wet, but I thought I would throw that out there for discussion or gentle education!

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...