Digital Signal Transmission

[TomJ]

I think with this story the musician should be the one who prepares the meal and the result is a matter of taste. But the scientist can determine the quality of the ingredients.

[fas42]

34 minutes ago, pkane2001 said:

What do you think distortion is? Poison! :)

 

Everything is distorted! The very best system, that even I 🤪 could dream of, has distortion - what counts is whether that distortion, poison if you want to call it that, is significant to the listening experience!

[Confused]

It’s always wrong to blame the meal.  Get your fork working in the right zone and your taste bud brain system will compensate for any anomalies in the food and let the true magic of the meal shine through.

[fas42]

4 minutes ago, Confused said:

It’s always wrong to blame the meal.  Get your fork working in the right zone and your taste bud brain system will compensate for any anomalies in the food and let the true magic of the meal shine through.

 

Yes, mood plays a part. And of course, good ol' red wine climbs on board as well, 😉 ... but one can stand fully apart from the 'distortions' of mood and drugs, and decide whether the food was as prepared as well as as could be reasonably expected - in the cool light of day.

 

The meal has to stand on its own. If it has a flavour imbalance, or a mistake was made in one step, then there's no getting around this - you can try covering this up with over application of a 'tubey' sauce, say - but those who know how good the raw ingredients can come together will just be irritated by this - this is where experience comes into it, so strongly.

[sandyk]

2 hours ago, TomJ said:

However, there is still a difference here - I trust the Linux nerds to know what they are doing. Here with the audiophiles, I'm starting to doubt it. 

That has been your assertion right from your very first post. So why are you REALLY here ?

It's certainly NOT to learn from the experiences of others.

[asdf1000]

5 hours ago, Miska said:

IMO, SINAD alone is as useless figure as any other number taken out of context. I don't think it helps saving $$$.

 

This is true but we all know which site made SINAD a focus but let's not forget his APx555 measures a bunch of other things too, not SINAD alone...

 

His measurements have caught one company getting maximum SINAD at the cost of much higher than normal output impedance.

 

[asdf1000]

7 hours ago, PeterSt said:

Sadly, No. There is no such thing as noise to be measured in the context of audio data, as the HF protocol frequency data *is* noise to the audio domain. There's really nothing to measure there.

Tom didn't mention anything about audio domain there though ?

 

He only said "- Noise reduction - could be measured"

 

And last update I saw a couple years ago was John Swenson is custom building a very high resolution A to D to be able to show these (noise?) measurements? If it was only audio domain then I don't think he would be going to such efforts...

[pkane2001]

2 hours ago, RickyV said:

What if the food wasn’t poisoned or not poisoned but it was slightly bitter. Would the scientists know what to do about it or wound the master chef know what to do?

 

The analogy is flawed. It's the fork we are discussing here, not the food. The food is the music, the art. The fork is the utensil that's used to consume food, just like audio equipment is used to consume music. Any individual that thinks that it's the fork that makes food taste better is certainly welcome to their opinion, but the fork doesn't make the meal. Does the chef know how to make a better fork to make his food taste better? Probably not.

[fas42]

44 minutes ago, pkane2001 said:

 

The analogy is flawed. It's the fork we are discussing here, not the food. The food is the music, the art. The fork is the utensil that's used to consume food, just like audio equipment is used to consume music. Any individual that thinks that it's the fork that makes food taste better is certainly welcome to their opinion, but the fork doesn't make the meal. Does the chef know how to make a better fork to make his food taste better? Probably not.

 

Wrong. The fork has practically zero to do with anything. Don't try to force the analogy to follow the audio chain sequence too closely, because it won't work. If you want something that roughly matches, consider the raw ingredients as the 'art' - and, that the kitchen has to recreate a certain, very specific dish from those , that ticks all the boxes ... it's the ability of the restaurant to duplicate precisely, in every area, how that plate of food matches what was served up, the very best time they, or someone else, did the dish.

[Rexp]

4 hours ago, pkane2001 said:

 

Are you seriously doubting it? Give you an example:   

 

observation by "normal" people = Earth is flat (I can see this with my own eyes looking out the window)

observation by "scientific" people = Earth is a spheroid planet rotating around a star that's part of a galaxy that itself is part of a local galaxy group that's part of a universe containing billions of galaxies

 

Which one do you think is of a "higher" standard?

Difference is, subjectivists can see the audio world is round unlike  blinkered scientific folks. 

[pkane2001]

You guys love to argue, but when it comes to providing any real evidence to support your arguments... crickets.

 

So, in bidding you adieu, always remember and never forget:  it’s not the fork that bends, it is only yourself ;)

 

[fas42]

24 minutes ago, pkane2001 said:

You guys love to argue, but when it comes to providing any real evidence to support your arguments... crickets.

 

So, in bidding you adieu, always remember and never forget:  it’s not the fork that bends, it is only yourself ;)

 

 

So sayeth the RWD ...

[sandyk]

1 hour ago, pkane2001 said:

You guys love to argue, but when it comes to providing any real evidence to support your arguments... crickets.

 

So, in bidding you adieu, always remember and never forget:  it’s not the fork that bends, it is only yourself ;)

 

As I suggested in another post, feel free to organise a series of correctly performed DBTs in the USA where the majority of members reside, as DBTs are supposedly the " Gold Standard" (except when the results don't coincide with E.E. and I.T. member expectations.)

 If I wasn't in Australia , I would be more than happy to demonstrate my claims using my own equipment.

In fact, when I was still living in Sydney I challenged Prof. Scott to organise with Sydney University's technical people to put my reports to the test.

 That didn't happen, so I seized the chance to contact Martin Colloms who I didn't realise was a member here until I saw a post by him, and he did put my claims to the test using comparison .wav files that I provided, with 6 separate sessions of 8 repeats in each session.

 

 Why is it so damn hard to do something similar in the USA where there are so many members ? 

 

I also recall that some years ago Chris organised a Symposium where the majority of members attending heard clear differences between SSD and HDD using the exact same Music files.

IIRC, Chris described the HDD as sounding a little muddy in comparison with the SSD .

[PeterSt]

4 hours ago, asdf1000 said:

And last update I saw a couple years ago was John Swenson is custom building a very high resolution A to D to be able to show these (noise?) measurements? If it was only audio domain then I don't think he would be going to such efforts...

 

4 hours ago, asdf1000 said:

12 hours ago, PeterSt said:

Sadly, No. There is no such thing as noise to be measured in the context of audio data, as the HF protocol frequency data *is* noise to the audio domain. There's really nothing to measure there.

Tom didn't mention anything about audio domain there though ?

 

The only thing what one could want to measure is noise in the realm of audio. Please remember that Tom responded to my outlay of what causes jitter in the DAC's clock signal (noise). Next he wants to measure the noise in the USB signal (he will correct me if I'm wrong) but there is nothing to measure because that is nothing but noise because of how the (protocol) signal is set up (or "is"). It is data as such comprising of 0's and 1's passing at a frequency of 24 MHz. The audio is embedded in there.

No analyzer can see noise around audio there (whether audio is playing or not). Only jitter can be seen (eye diagram).

 

What I showed was a sinus of the in-DAC clock signal. So there theoretically you could see noise, weren't it that that sinus is 22MHz to begin with, with superimposed noise which can be seen (like in the picture I showed) but not measured. It could only be measured when the clock signal is off so the noise floor could be seen (FFT). Paul will have his doubts because everything can be measured, but he won't do it because he is not interested in those frequencies (this is not a stab, Paul, it is just so, including that it can be measured with crazy $). But with that signal off, all implied distortions (noise) will be out of the way as well. Also, the fact that audio plays or not, will make a difference.

 

Below you see three pictures of my own measurements; 

1. An 1 KHz signal at -0dBFS.

2. Music playing at -0dBFS.

3. Music playing at -144dBFS (silence).

 

 

 

 

 

It shows that jitter changes with music playing. This is nothing new, if we only realize that this is so. Thus, a system does not only show a jitter figure (number) it also shows a jitter signature depending on the music playing (as in: the one instrument may exhibit a different jitter figure than the other).

This is all 99% depended on the system as a whole, thus including the PC (or streamer - just saying, it is a computer just the same), the software impeding computer noise (or less), the cable towards the DAC, its capability to shield noise, its capability to prevent its own noise to escape from it (and pollute its surrounding), a possible isolator to block remaining noise ...

etc.

 

The 1% remaining is in the DAC itself and it is about its capability to block jitter as such, but also its capability to block noise again because else jitter is readily created again (my earlier story around the clock signal noise). But, that 1% is only 1% if relative to the above some improvements have been applied, so that this remains:

 

 

(the earlier 3 pictures play 44.1KHz Redbook and the last one 352KHz DXD, but this does not matter).

 

Might someone be interested, here's that often mentioned "eye" from this situation :

 

 

The "noise" I mentioned is visible through the eye being larger (less noise) or smaller (more noise). In the corners you can still see noise per the deviating lines from what should be (the thickest lines is what should be).

Btw, it is easy to confuse noise with jitter here because the eye shows jitter for the time axis (horizontal) and noise on the voltage axis (vertical). This requires a bit of experience, like observing the crosses and the sideways deviations.

 

This was the difference occurring after i2s was galvanically isolated from the front end (say the USB input).

Later, the upgrade with the galvanic isolation in USB itself (in-DAC) was created, but I did not measure things like this over it because for me it is sufficient to know the theories of what will happen, plus I (seem to) know what to listen for. Also, as this isolation is an option which can be by-passed, everybody can most easily hear the difference (and check what they payed for).

So I don't sell figures (that I recall), but good sound. I make the measurements all right when deemed necessary, but I merely make them for laying out that it's not really rubbish what is being done. And as some will know, the wires in my own DAC to apply USB cable measurements are soldered in there now for what ... 4-5 years ? I simply never get round to really applying the measurements because they take huge amounts of time. Some day I will.

 

No more lyrics.

 

[PeterSt]

4 hours ago, pkane2001 said:

Any individual that thinks that it's the fork that makes food taste better is certainly welcome to their opinion,

 

Try a silver fork, and you will know. 🤑

[Confused]

7 hours ago, asdf1000 said:

 

His measurements have caught one company getting maximum SINAD at the cost of much higher than normal output impedance.

 

Slightly off topic I know (not that this thread is on topic anymore), but could I ask which manufacturer you are referring too in the above.  It is an interesting comment, because in my experience there is a very real relationship between output impedance and certain aspects of sound quality.  (I am assuming the above refers to an amplifier or maybe a headphone amp) 

[TomJ]

8 hours ago, sandyk said:

That has been your assertion right from your very first post. So why are you REALLY here ?

It's certainly NOT to learn from the experiences of others.

I am here get answers to the question wow digital signal transmission can change the sound and where.

So I learned, that the sound can only be changed while converting from D to A inside the DAC. And that network devices can not change the sound, but influence the converting. This is done by electrical noise from the network device over the electrical path. Jitter problems in Switches has nothing to to with worsen sound quality directly, but can produce more electrical noise.

I learned, that an Isolator do not block all the noise, just a percentage.

But if I prevent noise coming in the DAC and data is send correctly, the network is no longer a problem soundwise.

Right?

[sandyk]

1 hour ago, TomJ said:

So I learned, that the sound can only be changed while converting from D to A inside the DAC. 

No. It can also  be changed by the PSU area of the source due both to the components used in the PSU itself, but also as Peter has explained by different software loading of the PSU area. Even the external PSU used to power Regens and switches etc. can change the sound qualities to some extent, which is why some members daisy chain ultra  low noise LT3045 voltage regulators in order to increase apparent HF detail. This is due to the output impedance of these types of  regulators being much lower at >100kHz than in the lower frequencies.

[TomJ]

43 minutes ago, sandyk said:

No. It can also  be changed by the PSU area of the source due both to the components used in the PSU itself, but also as Peter has explained by different software loading of the PSU area. Even the external PSU used to power Regens and switches etc. can change the sound qualities to some extent, which is why some members daisy chain ultra  low noise LT3045 voltage regulators in order to increase apparent HF detail. This is due to the output impedance of these types of  regulators being much lower at >100kHz than in the lower frequencies.

I mean only during the conversion of D to A, or then in the analogue chain. The conversion is influenced by HF of PSU entering the DAC. But before conversion, there is no sound alive to change😀 - only data (and electrical noise/HF) and with TCP/IP and USB we get the data without manipulation into the DAC.

[asdf1000]

3 hours ago, Confused said:

Slightly off topic I know (not that this thread is on topic anymore), but could I ask which manufacturer you are referring too in the above.  It is an interesting comment, because in my experience there is a very real relationship between output impedance and certain aspects of sound quality.  (I am assuming the above refers to an amplifier or maybe a headphone amp) 

 

It's a bit silly we can't post links to the site here but here it is and here is Amir's comment about it.

 

https://www.audio “science” review/forum/index.php?threads/smsl-m400-balanced-usb-mqa-dac-review.13732/

 

 

So he's not just talking about it's SINAD of 121dB, so I'm not sure why people just seem to talk like this is the only measurement he shares.

 

 

 

 

 

[pkane2001]

10 minutes ago, asdf1000 said:

 

It's a bit silly we can't post links to the site here but here it is and here is Amir's comment about it.

 

https://www.audio “science” review/forum/index.php?threads/smsl-m400-balanced-usb-mqa-dac-review.13732/

 

 

So he's not just talking about it's SINAD of 121dB, so I'm not sure why people just seem to talk like this is the only measurement he shares.

 

 

https://www.audioscìencereview.com/forum/index.php?threads/smsl-m400-balanced-usb-mqa-dac-review.13732/

[fas42]

10 hours ago, TomJ said:

I mean only during the conversion of D to A, or then in the analogue chain. The conversion is influenced by HF of PSU entering the DAC. But before conversion, there is no sound alive to change😀 - only data (and electrical noise/HF) and with TCP/IP and USB we get the data without manipulation into the DAC.

 

Correct. All the other circuitry can indulge in any sort of silliness that it wants to get up to, misbehave as much as it likes, electrically - you just have to stop it reaching across into the areas where the analogue waveform is directly created, or handled.

 

Your options are:

 

 * Assume all non-analogue circuitry is generating vast quantities of interference - and therefore make the analogue area bulletproof under all circumstances; have myriad layers of shielding, and isolation built into the analogue areas.

 

* Fiddle like crazy with everything that works in the outside of analogue areas - trying to make the circuitry be as benign as possible, as regards to creating the type of noise that impacts the analogue parts. This is the reasoning behind these ultra specialised streamers, PCs, etc - try and tame the mechanisms that spew out the really bad interference.

 

* Use a combinations of these approaches. In the real world, this is what most people do - and it's probably the most cost effective  way.

[fas42]

13 hours ago, Confused said:

Slightly off topic I know (not that this thread is on topic anymore), but could I ask which manufacturer you are referring too in the above.  It is an interesting comment, because in my experience there is a very real relationship between output impedance and certain aspects of sound quality.  (I am assuming the above refers to an amplifier or maybe a headphone amp) 

 

Was just reminded of the audio friend up the road ... he's been in a 'battle' for years trying to simplify the system, and still get the quality without relying on a buffer, anywhere. It's nearly always been a trade-off - the buffer kills some of the quality; but using it means that higher sound levels don't start to compress, when using the volume control of the source device. The problem is that the higher current drive needed for higher volume starts to expose issues with the power supply of the source unit - here, just getting a better quality source component is probably the easiest solution.

[sandyk]

11 hours ago, TomJ said:

I mean only during the conversion of D to A, or then in the analogue chain. The conversion is influenced by HF of PSU entering the DAC. But before conversion, there is no sound alive to change😀 - only data (and electrical noise/HF) and with TCP/IP and USB we get the data without manipulation into the DAC.

 Your reply is from your closed minded perspective. In  fact, the PSU area of the source, including the PSU area for the SSDs  can also affect Digital Video the same way that it affects Digital Audio, just as the same data on SSD and HDD can sound a little different, even when played directly from System Memory. I am also able to demonstrate this to those who have a decent internal video card and a good quality external monitor, as well as a better than average main system.

 I have previously had my PSU reports in this area verified with a series of 6 separate DBT sessions with results publishd in HiFi Critic Vol. 6 No.1

 It is becoming increasingly clear that you came here only to TROLL, and not learn  from the  listening experiences of members here.

Neither are you interested in any of the proof that Peter , who BTW, is the designer of one of the most highly acclaimed DACs commercially available, as well as the highly respected XXHE S/W for Windows,(http://www.phasure.com/)

went to a lot of time and trouble to provide.

[sandyk]

1 hour ago, fas42 said:

Correct. All the other circuitry can indulge in any sort of silliness that it wants to get up to, misbehave as much as it likes, electrically - you just have to stop it reaching across into the areas where the analogue waveform is directly created, or handled.

 You have a short memory span Frank, as you have personally heard and verified obvious differences generated in the front end by the PSU area .😉

 You are also neglecting that what audible differences Peter is able to cause with his S/W ,remains after conversion from D to A in the DAC as verified by Manishander in the Red Pill, Blue Pill thread..