Linux vs Windows OS for sound?

[One and a half]

Hi One and a half (interesting name), We can keep going lower and lower on the IT stack to realize that everything contributes to sound quality. So, if I read you posting correctly, changing an OS has a knock on effect on the hardware components acting on a process that changes the noise level/floor. What combination of OS and hardware runs the quietest? Which OS can be configured to have the least amount of extraneous processes? Maybe that's a benchmark that can be tested/validated? Is this leading to the conclusion to procure a dedicated appliance, like Bryston BDP-2 and throw away this idea of using software based devices like ROON that run on traditional PC based hardware?

 

 

What combination of OS and hardware runs the quietest? = Unobtanium

 

Which OS can be configured to have the least amount of extraneous processes? And still function properley, none really.

 

No benchmark or tests available, the combinations of OS and applications would make an accurate test difficult.

 

Here's a great article (1 of 3) which explains the ground plane.

 

Q&A with John Swenson. Part 1: What is Digital?

Read more at Q&A with John Swenson. Part 1: What is Digital? | AudioStream

 

There are two other parts to the article, all well worth reading. Digital signals are actually voltages, therefore are analog in nature. It's the crud that lives on the digital signals that cause the problem, very much like remora on sharks.

[firedog]

The timing is created by the clock in the DAC. The computer has nothing to do with it.

 

Noise and interference can affect the clock in the DAC.

[mansr]

No, it is a plain as day fact. Go ahead, try and read a memory byte without it being clocked. Just try to run an instruction without it being clocked. Lets' see, can you actually do anything on a data buss without it being clocked? How about a serial connection - that would be fun to watch...

 

I am sorry, but you seriously need to go back and rethink what you think you are saying. It's pretty obvious what you think you are saying and what you are saying are two different things.

 

-Paul

 

You're both right, but you're talking about different clocks. Of course a computer uses various clocks to do what it does. Those clocks, however, have absolutely no relation whatsoever to the timing of samples in the DAC during playback. All that matters is that the computer can send data to the DAC fast enough, otherwise there will of course be dropouts.

[Paul R]

That's a little cryptic mate.......go ahead and correct me. Forums are places for learning and discussion. I'm only ears if you are able to point out eloquently and comprehensively why I am wrong.

 

It would be easier if you didn't selectively quote to make yourself look better.

 

It's pretty obvious you do not know what you are talking about, and are extrapolating that ignorance into other areas.

 

"Asynchronous USB" does not change the physical interface - all the electrical signals are the same. What it does do is take control of data transmissions and buffering on the DAC side. In other words, the clocking.

 

Once you do that, you have issues on the DAC that make a difference, including clock drift in the DAC, trigger level variance caused by electrical noise on the physical USB cable - some of which is or at least can be injected from the PC.

 

Instead of being asinine - go put together a test of your own and then report the results. If you are so sure that you cannot get noise into a DAC from a PC, go measure some. You will probably have to pay for lab time to get the equipment you need to do that though, as the measurement gear is rather expensive.

 

If you don't want to do that then... well... free advice. It is not wise to poke the bear here. Most people on this forum know more (*much* more) than the average college graduate, and have the experience to back it up as well.

 

There is another forum that delights in pseudoscientific egomania, but occasionally has some really good postings. You might try over there.

[Size]

Some thoughts in response to your points. I'm not asking for a Nirvana solution, but which combination "sounds" the quietest when listening through an audio system all other components remaining the same? For example, Raspberry Pi 2 (or 3 without wifi turned on) vs a fanless box PC device running Linux (Gigabyte BRIX)? How about including the devices marketed as audiophile types? In respect to "crud" it doesn't interfere with digital content being sent otherwise data integrity would be lost, but it does show up through the same "pipes" like RFI interference.

[Jud]

As I said in my first post - it is possible that a computer/OS could be configured in a way which would have a great impact on sound, but I would class that as improper configuration. Certainly anyone using this site should be more than capable of configuring a computer correctly.

Something as fundamental as an application's memory buffer handling can affect noise on the ground side and thus potentially sound quality. Designing hardware and software to minimize noise on ground is not a trivial matter. If you would like, I can give you a link to web forum correspondence discussing this from the fellow who developed the firmware code responsible for popularizing the input your DAC is likely using.

 

In my scenario, I could. It uses wi-fi.

 

Certainly in terms of eliminating EMI from a computer, this is a good thing.

Two points here: "All this RFI" (agree) and "change the data" (never!).

You would most certainly hear the difference if all this RFI actually did change the data.

But do you really think there are computers out there which are "changing the data"? I wish my bank had one of these magic computers........maybe it could change my $1000 deposits into $2000 ones? That would involve (after all) the simple misrepresentation of a single 0 for a 1. After all, data centres must be pretty confusing places with all that RFI

I think you were overly anxious to find differences between our positions. I agree if the level of noise were extreme enough to change the data, you would very certainly hear it, and it wouldn't just be a matter of lower sound quality - you might not get sound at all.

 

What I'm saying though is that once we've assured the data isn't changed, we are *not* done. There are ways distortion may be caused in digital audio that do not involve any data changes. (Jitter was probably the earliest known example.)

I think you mean synchronous.

 

Nope. Async is better at minimizing jitter, but not perfect. At what level jitter is audible is one of those never-ending discussions.

 

But of course!! Yes, jitter would cause a measurable impact on sound quality. But if anyone here has configured a computer to be affected by jitter has simply mis-configured their computer. The timing/clocking of the data should be done by precision (audio grade) buffers and clocks, outside the computer.

Yes, clocking with async inputs is done at the DAC. Now let's suppose there's ground noise that enters the DAC at a very low level. What is the DAC clock's reference voltage? Ground. So that reference varies, causing jitter *at the DAC clock*. See - not so simple.

 

I don't doubt audibility for one second. In your scenarios above, there would be audible differences. It is just that I would class each of your examples as configuration errors (or audio equipment design errors in the case of improperly isolated USB inputs).

The computer and its OS (intrinsically) will have no impact on the sound quality.

 

Many folks think any properly designed and configured computer and DAC will render perfect audio for all practical purposes. Others do not. I'm not going to argue one way or the other; I will just suggest it is not a trivially settled matter.

[lmitche]

Many folks think any properly designed and configured computer and DAC will render perfect audio for all practical purposes. Others do not. I'm not going to argue one way or the other; I will just suggest it is not a trivially settled matter.

 

Jud, would you agree some computers sound better then others and that the difference may indeed be due to the design and configuration?

[Paul R]

Some thoughts in response to your points. I'm not asking for a Nirvana solution, but which combination "sounds" the quietest when listening through an audio system all other components remaining the same? For example, Raspberry Pi 2 (or 3 without wifi turned on) vs a fanless box PC device running Linux (Gigabyte BRIX)? How about including the devices marketed as audiophile types? In respect to "crud" it doesn't interfere with digital content being sent otherwise data integrity would be lost, but it does show up through the same "pipes" like RFI interference.

That all sounds like a pretty reasonable theory to me, and doesn't conflict with any known facts.

 

Just my opinion here, but using a nice "high powered" server to store your music, stream and buffer from the internet, and serve music (and video, etc.) to essentially lower powered electrically quiet devices over a network seems to produce the most accurate and pleasing sound to me. It's also cost effective, which in the audiophile world, is jaw dropping amazing.

 

-Paul

[Jud]

Jud, would you agree some computers sound better then others and that the difference may indeed be due to the design and configuration?

 

As I said, my intention isn't to try to persuade anyone of a particular position, but to get them to consider whether a model that completely discounts such factors may possibly be oversimplified. If someone concludes these things will not have an effect, that's fine. I just would like them to understand that taking them into potential consideration isn't inconsistent with good science and engineering.

 

 

Sent from my iPhone using Computer Audiophile

[lmitche]

As I said, my intention isn't to try to persuade anyone of a particular position, but to get them to consider whether a model that completely discounts such factors may possibly be oversimplified. If someone concludes these things will not have an effect, that's fine. I just would like them to understand that taking them into potential consideration isn't inconsistent with good science and engineering.

 

 

Sent from my iPhone using Computer Audiophile

Lol, rather like climate change denying, got it!

[amey01]

It would be easier if you didn't selectively quote

It's pretty obvious you do not know what you are talking about, and are extrapolating that ignorance into other areas.

being asinine

 

 

What would be easier....is if you left your patronising, name-calling rubbish at the door......

 

Most people on this forum know more (*much* more) than the average college graduate

 

I don't know about college graduates and people on this forum, but the guys that design this stuff know more (*MUCH* more) than the average keyboard warrior on this forum. Seriously - the guys that design this stuff aren't stupid. There is biphase mark coding, which is part of the redbook specification and exploited by designers to correct bit-shift. Every single DAC oversamples.

 

You have yet to tell me how "noise" from a *properly configured* PC, into a *well designed* DAC will impact sound quality.

 

But if it does, use optical. Use wi-fi (as I do).

 

But regardless - this thread is not about that. This thread is about how the operating system can impact sound quality.

So answer this: Let's assume the "noise" does impact sound quality. We use wi-fi to our streamer. Tell me then how possibly a Debian system could sound different to a Windows 10 system. HOW? Your turn.

[amey01]

Something as fundamental as an application's memory buffer handling can affect noise on the ground side and thus potentially sound quality. Designing hardware and software to minimize noise on ground is not a trivial matter. If you would like, I can give you a link to web forum correspondence discussing this from the fellow who developed the firmware code responsible for popularizing the input your DAC is likely using.

 

Interesting, constructive post. I'll do some research on that.

 

(Jitter was probably the earliest known example.)

 

It is the only known example as far as I know.......??

[Paul R]

What would be easier....is if you left your patronising, name-calling rubbish at the door......

Yawn - I am sure you are right. You win.

 

But did I mention most of the folks who design and market these things *are* actually here on the forum? They tend to hide out when loudmouth know-it-alls start bragging about how much they know on the system. As I said originally, you should go check out the gasbag forums - probably more your style.