Do any physical aspects of the computer have any bearing on the sound quality?

Is that the same old guy that reckoned that his Linn Sondek sounded different with the drive belt the other way up?

 

Except no one has claimed to have any problem or proposed that there might even be one. Paying for an entirely separate device just to stream files - now THAT is a problem

 

Heh!

It's like trying to figure out if the side door of your average U.S build sedan car will have any significant impact on its aerodynamic performance.

 

I was skeptic too until I've tried myself.

 

Right....let me check if my other leg has any bells attached to it.

 

I'm confused on what you are trying to do. Can you describe it more? Is this for a second separate listening area while you work at the computer?

 

What processes / applications do you kill in order to improve the sound?

 

You're wasting your time. This is the same mentality as the infamous "green felt on CDs make them sound better" incident. Pure BS.

 

Where would I plug my Audioquest King Kong USB cable, or would I opt for their ethernet cable?

 

I have learned over the years to take "Audio claims" with a grain of salt.
Most are made with the stipulations of "Well my ears hear it", "If you have better equipment you would hear it", or the worst "I trust what I hear"

And of course none of the claims are ever able to be validated.
In fact even questioning the claim makes the claimant go round and round...........

 

Yes, that's right!

I use a Western Digital Elements hard drive as a sort of go-between the two systems.

I had no speakers when I first got the desktop. I have been experimenting with speakers and DACs.

I couldn't see any reason why the computer itself should have much effect on the sound but it doesn't do any harm to check.

 

In that case I think the DAC, bookshelf speakers, and listening area would not be enough to reveal any changes to a computer that "may" be perceived by some systems in a perfect environment.
Note, as long as you are using on outboard DAC and not the computer analog output.


Aka, I wouldn't sweat it.

 

I was using a Windows computer with a USB sound card (Focusrite Scarlet 2i2) to digitize vinyl (needledrop). I was doing some bench measurements of the distortion of the 2i2 and noticed every now and then, a jump in the distortion. It took me a while to figure out what was going on, but I finally found out that every few minutes or so, the needledrop would have a very short drop out or gap in the signal. After much research, I discovered something called DPC (Deferred Procedure Call) Latency. This is a measure of how long it takes windows to execute certain processes after it has to hold off the execution to do something else. Focusrite has a good description of it here:
Troubleshooting DPC latency
My computer was taking too long to process certain things and as a result, dropped some of the real time information (my audio), resulting in a very brief gap in the audio waveform that showed up as a spike in the distortion meaurement.
I spent a lot of time trying to track down the cause and eliminate it. The web is full of suggestions. I spent weeks working on it, trying every suggestion I could find on the web. Mostly involving shutting down as many non-essential processes as possible (battery charge monitor, you name it). In the end I could get my system to work almost 100% of the time, but I had so many windows processes disabled, that the computer was not usable for anything else. The bottom line is Windows is not designed to be a Real Time Operating System (RTOS) and tries to emulate an RTOS by handling interrupts as quickly as possible without dropping any real time information. I thought about buying a faster machine (even though mine was a pretty powerful one to begin with). So I tested a few machines in the office using the Latency Monitor Program:
Resplendence Software - LatencyMon: suitability checker for real-time audio and other tasks
The bottom line was even very fast laptops with solid state drives often reported marginal or unsuitable latency for real time operations. So I ditched using a Windows machine entirely and went to a dedicated digital recorder.
That was a USB ADC. Would it also be true for a USB DAC? I don't know, but I can certainly see how it could be.
So I am not so quick to dismiss those that claim to hear differences in computer audio based on other processes running. I suspect it could well be true. For me the answer is to not use a computer when a real time signal is required (like a USB DAC). But it is a totally different issue when a packet interface (like ethernet or wifi) is used that doesn't require real time operation and ensures no lost packets. In that case, I don't see how anything going on in the computer can affect the audio quality.

Sorry for the long, rambling response. Bottom line is look into DPC latency before you totally write off the effect of other processes running on the computer.

Terry

 

I thought I'd give the OP a chance to explain how he achieved his gains and if it wasn't too complicated or risky I'd give it a shot. I've got time to burn lol .....

 

Probably explorer.exe

I was watching the CPU meter while playing a 192/24 FLAC on a circa 2012 i7 PC. It was averaging 5-7% with some peaks hovering around 11%. When that music stopped the CPU meter averaged...5-7% with peaks around 11%. The culprit: Chrome browser with 5 tabs opened. CQFD decoding FLAC and playing audio is a non-event for any PC .

BTW, @TerryS an RTOS isn't an OS which process events in real time but one which is able to process events in a consistent and predictable short span of time. Your problem was most likely caused by a less than optimal USB driver and also probably because you can't buffer an analog signal so ADC is different from DAC, but I'm pleased to know that you found a solution that works for you.

 

Didn't realise that was at all controversial.....

 

Which way up do you prefer it, and why?

 

In my experience, the sound card/DAC, power supply, and even the case can impact the sound you hear.

The sound card/DAC is self-explanatory.

The power supply, when adequate, should deliver clean power to the CPU and other parts. For the CPU, it reduces any incidence of calculation errors. And, today's PSUs are much quieter than older ones. The fan only spins on some models if it reaches a certain load threshold. There are some models that are fanless. Finally, most computers only draw about 230 watts on average, but it's always good to have am little more for reserve, and depending on your computer configuration.

Many computer cases today have a lot of holes for ventilation/fans/radiators. There are a few models that are insulated for sound, including many older cases. In most configurations, the most fans that are needed are two case fans. In my current case, i'm just running one exhaust fan. The computer still runs quite cool, especially since the power supply is mounted on the bottom rather than the top, as in most older units.

 

Firstly, make sure the side that has been precision ground, goes on the inside, and then which way "up" that sounds better. Some Linn dealers used to mark the belt on the outside with direction arrows with a grease pencil, others put in the tray ready to be fitted. How, why? By listening to it.

If you look at the Incredible lengths manufacturers go to reduce motor vibrations, the machining of the belt is very important to performance.

 

Others may be skeptical, but i've experienced it time and time again. There is no substitute for personal experience.

 

A good power supply will also help ensure a clean current to the CPU so it will reduce or eliminate processing errors. In fact, one sign that a PSU is failing is that your computer will "act up". It will freeze and delete or corrupt files. And, yesm EMI is always an issue.

Again, we can agree on overall things, but disagree on particulars. Experience is everything.

I agree that stopping processes is bologna.

 

Just a comment on case fans. I had been using a cheap Antec case fan and it was pretty quiet. I recently bought a Corsair magnetic fan for $24. It is not quiet. It may be because I have it connected to the case fan 2 socket because it, like the Corsair, is a four-pin. The case i socket is a three-pin because it's an older mobo. What I need to do is fix the connector to the Corsair fan so I can use the three-prong socket. The reason is that the #1 socket provided a variable speed control.

I accidentally ordered the Corsair 100R instead of the quiet edition. So, I looked for a cheap sound dampening solution, and what I used doesn't seem to be helping. One of these days, i'll have to order the 600Q, or at least the 100 quiet version and transplant the computer innards again.

 

OK, this makes sense. My apologies to @beppe .

 

Ya, I don't claim to be an expert on RTOS, but using a couple of the DPC latency checker tools I found on the web showed that the Windows machines I checked were either unsuitable or marginal for real time operations. Like I said, these were pretty powerful laptops with solid state drives, so it looked like it was not really practical to buy a new machine to improve the situation. Possibly one of the high dollar gaming machines would perform better, but that didn't seem like a good gamble either.
My point is that I tend not to discount the effect of all the dozens of processes that are running in the background of your Windows machine on things that need to happen in real time (like a USB DAC). It scared me enough that I won't use a PC in such a situation.

Terry

 

not that I know of - he seems wedded to digital