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

Discussion in 'Audio Hardware' started by oxenholme, Jan 3, 2018.

Thread Status:
Not open for further replies.
  1. Mike-48

    Mike-48 A shadow of my former self

    Location:
    Portland, Oregon
    I checked my desktop -- a 6-yr-old i7 (4 cores, 8 threads, 8Mb cache) with 8Gb RAM and SSD. Latency reported fine for audio.

    My laptop is 7 yr old with mobile i7 (2 cores, 4 threads, 4Mb cache), 8 Gb RAM, and newer SSD. Latency reports unsuitable for audio, even when it's set to the "low latency" power setting. I have used it for recording, but perhaps results have not been optimal.

    I have thought that the laptop sounds good, but not great, when used for playback into my better audio system. Hmmm....
     
    Last edited: Jan 5, 2018
  2. beppe

    beppe Forum Resident

    Location:
    Venice, Italy
    It's difficult to me to understand why we (mostly all) believe that a different pure copper cable, a different electrolitic condenser with different brand but same characteristic, different brand of spikes, different plugs, etc etc change the way we perceive sound(music) from our systems, but most of us won't even try to stop some unneeded Windows process in our pc used to transfer bit to our dac. With a single click om my .bat file my Windows xp laptop works with only 24 background processes, without latency problems and flawless.
    In my system, with my ears difference is perceivable. Sorry for my bad english :)
     
    CrazyCatz likes this.
  3. anorak2

    anorak2 Forum Resident

    Location:
    Berlin, Germany
    There's a difference between analogue and digital audio. Even though I have doubts that changing a wire makes a difference, it's a given that cleaning the record, upgrading the cartridge, avoiding mechanical bumps etc. influence audio quality (and similar "tweaks" are effective for tape). This is so because basically there are always distortions present, and no matter how much you try to eliminate them, you'll never get rid of all of them. So there is always room for improvement.

    But the whole point of digital is that the physics and the mechanics of the system make absolutely no difference, as long as it works.

    If "tweaks" would make a difference, there are only two explanations:

    - Either the system wasn't working properly before. But that would have been quite obvious and not at all subtle.

    - Or that digital audio works differently than designed, but still works. And that is illogical.

    This is also true if your "tweak" consists of stopping other processes in your system. As long as they're not slowing the system down so much that the music playback becomes stuttery or stops altogether, they can't effect the audio, because that is just how computers work. Therefore stopping them will make no sonic difference.
     
    nosliw likes this.
  4. jkauff

    jkauff Senior Member

    Location:
    Akron, OH
    If you really think you can hear process interference in your audio recording/playback (and some people claim they can hear when Error Correction kicks in on CD playback), use Task Manager or a program like Process Lasso to set all your audio-related processes to Real Time, using a single core (your least used) if you have a multi-core system. Problem solved.
     
  5. Cliff

    Cliff Magic Carpet Man

    Location:
    Northern CA
    Please post an A and B clip of a song. One with all the normal processes running and one after you run your .bat file.
     
  6. beppe

    beppe Forum Resident

    Location:
    Venice, Italy
    Not possible because the file is the same. I've tried with task manager killing one by one every process not related to audio, network and of course processes needed from pc to work properly. Then I've made my own list in a .bat file.
    Years ago many audiophiles were used to use a "only transport" cd player. Different brands do provide different sound signature? Bit is bit and zero is zero. And, if is it true that in digital domain there is no difference, as long as it works, digital cables are all the same? And where's the difference from coaxial and optical? Bits are bits. Not to pick fights, only to increase my knoledge
     
  7. TerryS

    TerryS Forum Resident

    Location:
    Peyton, Colorado
    Keep in mind that even though the Latency software reports that a system is suitable or not for real time operations, it is just 'guessing' since it doesn't know what real time operation you are planning to run. It can only determine the latency in milliseconds. Depending on what you are trying to do, the latency time may be short enough, or it may not. An audio sample rate of 96kHz would require half the latency time of a 48kHz audio waveform. And 192kHz would require half of that. So what is the latency check program calling good enough?

    I used the latency check programs (there are a couple I found) as a rough indication as I tried shutting down various processes, but to actually confirm that the latency time I ended up with was good enough (or not) I actually recorded audio and examined it for drop outs.
    I would record about half an hour of audio consisting of the output of a signal generator (sine wave). Then I used ClickRepair software to quickly scan the resultant file and flag any drop outs. It turns out the drop outs look very much like small clicks or discontinuities in the waveform that ClickRepair could flag automatically.
    I often found that the first ten minutes or so of audio would be glitch free, but then glitches would start. I was never able to identify a single process that was responsible for the glitches, but that as I shut down more and more processes, I could run glitch free longer and longer.
    I thought about using a batch file to shut down processes when I was recording, or a dual boot system, but I found that even BIOS changes mattered. That's when I gave up on having a PC in the path at all.

    Terry
     
    Randoms likes this.
  8. Strat-Mangler

    Strat-Mangler Personal Survival Daily Record-Breaker

    Location:
    Toronto
    Exactly!

    There is such a thing as overanalyzing and looking for secret voodoo when there isn't any.

    Using ASIO or WASAPI bypasses the Windows mixer entirely, resulting in bit-perfect playback. If you own an external DAC, you output the data stream from the PC/laptop via USB or Optical. The signal can't be purer than it otherwise was by stopping processes or changing buffer settings or anything that is proposed. That is the nature of digital. Changing DACs will have an immediate and impactful difference, though.

    EDIT : By the way, USB 1 is fast enough to carry even hi-res FLAC data stream so switching to USB3 won't improve anything, either.
     
    Last edited: Jan 6, 2018
    nosliw likes this.
  9. TerryS

    TerryS Forum Resident

    Location:
    Peyton, Colorado
    I tried both ASIO and WASAPI. ASIO was definitely better, but both would drop samples depending on what processes were running. This was using USB2 and an external ADC (Focusrite Scarlett 212). If the issue doesn't exist, why is there so much info out there on how to correct it out there? Look into DPC latency.

    Troubleshooting DPC latency

    Terry
     
    Randoms and CrazyCatz like this.
  10. Strat-Mangler

    Strat-Mangler Personal Survival Daily Record-Breaker

    Location:
    Toronto
    Because most users do not know how to ensure all components work flawlessly. A component or driver conflict can artificially create latency. That isn't fault of the manufacturer.

    I've had ASIO working perfectly on a 10-year old PC with zero drops in latency.
     
  11. TerryS

    TerryS Forum Resident

    Location:
    Peyton, Colorado
    I have no doubt that it is possible to get a system working 100% bit perfect. But it isn't a 'gimme'. Background processes the user has zero knowledge of can affect the ability of the system to keep up with real time processes. Dropped samples result when it can't. To assume that a PC will operate bit perfect in a real time system is optimistic. It might. It might not. Depends on the PC and the real time process.
     
    Randoms likes this.
  12. Grant

    Grant Life is a rock, but the radio rolled me!

    And data transmission through cables.
     
  13. TerryS

    TerryS Forum Resident

    Location:
    Peyton, Colorado
    The only way I can see that data transmission through cables would affect things is if they were bad enough that data got corrupted. If that was the case, then if you tried to save a file using that same cable, it would be immediately obvious that something was seriously wrong. I don't think USB has a retry mechanism, so I don't think a marginal cable would slow the transmission rate by forcing retries (but I'm not sure of that).

    Terry
     
  14. Grant

    Grant Life is a rock, but the radio rolled me!

    I was thinking about worn cables. yes, I know, either digital works or it doesn't, but bad cables and-or connector can corrupt the data if it still works.
     
    CrazyCatz likes this.
  15. TerryS

    TerryS Forum Resident

    Location:
    Peyton, Colorado
    According to Wikepedia:
    USB - Wikipedia
    USB uses a CRC to check the data for errors, so I don't see how a worn cable could affect the audio.
    "If the data is corrupted while transmitted or received, the CRC check fails. When this happens, the receiver does not generate an ACK, which makes the sender resend the packet."

    I confess I don't know enough to say for certain, but it seems like the system was designed to prevent that. It shouldn't allow corrupted data across the link.

    Terry
     
  16. ralf11

    ralf11 Forum Resident

    Location:
    Earth

    or if elec. noise got to the DAC

    BTW, never use a DAC inside a computer - get an outboard DAC and electrically isolate it if using USB
     
  17. Grant

    Grant Life is a rock, but the radio rolled me!


    I wasn't thinking of just USB. I was also thinking of internal connectors.
     
  18. Grant

    Grant Life is a rock, but the radio rolled me!

    Today's motherboards and PCUs are built better and better isolated. EMI isn't much of, or not an issue.
     
  19. ralf11

    ralf11 Forum Resident

    Location:
    Earth
    how did you test?
     
  20. Grant

    Grant Life is a rock, but the radio rolled me!

    Software audio programs. If there is EMI, you can see it that way, and use your ears. I have a well-built motherboard that is almost a decade old (Gigabyte). But, when I started getting tons of electrical noise in my sound, and the computer started acting up, the power supply was the culprit. So, as soon as I replaced it, the problem went away.

    There people on this thread who are poo-pooing the idea that a bad PSU is an issue, but it is a major factor in noise.
     
    CrazyCatz likes this.
  21. Mike-48

    Mike-48 A shadow of my former self

    Location:
    Portland, Oregon
    Terry, what I have not yet asked and I don't think you've yet said is: Were the increases in measured distortion associated with clearly audible artifacts? I'd expect dropped samples to cause clicks, for example.
     
  22. ralf11

    ralf11 Forum Resident

    Location:
    Earth
    I would never poo-poo a bad PSU
     
    Grant likes this.
  23. TerryS

    TerryS Forum Resident

    Location:
    Peyton, Colorado
    They were audible, but very brief. As I got the processes reduced further and further, they became fewer. It became tiresome to listen for them. That is when I developed the method of letting ClickRepair search them out and identify them. I could search 30 minutes worth of audio in a few seconds using ClickRepair.
     
  24. Mike-48

    Mike-48 A shadow of my former self

    Location:
    Portland, Oregon
    Interesting. Any latency issues in my tracks have been inaudible; however, Wavelab Elements sometimes marks a few spots as possible errors. I wonder if those spots are what you are talking about. I'll have to take a very close-up look at the waveforms some time soon.
     
  25. anorak2

    anorak2 Forum Resident

    Location:
    Berlin, Germany
    Not when it's broken or undersized for the computer. Then the whole machine will be acting up, so it's not a case of "it works". Of course replacing the power supply will make it work again. But if the computer works fine already and all components are adequate, changing the power supply can't make a sonic difference.
     
Thread Status:
Not open for further replies.

Share This Page

molar-endocrine