Ad Widget

Collapse

Announcement

Collapse
No announcement yet.

Computer can't keep time?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    No... Windows' own audio system is hopeless and can easily screw things up, you need ASIO.

    Going out of sync is caused by dropped samples. Either the computer can't process them fast enough, or the hard disk can't deliver them fast enough, or sometimes Windows just drops some because it feels like it, which ASIO fixes.

    Recording something while playing back something else demands twice as much data rate, so that's usually when it breaks. The Windows audio system doesn't even have an API for properly synchronized full duplex, another vote for ASIO or DirectSound.
    "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

    Comment


    • #17
      I second Steve's suggestion. Stay the hell off of the internet with your DAW. Do a really good scan, then ditch your virus protection, firewalls etc... If you are not on the internet, you don't need to scan all of your files in the background all the time.

      If you must use this computer for other things, then set up a dual boot system with minimal OS options and programs for the DAW partition.

      Here in Houston, used computers are dirt cheap. You could get something usable for under $200 with the operating system. Get a KVM switch and you are in business.

      Comment


      • #18
        Yes, Steve A. and OC Disorder have good advice there. I've been playing with computer recording since Windows 3.1 and what they say more or less mirrors my own experience.

        I couldn't get anything to stay in sync until I bought a M-Audio Delta soundcard with an ASIO driver, and from then on it just worked perfectly from Windows 95 through to XP. Even on an old 600MHz machine it could easily record four tracks while playing back another four.

        When XP got end-of-lifed, I decided I didn't want to learn how to optimize Vista and Win7 for audio all over again, so I ditched my PC gear and bought a Mac. MacOS has an ASIO-like pro audio driver framework built in, and because it's a component of the OS, keeping it working reliably is Apple's problem, not mine. It worked straight out of the box and hasn't broken yet.

        Of all the DAWs I've tried, my favourite was Pro Tools M-Powered and the one I hated most was Cubase SX.
        "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

        Comment


        • #19
          I am still using XP for the same reason. My hobby is making music, recording and tinkering, not upgrading computer drivers!

          Comment


          • #20
            Originally posted by Steve Conner View Post
            No... Windows' own audio system is hopeless and can easily screw things up, you need ASIO.

            Going out of sync is caused by dropped samples. Either the computer can't process them fast enough, or the hard disk can't deliver them fast enough, or sometimes Windows just drops some because it feels like it, which ASIO fixes.

            Recording something while playing back something else demands twice as much data rate, so that's usually when it breaks. The Windows audio system doesn't even have an API for properly synchronized full duplex, another vote for ASIO or DirectSound.
            That would make sense if it were the computer going out of sync with something else. But he's talking about two tracks going out of sync with each other within the software. The summing of the audio will happen before it's ever sent to the buffer, so an ASIO issue is out of the question.

            Comment


            • #21
              Well, I've been here and had the same issue, and ASIO fixed it.

              What's happening is that, during recording, the audio coming into the software is getting out of sync with the audio that it's playing out for the musician to use as a timing reference. That can happen easily if the playback engine drops some samples now and again but the recording engine doesn't, and that's quite possible in Windows MME. (I've only seen this first-hand with audio code I wrote for Windows Mobile, but I've seen the symptoms on all Windows flavours, and they share the same audio API.)

              When you replay your new multitrack recording, the two tracks play out in sync, because they're mixed in software as you say. But the timing error already happened during the recording process.
              "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

              Comment


              • #22
                Originally posted by Steve Conner View Post
                Well, I've been here and had the same issue, and ASIO fixed it.

                What's happening is that, during recording, the audio coming into the software is getting out of sync with the audio that it's playing out for the musician to use as a timing reference. That can happen easily if the playback engine drops some samples now and again but the recording engine doesn't, and that's quite possible in Windows MME. (I've only seen this first-hand with audio code I wrote for Windows Mobile, but I've seen the symptoms on all Windows flavours, and they share the same audio API.)

                When you replay your new multitrack recording, the two tracks play out in sync, because they're mixed in software as you say. But the timing error already happened during the recording process.
                Oh, I'm not arguing against ASIO. Actually everything that's been said about DAWs is really good advice. I'm just not sure it will solve the problem in this particular instance. He mentioned that recording a pre-recorded track to a new track caused them to be out of sync, but unless he's looping it back in the hardware that shouldn't involve the output buffer anyway, at least I don't see why it would.

                Comment


                • #23
                  Originally posted by uvacom View Post
                  unless he's looping it back in the hardware that shouldn't involve the output buffer anyway, at least I don't see why it would.
                  Well that's what the guy must be doing, looping it back in the hardware. It's a common test that I've done many times for just this reason, it shows up any problems that you'd run into when multitracking later. You record a click track for half an hour, loop it out of the box and back in again, and record that second track for half an hour, and then look at the very end and check the clicks haven't wandered out of phase compared to the beginning. A properly set up DAW shouldn't slip by even one sample.
                  Last edited by Steve Conner; 04-01-2010, 06:21 PM.
                  "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

                  Comment


                  • #24
                    Originally posted by Steve Conner View Post
                    Well that's what the guy must be doing, looping it back in the hardware. It's a common test that I've done many times for just this reason, it shows up any problems that you'd run into when multitracking later. You record a click track for half an hour, loop it out of the box and back in again, and record that second track for half an hour, and then look at the very end and check the clicks haven't wandered out of phase compared to the beginning. A properly set up DAW shouldn't slip by even one sample.
                    Steve:

                    Bingo! If you reread the original post it sure sounds like that was exactly what he was doing:

                    "What I did to test this was record a song played by my computer and make it track 1. track 1 was then recorded to track two, and at the end of a 7 minute song was often 7 or 8 seconds behind track 1. This obviously makes recording impossible."

                    It sure sounds like he was taking the output of his SB X-Live card and routing it back to the input.

                    Speaking of SoundBlaster I have yet to see one of their cards that I would recommend to anyone for a DAW rig. Not that they aren't out there, just that I haven't seen them. An e-buddy was making 24 bit vinyl rips using one of the current model SB cards and when I looked at the waveforms in Audition 3.0 there was absolutely nothing over 25khz. (With 24bit/96khz vinyl rips you usually get some content all of the way up to the Nyquist point of 48khz- especially if you have a scratch.)

                    Steve Ahola
                    The Blue Guitar
                    www.blueguitar.org
                    Some recordings:
                    https://soundcloud.com/sssteeve/sets...e-blue-guitar/
                    .

                    Comment


                    • #25
                      Oh god, don't even get me started on resampling, I hate it.

                      To cut a long story short, Soundblaster cards are famous for running the converters at 48kHz all the time, and resampling inside some custom chip to whatever rate the computer asks for.

                      And nowadays many OSs will do the same even if the hardware doesn't.
                      "Enzo, I see that you replied parasitic oscillations. Is that a hypothesis? Or is that your amazing metal band I should check out?"

                      Comment


                      • #26
                        Well, then I must concede that we've probably found the culprit - or rather, you guys did quite some time ago and I over-analyzed the problem just to arrive at the same conclusion.

                        SB cards definitely suck. I remember the older ones advertised 96kHz but as you said Steve the internal DSP processes at 48kHz no matter what. There was actually a class-action lawsuit against Creative for this.

                        Comment


                        • #27
                          Computer can't keep time?

                          Well, I agree with Steve Conner that you really have to replace your old computer with faster and/or much better sound card

                          Comment


                          • #28
                            You don't need a new computer, just change your sound card and upgrade your RAM (to at least 2GB) and you will see that the quality is increased and the problems will disappear. It worth to invest in it.
                            Product Development Services and Electronic Circuit Design

                            Comment

                            Working...
                            X