• Register
  • Help
Page 1 of 2 12 LastLast
Results 1 to 10 of 16

Topic: Is 160 UNIQUE voices really possible??

Share/Bookmark
  1. #1

    Is 160 UNIQUE voices really possible??

    We were discussing this, Thomas J, Maarten and I on the IRC #MIDI-Mockup (Efnet, be sure to drop by) channel tonight. I was talking about getting MSG32 crashes like so many others. I notice this happens when GS is really getting stressed and beginning to have weird note dropouts and hickups.

    It is not a CPU problem. When GS does 160 voices it uses about 40% CPU. It is not a HD TRANSFER problem, new Barracuda IV harddrives, UDMA 100. It is not a PCI bandwidth problem, using the SiS chipset which has the best PCI bandwidth.

    So what is it? Well our conclusion, and pardon me if I am naive that we have found a reason which nobody else seems to have thought of, was that it boils down to HD seektime. If your HD, in addition to streaming the data to RAM, has to find 80 different locations on one drive (or two, in my case), it will use at LEAST 10 ms on each seek - probably more, because the samples are spread out across the drive(s) - which amounts to 800 ms. of seektime per buffer it has to read. I am not sure how big the buffers GS reads are, but I think somewhere around 32k or 64k - which is less than a second of data. So each second 800ms will have to be used on seeking alone, leaving 200ms. for the actual reading.

    Maybe I am not taking some things into consideration here, but if I am not far off, this should make it impossible to read 160 unique voices. If you use sample repetitions across channels or use the same short samples many times, this will be less of a problem, since the data will be cached in the VCache (or where else), which is also why, in SOME performances I have no problem getting 160 voices all the time, whereas in others I begin getting into trouble and crashes around 140-160 voices.

    Does anyone have a view on this subject?

  2. #2

    Re: Is 160 UNIQUE voices really possible??

    Simon,
    Hi. Interestign ideas. I\'ll have to think about that a bit. A couple of things come to mind:

    1) 10mS seek time isn\'t a hard number. It changes based on where you are on the disk. (Might be shorter or longer.) Anyway, possibly some performances stress this harder? Or stress it less? (Based on where the samples are on the drive.

    2) Independent of how large the read request is from GSt, you can effect this greatly based on how you\'ve fdisked the cluster size your drive. (A 4K cluster could be 8 seeks for a 32K request)

    3) Are all your samples on a single drive? This would seem fit the model you propose, but would imply that spreading samples across a couple of drives might help. (Horns on one, piano on a second.)

    Anyway, interesting thoughts.

    Take care,
    Mark

  3. #3

    Re: Is 160 UNIQUE voices really possible??

    Not to crack wise, but remember that bumblebees can\'t fly, either...

    Which is tosay there are predictive lookahead algorithms, buffer management tricks, all sorts of things Tascam/Nemesys can use to get the track count up there.

    Any sufficiently advanced technology is indistinguishable from the Supreme Court\'s decision to award the election to Dubya...

    Dasher

  4. #4

    Re: Is 160 UNIQUE voices really possible??

    Simon
    A couple things come to mind reading your post.
    You should be able to adjust the buffer cluster size , try different settings and see if that makes a difference.
    I have a Darla 24 and the setting is under my driver/mixer settings icon.
    And as already mentioned defraging helps keep your drive running at peak performance.
    Also the sis chip has the best bandwidth between the chip sets , but once you factor other elements in the chain one the current sis equipted motherboards they are not always the fastest.
    I read a few reviews comparing sis , kt 333 etc.

  5. #5

    Re: Is 160 UNIQUE voices really possible??

    One thing about the fact that you never move your GIGs around, so defrag is not an issue: while this may not really address your question, Windows is always moving files around on the HD. Even if you don\'t move it, Windows does all kinds of optimization that causes blocks to be rearranged. Which is basically why things can get corrupted even if you never, ever screw with anything.

    This could actually support the need to answer your question. It might also help answer it, inasmuch as there\'s prolly nothing you can do to keep Windows from constantly shifting blocks around.

    Best,

    bt

  6. #6

    Re: Is 160 UNIQUE voices really possible??

    Originally posted by Bolt Thrower:
    One thing about the fact that you never move your GIGs around, so defrag is not an issue: while this may not really address your question, Windows is always moving files around on the HD. Even if you don\'t move it, Windows does all kinds of optimization that causes blocks to be rearranged. Which is basically why things can get corrupted even if you never, ever screw with anything.

    <font size=\"2\" face=\"Verdana, Arial\">Sorry, but that is simply not true....

  7. #7

    Re: Is 160 UNIQUE voices really possible??

    Originally posted by Kenn159:
    Simon
    A couple things come to mind reading your post.
    You should be able to adjust the buffer cluster size , try different settings and see if that makes a difference. I have a Darla 24 and the setting is under my driver/mixer settings icon.

    <font size=\"2\" face=\"Verdana, Arial\">Cluster size on HD\'s have nothing to say here. I am talking about the buffersize that GS uses. GS reads in XX kb at a time no matter what. I dont know how big that buffer is, but it is obviously something that we have no control over, and is coded into the WVSTREAM device. Buffer size on your soundcard has nothing to say here either.

    [quote]

    And as already mentioned defraging helps keep your drive running at peak performance.

    [QUOTE]

    I hardly ever delete or move anything around on my GIG drives, so this is not an issue here either.


    Also the sis chip has the best bandwidth between the chip sets , but once you factor other elements in the chain one the current sis equipted motherboards they are not always the fastest.
    I read a few reviews comparing sis , kt 333 etc.
    <font size=\"2\" face=\"Verdana, Arial\">What \'elements\' are those? I am specificually talking PCI bandwidth - an overall system performance is usually benchmarking the CPU - which has no say here. The fact is that SiS has more than PLENTY of PCI bandwidth to not make the PCI transfers the bottleneck. Now other new chipsets are following suits, KT266a, nForce etc.
    Maybe KT333 has a bit more memory bandwidth than the older SiS 735 chipset, but that is not the bottleneck either. So it has to be the harddrives or GS code.

  8. #8

    Re: Is 160 UNIQUE voices really possible??

    Ya I frequently defrag my gig files..
    Doesn\'t seem to do anything, but I just do it anyways.. It\'s just a big wall of blue right now.. Hmmm 65% full on a raid 0 array.. What am I gonna do once it\'s full??
    Bri

  9. #9

    Re: Is 160 UNIQUE voices really possible??

    Talking to a tech at Echo Audio, he stated that Sis chipsets do NOT work well with any device using Motorola DSP. I hope he\'s wrong, cuz my new Giga rig uses the Sis 650.

    But if he\'s right, could this be part of your problem?

    Dasher

  10. #10

    Re: Is 160 UNIQUE voices really possible??

    Just out of curiosity, all of you guys that are offering solutions and ideas....do you guys have actually documented getting 160 voices in your set-ups?

    I mean \"really absolutely 160\" or \"well, I think I got 160\"?

    I\'m just curious!

Go Back to forum

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •