Vienna Symphonic Library Forum
Forum Statistics

181,889 users have contributed to 42,190 threads and 254,623 posts.

In the past 24 hours, we have 3 new thread(s), 17 new post(s) and 57 new user(s).

  • How many instances of the VI can you run performance wise?

    I was just wondering how much CPU power an instance sucks out of
    your PC.

    Could you give an example of how many instances can be run simultanously in a sequencer, eg. Sonar, on a PC with say a P4 3Ghz ?

    Given the fact that one VI equals one instrument, you obviously need a lot of instances to run a symphony orchestra.

  • bump

  • It's a bit of a "how long is a piece of string" question. I do recall though somebody mentioned CPU was about 1-1.5% per instance on a machine not dissimilar to yours.

    Then of course it depends upon the pathces you are loading, if you use the ramsave feature and of course how many VSTi's your host supports - e.g. CubaseSX3 has 64.

  • Thx Timkiel - yes I know its a indefinite question, as it varies with
    the machine, os , setup etc.

    Anyway 1-2 % per instance sounds nice.

  • Also, bear in mind it's been quoted that the new VIs might load as many as twice the samples you could get into GS3 on the same PC. So you should need less instances anyway.

    Colin

  • No - the point is that an instance will only play 1 instrument - its not multitimbral/multichannel - so to have a symphony orchestra standing by, you'll need at least 15 instances for a very simple setup ( 4 woods, 4 brass, 5 strings, 2 perc )
    So number of instances does count

  • I'm just curious why that's a problem. To me it seems a lot simpler to deal with one MIDI channel for a solo violin than 25.

  • Hi Nick
    Its not a problem, its nice as you said - but it puts into focus how many instances you can run at one time. If you could only run say 2-3 instances
    before the CPU got choked - its not much of an orchestra you could have,
    more like a trio [:D]

  • I'm with Nick on this one. I can't see the problem in having one multiarticulate channel as opposed to multiple tracks. And applying premix effects to one track as opposed to many must surely decrease the effect plugin tax on the system. One preverb plug as opposed to 25? And that's before we know what MIR will do for us.

    Fact is, once anyone is practised in using the VI instinctively, the music time will go up, and programming time will plummet. I can envisage writing lines continuously, and a little end tweaking, instead of writing lines and cut and pasting, or rerecording articulations one at a time. If a VI cuts the time by 50% or more, then a lot more music will get written in less time. (IMHO). And with ram optimisation, you just keep going until you finish. (Dependent on computer capability)

    I think that's a great idea.

    Regards,

    Alex.

  • last edited
    last edited

    @Nick Batzdorf said:

    I'm just curious why that's a problem. To me it seems a lot simpler to deal with one MIDI channel for a solo violin than 25.


    This is exactly the point and I too am mystefied by groanings hither and yon regarding this very positive change in ease and work flow.

  • last edited
    last edited

    @Sapkiller said:

    No - the point is that an instance will only play 1 instrument - its not multitimbral/multichannel - so to have a symphony orchestra standing by, you'll need at least 15 instances for a very simple setup ( 4 woods, 4 brass, 5 strings, 2 perc )
    So number of instances does count


    My point was that users are currently utilising up to 60 or 70 instances of EXS to create their orchestra, with single instruments taking maybe 16 channels, each with a different articulation. The new VIs do away with this by combining an instrument or section onto a single channel, just like Nick & Alex said.

    15 instances has to be more efficient than 70 - no?

    Colin

  • Hi Colin
    Yeah - with information I have now ( 1-2% ) its clear that you will be able to run a lot of instances - but I was'nt sure, for all I knew it could have been a CPU hog calculating all those predictive speed things and maybe using sample morphing techniques for the crossfades ( which it clearly does'nt use with that low a CPU consumption ).

    Anyway I'm all for the one instrument - one channel approach, makes life much easier.