Help! LFO Crazyness. Cubase Download Included.

  • Running Cubase 4.5.2, Latest Virus OS, TI 1...

    Problem: LFO Changes it's perseved rate when adding more voices, load to Virus.

    Download Project

    I'm using LFO 3 set to max, Modulating Osc 1&2 Pitch at 50%. This creates a heavy detuned gaga lead on 1 square wave instead of using lots of unison detuned to achive the sound. I've got unison set to twin, 64 detune, 50% pan spread just to create some width. The idea is the modulation moves the pitch so fast that the key doesn't change but realy detunes your sound (best atempt at explaining).

    When adding more patches/voices, load to the virus, lfo3 seems to slow down, although it's still set to max and i'm not touching/editing that part, making it's modulation of the pitch noticable, dropping the lead out of key as the lfo flows.

    Can some one try the project on there virus, please. Would also be good for someone to try it on a ti 2 to see if theres a difference.

    In the project there's a simple few bars loop and 6 voices. Virus voices only. Track one is the gaga lead with the other tracks just been the same notes, different voices of the original part. When you unmute the other parts you will notice the gaga lead start to behave incorrectly.

    Any info or help here would be much loved.

    Thank you

  • I only had a minute to test it, but from what i could hear and by what you desribe it sounds like something that i'm pretty familiar with. For me it basically happens in two situations, one is with special combinations of unison and amplitude release. This is what i've sent to the support, you can hear the sound's totally going wonky when you add a longer release:

    No news regarding that one till today, the support said that they could reproduce it, but couldn't telly me any more yet (Mai 2010). The other situation is, and that seems to be what you are experiencing, when you have two or more sounds on either even or uneven parts producing a lot of load. According two the support the even parts are processed by one DSP, and then uneven parts by the other DSP. So when you have a lot of load on only even or uneven parts, ie on only one DSP, then the parts begin to fight for the available processor power for the LFOs, and then you'll end up with stumbling that we are experiencing:…

    You can test it in your project by muting all parts except for part one, three, and five, then toggle part three and five and you'll notice that this will toggle the problem. Try the same with the even parts and you'll notice that they do not have any effect on it.

    According to the support (October 2010) there is unfortunately currently no other solution for this problem apart from organizing your sounds on even and uneven parts so that the load is distributed more or less evenly across the two DSPs.


  • Thank you. Just downloaded your links.

    I'd heard about the odd and even parts a few months back and found that to be very useful. I think I mentioned my problem then in the thread too but never made a link.

    So I wonder if the lfo is actualy changing speed due to lack of power or if it's just apears that way..

  • Guess that's a question for the support... IMHO the pure lack of power beeing the culprit seems a little odd, i mean, even when i strip down the sounds in my example so that they are not using any effects, only a single OSC, etc., the LFO is still stumbling depending on the release of the sound in part 1. One should think that the DSP would be able to handle such little load, but who knows, maybe the LFOs sit so far at the bottom of the DSP priority list so that they get very easily interrupted.