Vienna Symphonic Library Forum
Forum Statistics

182,049 users have contributed to 42,201 threads and 254,656 posts.

In the past 24 hours, we have 2 new thread(s), 10 new post(s) and 43 new user(s).

  • VEpro 6 External Midi Latency

    Hello,

    I know there is a warning that external midi is "experimental", however for various reasons, I have to use MoL to send to the VEpro server. I bought VEpro6 update in order to see if the midi latency was improved but alas no difference.

    [url=https://www.dropbox.com/s/y3txea2u8u11qta/VEP%20Latency.mp3?dl=0]Here[/url] is what I get. The click is generated from logic, the piano is playing in bidule receiving the same midi signal, and the pizzicato strings is from VEpro receiving the midi but returning the audio via the network.

    That's quite some latency. I am not generally bothered by latency but this maybe a touch extreme even for me. :)

    Is there some way around this? Why is there so much latency involved when the server version receives it? The standalone behaves like bidule - nice and tight.

    FYI I cannot use the plug-in midi:

    - cc7 is captured by the SI object in Logic. I am staying with LP9 for serious work LPX (where this can be switched off) has too many problems for me that affect my workflow.

    - I need to retain backwards compatiblity. It becomes too complicated to restructure my template so dramatically and I have certain workflows optmised around external midi.

    But I don't see why there should be so much latency on receiving MoL. Any ideas?


  • Well I have heard back from Marnix Veenenbos at support, but his answer doesn't make a lot of sense. Is there anyone else out there that can help?

    Basically he is saying that because I am not using the VEpro midi, there can be no latency compensation. Well, to be frank, that's not telling me anything I didn't know. Of course I am not expecting latency compensation!! I don't get latency compneation in the standalone modes either, but I don't have THAT amount of latency.

    A further curiosity is that when I send VEpro midi to dummy instrument within an instance, the MoL midi is suddenly perfectly in time. That can't be explained by lack latency compensation on the MoL side.

    What I am doing is setting up a track in Logic outputting crochets to each of the SI's in the environment. I have a dummy instrument in each instance in VEpro6 server. Bingo...all the MoL stuff is in time.

    And, once it is in time, playing live with the sequencer stopped works fine as well.

    I don't agree that this huge latency should be "expected". If it was normal that there should be over 500ms latency then forcing VEpro6 to play back in time using a dummy region should be a bug!