@hose said:
Will it be possible to create a "duplicate channel" feature?
This is a good point, I'll see if I can squeeze it in before the release.
194,424 users have contributed to 42,920 threads and 257,967 posts.
In the past 24 hours, we have 4 new thread(s), 11 new post(s) and 74 new user(s).
Thanks for your answer! Does the OS typically schedule the threads evenly, or does it leave much to be desired?
The reason for my question is that I have some very CPU-intensive kontakt instruments. I am considering running each instrument in a separate VE instance to try and "force" the OS to schedule each to a different CPU core.
Most of the time, the OS is really very good at scheduling threads. As long as you (as a developer) try to keep a sane threading model in your application - you can usually trust the OS to do the proper scheduling, as long as you have thread priorities set correctly. There are of course times when thread context switching can cause performance losses, but to my experience - the OS scheduler is remarkably good and quick. I suppose you are using Kontakt in a sequencer host today, I would recommend to have the sequencer handling any threading or affinity.
When it comes to threading in audio applications, and mixers in aprticular - there are several things to consider. At one or several points the threads need to be syncronized, for submixing, bussing, sends etc, and this can create some issues if it is not done properly. I know some people, running multiple audio input objects in Logic, are having quite some issues with most of processing plugins ending up on the same mixer thread, overloading even an 8-core machine with only a few plugins.
Thanks for your answer! Does the OS typically schedule the threads evenly, or does it leave much to be desired?
The reason for my question is that I have some very CPU-intensive kontakt instruments. I am considering running each instrument in a separate VE instance to try and "force" the OS to schedule each to a different CPU core.
Most of the time, the OS is really very good at scheduling threads. As long as you (as a developer) try to keep a sane threading model in your application - you can usually trust the OS to do the proper scheduling, as long as you have thread priorities set correctly. There are of course times when thread context switching can cause performance losses, but to my experience - the OS scheduler is remarkably good and quick. I suppose you are using Kontakt in a sequencer host today, I would recommend to have the sequencer handling any threading or affinity.
When it comes to threading in audio applications, and mixers in aprticular - there are several things to consider. At one or several points the threads need to be syncronized, for submixing, bussing, sends etc, and this can create some issues if it is not done properly. I know some people, running multiple audio input objects in Logic, are having quite some issues with most of processing plugins ending up on the same mixer thread, overloading even an 8-core machine with only a few plugins.
Thanks again for your answer!
As you have imagined, I'm running kontakt in Logic Pro 8. To my amazement, Logic frequently runs many Kontakt instances all on one CPU core, while other cores idle. So, I am stuck using PC servers for the moment.
It is my hope that VE Pro will be a way to force each Kontakt instance onto a different core, but still be able to mix inside logic -- so I can get rid of my servers and run everything on my mac pro.
Do you think my hopes are reasonable?
AFAIK there is no requirements to run a VI in VEPRO to make it work, and it has been hinted that you don't even need to own a VI to purchase VEPRO when it becomes available so in short the answer should be yes!
The current model for VE3 network is that each purchase gives you three licences, so you can have three slave machines connected to 1 DAW (or two if you are going to run VE3 locally). You need a dongle for each slave computer.
Best
Tim
Hi everybody,
we are focussing on VE PRO for the next month, and I believe that it cannot be released before mid/end of may, realistically.
We need to make sure that VE PRO is rock stable - for all of you [:)], and that includes some serious testing and fine-tuning.
Thanks for your patience and enthusiasm!
Best,
Paul
Given my recent travails with updates from other libraries - take your time. [:'(]
welcome robjohn33,
no, you don't need to have a Vienna Instruments Library registered for purchasing VE PRO, however it will take another month or so until it will be released ....
kompakt is a player for certain libraries - from what i've heard it will run as well as kontakt
christian
Absolutely. However, I AM presuming that by streaming older sample libraries from the G5 and relying upon my MacPro for all the CPU grunt work, things should run effectively? (I guess unless the actual Vienna Ensemble Pro program itself uses a great deal of the G5's resources?) - robjohn@cm said:
however, please don't expect more from your G5 with VE PRo than the machine can deliver without it ... christian
The MacPro will not be doing the CPU load if you put it on the G5, the G5 will be doing the load.
Whatever machine you host VE Pro from, you will be loading the samples there. Therefore it will be the machine that will take the load. It will then pipe all the audio over it's ethernet port to the DAW machine. But the actual processing is done AT the machine that is hosting the samples.
Did you watch the VE3 demo? It will clear this up for you. However I re-read your post and it seems like all you want to do is sample your older libraries from that G5 using VE3 Pro.
What I would consider doing is the idea of making your G5 your Logic machine (or whatever DAW you use) and then have the MacPro be your VE3 machine since it will be the one that needs all the horsepower.
I have a macbookpro with only 2GB's of ram. I can run an entire orchestra from my Vista 64-bit VE3 machine. Since it is a power house and does all the processing, my macbookpro takes almost zero beating at all. And my 64-bit machine has lots of head room left. And I mean lots.
Maestro2be
@cgernaey said:
What I would consider doing is the idea of making your G5 your Logic machine (or whatever DAW you use) and then have the MacPro be your VE3 machine since it will be the one that needs all the horsepower.
By all means test this theory, but I personally disagree in the idea of using the G5 as the main Logic machine & the MacPro as the sample slave. Playing sample libraries like VSL does require some CPU, but the actual bottleneck for streaming samples is not CPU, rather RAM & Disk speed. The VI player is VERY efficient, and so is VE. In fact, I am ALMOST able to stream an entire orchestra from my Dual 2.5GHZ G5 alone! However, 2 instances of Amplitube (guitar amp FX plugin) brings my G5's CPU to its knees. If you plan on running MIR or any other high processor reverbs, you will only be able to do this on a MacPro. The G5 processor just cant handle it (MIR) and/or the algorithms are not being coded for PPC (see the new vienna reverb in VS, which unfortunately is intel only!) BTW, I use a dual 2.0 Intel mac mini as a VE slave, and VE is running at 25% CPU, yet my poor FW400 external disk & 2GB RAM are maxing out, further confirming that my bottleneck is not CPU, rather disk speed/RAM. <br><br>
Thus I personally recommend using the G5 as a sample slave (max out RAM & add additional disks via SATA card), route the separated audio signals into Logic via aux tracks, and use the Mac Pro as your CPU machine for reverb/FX. At least this is my plan.