Ok folks, I am admittedly new to all this and I am still trying to figure out whether to run Gigastudio 3 or Kontakt 2, and what kind of rack mount PC to build for whatever I choose to run.
So I need some clear explanations and opinions. I will use this machine mainly for live sample playback and sequenced sample playback, though for the sample playback I figure I could probably render it to audio and then just load the audio into my keyboard (Triton Studio 76). But for the live stuff I am confused. I will not have time to load samples during a gig. My band pretty well plays from one tune right into the next and doesn't want dead time while I screw around with the DAW. In addition, we do not use set lists. We call them as we go so I don't know what's coming up in advance.
My Triton Studio only has 1 midi out. I also gig with an old Korg T2, which surprisingly enough, has two midi out ports (go figure). I read that with Kontakt you can preload 128 instruments, assuming you've got the horsepower and the memory, and then select them by using program change commands.
So here are my questions:
1. Can I reference all those instruments with only one or two MIDI output ports? If not, what does it take? Do I need keyboards with lots of physical midi outs? There aren't too many of them around?
2. Pertaining to question number 1, do I need an audio interface card on the DAW that has lots of physical midi in ports? If so, what audio interface cards would you suggest?
3. If I am not actually using Sonar 4 PE at the gigs, would I still want to run Kontakt 2 as a VST plug-in? How would that work? Would I want to use Kontakt 2 as a plug-in at home while doing the production work and then use it standalone at gigs, if the sequences are really running from the Triton Studio (with audio tracks as part of the sequence)?
4. I have read that Kontakt now supports multiple processors. Does that mean it supports both dual core processors and dual processor systems? Does it really take advantage of them? Do you really get anything like twice the performance, CPU wise, by having a dual core processor with Kontakt?
5. How much RAM will actually buy you improved performance? Can Kontakt 2 really do anything with more than 2Gb? How about Sonar 4? I will more than likely be running an Athlon processor (maybe an X2 dual core or an FX-55 or FX-57 single core) and will be running Windows XP. I will buy however much RAM the software will do something with.
As I said earlier, I am trying to accomplish two things: one is to decide on Gigastudio 3 vs. Kontakt 2. Yeah, I know, it's an ongoing debate which makes it hard for guys like me that are just trying to get into soft sampler work for the first time. Secondly I am trying to determine what kind of PC to put together (rack mount for live gigging) to support all of this. I don't want to build something that will be practically obsolete in a year or two. I know that new stuff comes out all the time and I can't do anything about that. But I want all the processor power, RAM, and disk drive performance that I need to really have these products (Sonar 4 PE / soft sampler) run at their maximum efficiency. I am not too concerned about the price of the DAW but I only want one machine. I'm not interested in configurations that have me putting together 2 or 3 PC's into a congfiguration. Oh yeah, and I realize that performance depends a lot on disk drive configuration and performance. I was thinking of something loike one drive for the OS and software products, 2 drives (RAID 0) for the samples, one drive for audio, and maybe one drive for the Windows pagefile.sys. I will probably set everything up to support SATA II, not so much because I think the additional bandwidth capability will buy me anything (I know the limiting factor is the data transfer rate of the disk drives) but more to use the NCQ (Native Command Queuing) feature, which I believe will improve performance.
I really need everyone's help. If you would please remember, if you are nice enough to respond to this thread, that I am a beginner and need things explained in some detail to thoroughly understand. On the hardware side I am experienced. I am in the business. But on the software side this is all new. For 30 years I have done everything on hardware work stations (like the Triton).
Thanks in advance to everyone willing to take some time with me on this.
Basically. K2 will do all the things your decribing. Each Midi port (theres 4 in K2) can hold 128 channels and you can send program changes from your Triton to scroll through them. Unless your using the convolution in K2 it doesnt take up a ton of CPU. 2 or 3 GB is enough. You can run it standalone. OR i highly recommend, Brainspawn forte to host it along with all yur other VSTis.. its made for Live playing where you can set up "scenes" and switch between them on the fly..K2 is perfect for live because the FX in it are neat...GS3 is also good cause it will host your FX but limited in that is doesnt always play nice with the rest of your VSTi`s.... If i were just running a Piano then maybe GS3 but for everything else K2... My Opinion,, Rich
You need to decide how many samples you plan on loading to play a gig. There is no time to really load most instruments real time with either Giga or Kontakt.
The other thing that makes it a little trickier is the fact that you want to use sequences. How many sequences are you going to use. Sonar usually loads the sounds when used as VST which will take some time. The other option is to use Kontakt as a standalone instead of as a VST.
GigaStudio may or may not work well on the same machine with Sonar. Tascam recommends running Giga on it's own dedicated machine.
I was thinking that for gigging I would use Kontakt (or Giga) as a standalone program and then use Kontakt as a VST or use Giga with rewire or the VST adapter when home creating the music. The band will have about 20 sequences in a night. Some of them overlap in sounds (i.e. brass section, string ensemble, etc). But many will be unique to a sequence as well. I would want to have as many loaded in the beginning of the night as possible. The whole series of questions in my original post was geared to figuring out exactly how to configure for the most preloaded samples to be accessed as well as how to get the most performance out of the hardware. I would love to have the questions answered.