Setting a really complete single rackspace

Hi, I am a keyboard player. I play mostly in church. I use two keyboards. Before I use GP and vst plugins, I use technics and roland keyboard and I am used to switch from one sound to another quite often according to situation and songs. Like, from an upbeat disco song with brass section, to RnB with B3 sound, then playing lead melody in ballad with saxophone, then drastically playing some orchestral arrangement (strings, horns, woodwins) for a majestic hymns.

In GP, I made one single rackspace with all likely to be used sounds in it, plus necessary effects. Currently I have a grand piano, electric piano, strings, pad, brass section, B3, lead synth, Sax, Flute, and also a bass (in case the bass player doesn’t show up).

I put on/off switch and volume knob plus some other widgets for each sound. All in a single rackspace.

My question? Is it a good practice? Does GP designed to do this kind of complicated things? Actually I did this because many times, the situation and song is unpredicted. The leader can suddenly sing an out of list song. It will be frustrating if I need to make one rackspace for each single song.

Any ideas? Thanks.

I am using variations to switch between sounds used in a song, so only 1 rackspace per song.
Technically you can just use 1 Rackspace and many variations.
You have to mute/unmute your plugins in the variations (but have an eye on haning notes!)
Or you use the filter plugin to filter out NOTE ON messages, this way you avoid hanging notes.

But because of complexity it is much better to use multiple rackspaces and with the SetList Mode you can define your songs and songparts and can reuse Rackspace Variations.

Many possibilities, but just 1 rackspace for all you songs it not the best solution.

What’s “best” is what works best for you.

Back in the old days, when we all had two or more keyboards that each made their own sounds, I think most of us got used to changing patches on each keyboard individually for whatever sounds we wanted at the time. The only way to really replicate that (that I can think of) is jamming everything into one rackspace. I tried that initially, and it can work, but ultimately I moved to a different approach.

Right now I’m using rackspaces designed around the “main” sound (e.g., piano, Rhodes, B3, synth) and variations for the “auxiliary” sounds I might use with that. For example, I have three different VSTs I might use for the bass. I can switch among them with widgets that put the VSTs on bypass and/or midi in blocks and filters.

What’s optimal really depends on your individual situation. For me, where I have a sample-based VST that takes a long time to load and sucks up a lot of memory I try to minimize how many instances of that I have. That initially led me to that “everything in one rackspace” approach, because I was using Omnisphere for my Rhodes and the Yamaha piano (and some of the bass sounds). Putting those in five or six rackspaces made loading times too long for my tastes.

Then I switched to using Pianoteq for all my piano and Rhodes sounds (and Rhodes piano bass), and Modo Bass for my electric bass sounds. Thus modeled instruments vs. sampled instruments. So now I can load 20 instances of Pianoteq in less time and less memory than one instance of an Omnisphere piano.

I still use Kontakt for guitars (Orange Tree) and stick those in one rackspace with variations to switch between Fender, Strat, clean, dirty, and effects options. They take way too long to load for me to set up different rackspaces for all the permutations I might use.

Off the top of my head I’d say I use about 8 different rackspaces just for piano (all with Pianoteq). Most of them with only have one or two variations. I found it more practical to have separate rackspaces for Steinway D “pop”, Steinway D blues, Steinway D classical, Yamaha CF, Yamaha CFX, etc.than to have 50 widgets to control every parameter that changes between variations.

It’s different for my other keyboard sounds. I have all my Rhodes variations in one rackspace, all my B3 variations in one, all my Vox Continental in one. In each of those I have about 3 other VSTs that I can switch between using widgets/variations for the different bass or auxiliary sounds (that in the old days I’d access by changing patches on the other keyboards). That works well for me, but if I were setting it up again I might just skip the variations and have different rackspaces for something like John Lord Hammond, Green Onions Hammond, and maybe some more generic clean or dirty Hammonds.

In the end, “everything in one rack” didn’t work for me, but “separate rack for every combination of VSTs” didn’t work either. The nice thing is GP gives you a lot of flexibility to play around and figure out what works best for you.


So, may be it will be better to make a list of possible combinations, for eg. Piano + strings, Piano + B3 + Brass, Piano + Pad + Sax, Woodwind + Brass + Strings, etc, then make one rackspace for each combinations with necessary variations. Is it right? Will it make cpu load lighter and give me possibilities for better buffer size/latency?

And, when I use same particular vst on several rackspaces, for eg. a grand piano vst with same settings on 3 rackspaces, will GP load 3 instances of that same vst or will GP load it once and use it whenever each one of those rackspaces is called?

I would like to point out that if you really want to individually control two keyboards - you may want to actually use two instances of Gig Performer. You will get much better overall performance and ability to individually change sounds on each keyboard. Just mentioning this as this option with multiple instances gets often overlooked because there is no other software that can do this properly as far as we know.

1 Like

Yes, this is probably the more valuable solution to do so. And GP seems to be the only plugin host which is able to work like this. :ok_hand:

Please, think of a way to switch from an instance to the other (i.e. here from a keyboard controller to another one) when using GP in full screen mode and without computer keyboard, and GP will be the ultimate solution for jamming or for any situation where it is difficult to predict which sound combination you need on different keyboard controllers (I am thinking of a GP script function that could switch to any other instance).

Yes, that’s an excellent point. I tried that approach at one point, but then moved away from it.

Per David-san’s point above, it would be a “nice to have” if there were a structure to run multiple rackspaces in parallel or series within GP. A second rackspace in parallel is essentially what a separate instance already achieves. Add onto that the idea of an effects rackspace that sits after them in the sound chain. (So I don’t have to put the same half dozen effects VSTs in all my racks.)

How to pull it off visually, I’m not sure. Not sure how to tie it in with songs and setlists either. But that’s why you guys are the professionals.

Using 2 instances of GP, won’t it bring a lot more burden for the CPU? Will it work smoothly? And can I use the same audio interface?

With 2 instances each instance runs on a separate core.

1 Like

I use the RME drivers on Windows and there is no need of a second audio interface.

Using two instances of GP will generally increase your total VST capacity. For most people on most setups it’s probably not going to make a big difference, but if you find yourself running into clicks and pops with all your VSTs running in one instance you might get better performance if you can split them into two instances.

What audio interface are you using? Most leading brands have “multi-client” drivers, meaning multiple programs can talk to them at the same time. Some don’t, though, which would force you to use a virtual device wrapper if you wanted to run multiple instances (or do something like run GP and a DAW at the same time).

I use presonus studio 1810c. I don’t know if it will work with two instances of GP.

My Presonus Studio 24 works multi-client. The 1810c uses the same driver package, so I would be shocked if it didn’t support multiple instances of GP.

It is not a big deal to test it. And it is worth the try… :wink:

I was heavily using a Studio 1810 (which only differs from the 1810c by the color of the body and the physical USB connector) and multi-client ASIO worked like a charm :slight_smile:

Thanks everybody. I am using Lenovo Ideapad 320 with 500gb SSD, 8gb RAM, i5 processor. Is it enough for multiple instances? I want to try it soon though… I am curios. May be tomorrow. The only other problem I think of is that with only one instance of GP running, I can see the whole rig in a glance. With 2 instances, one for each midi controller, I can only see one view ( one GP ) at a time.

If you’re running Windows 10 you can have split screens.

I doubt there’s much extra overhead running two instances of GP vs. putting all the VSTs into a single more complicated instance.

From a processing power standpoint, you’ll probably be able to to get to lower latency using two instances. The 8 GB of ram could present a challenge if you use several big sample libraries.

Getting two instances on the same screen could be a challenge on a laptop. On my desktop I put them side by side and there’s plenty of space. On my laptop but I’d imagine it would be more readable with one placed above the other, but I haven’t tried it.