Sampletank within MainStage: Memory issue (RAM)

[pmurri]

Hello everyone,
I'd like to use a blend of virtual instruments from Logic Pro / Mainstage and Sampletank. That's doable using Mainstage, BUT each and every ST patch weights at least 1.8GB, even if the patch itself is small (say 0.2GB). ST eats also a lot of CPU, but that's manageable, at least on my mac pro.
Mainstage offers RAM optimization options, by using concert level sounds, but that's not enough neither flexible... HELP!
Any ideas here?
p

[Robert2316]

I’m having the same issue with ram maxing out using mainstage. I’m also wanting to use a mix of different patches through mainstage. I see where there is a live option within the Miroslav editor. I wonder if that may reduce needed ram. Let me know if you come up with a solution.

Kronos
Mac book pro 16g
Focusrite
Ik philharmonic
Mainstage

[pmurri]

Hi Robert,

I found a workaround. Still sampletank is eating a lot of RAM, but I am using only one instance for different patches.

Here's the workflow:

1. Create a sampletank channel strip at concert level, select "Multi-Output 16ch stereo" (NOT just stereo), with no audio output (important)
2. Within that sampletank create a MULTI with all the instruments you need
3. Within sampletank MIXER route each instrument to a different stereo output (from "3-4" onwards). I also suggest to kill any efx you don't need (eg reverb), since they eat CPU and RAM
4. In mainstage now click on the "+" button that appears near the volume slider of the sampletank channel and each time you click it the outputs from ST will appear sequentially on the right side of the slider
5. within mainstage rout each newly appeared ST instrument to a specific bus as output (eg: strings to bus 6, brass to bus 7, etc). I suggest also to rename the strip with the name of the instrument.

Now when you create a patch in mainstage you can add an audio channel strip that has as input the bus you need. Say you want the ST strings, select bus 6. There you have a patch with a ST instrument not eating extra RAM.

Note: I found that with this setting the "concert level" instruments will always have a priority / replace other instruments in the individual patches. To manage this you have to go into the layer editor of each patch (non ST instruments) and select within the Layer Editor "override parent ranges > add or replace". In this way you can layer the ST concert level patches with sounds from other sources (eg Mainstage patches).

A bit complicated, but it works!
Let me know if you find other solutions.

Ciao,

Paolo

[Robert2316]

Paolo,

Thanks for your solution! Yes it is a bit complicated, but it worked fine for me and I'm just going to stick with this setup. In my case I'm just using 1 concert level setup and I've setup 30+ dummy patches which will change my tempos and any click tracks that I use. As of now everything seems to be sounding amazing and the Ram/CPU meter is staying at half or less. Thanks again!

Rob

[pmurri]

You're welcome Robert, glad to hear that it worked.

p

[fgrati@gmail.com]

Hi, I followed your example - works pretty well! Lot of different sounds with little memory usage. Only thing, apparently I cannot use split with this setup because at patch level they are audio tracks and not instrument tracks. Any idea?

[Robert2316]

One option you may try is to setup your split at concert level under a different instrument patch and then just mute/unmute as needed. I’m out of town but I’ll check into it next week.

[fgrati@gmail.com]

Hi, that might work in general - but I use split points a lot, and they vary from patch to patch. I saw somebody implementing a solution using the standalone version of Sampletank and using MS only to store layers and split for each patch, sending MIDI data for each channel separately to ST. But this looks quite prone to problems.