Bouncing Synths

[jkoseattle]

I used to have one EastWest Play instance for every instrument in my project (violins, Cellos, Basses...). Now I am trying a different way, grouping a whole section into a single Play instance (one for strings, one for brass, etc.) and assigning each instrument to a different Midi channel. The aim here was to simplify and also hopefully save on RAM by having fewer total synths. HOWEVER... I didn't really realize what I was giving up. I can't adjust envelopes for any individual instrument, only for the whole section. I figured I'd just do all that tweaking after I freeze to audio. But now I'm realizing that freezing presents me with just a single audio track. This won't do at all. What is the best practice here?

Is it possible to bounce individual tracks in the same synth to their own tracks? How?

Am I really not gaining anything by grouping instruments to single synths rather than just having a separate synth for each?

 

[Blogospherianman]

You can either set up separate outputs from the single instance of Play which also requires adding Synth audio tracks assigned to each of the multi outs.  This allows for freeze to make multi audio rendering.

Easier way is to highlight (select) the source midi track of each part one at a time along with your main synth audio track and bounce the parts individually to their own audio track.  (Bounce to track) Once bounced you can disconnect the synth (saving CPU)  in the synth rack and it will retain all of your patches should you need to rebounce a part or even a single line of a part by shortening your selection.  I’ve had issues in the past with un freeze not bringing back patch parts. Plus you can’t add layers to a frozen track.  Plus if you had edited timing or crossfades and needed to change something, un-freeze wouldn't keep your edits.  That’s my few cents. Peace!?

Edited June 4, 2019 by Blogospherianman

[jkoseattle]

Thanks, that makes sense. I'm getting confused by this whole thing though. I'm also trying to set up track templates and synth presets etc., and am having a lot of inconsistent results. So now i'm wondering, is this even worth the effort? My old way was 1 synth per instrument, all on Midi channel 1, using simple instrument tracks. Is there any reason I shouldn't keep doing it this way, when I might have as many as 30 synths going? Is grouping instruments into a single synth really getting me anything?

[dcumpian]

19 hours ago, jkoseattle said:

Thanks, that makes sense. I'm getting confused by this whole thing though. I'm also trying to set up track templates and synth presets etc., and am having a lot of inconsistent results. So now i'm wondering, is this even worth the effort? My old way was 1 synth per instrument, all on Midi channel 1, using simple instrument tracks. Is there any reason I shouldn't keep doing it this way, when I might have as many as 30 synths going? Is grouping instruments into a single synth really getting me anything?

The answer is...maybe.  Using multiple instruments per VTSi instance should reduce the overhead required by each separate instance. However, depending on how efficiently the VSTi is coded, that may not be a significant savings.

I do use multiple instruments per instance and I wait to add envelopes until after I bounce each track to a separate audio track.  I normally make a new copy of the project before I start bouncing to audio so I can safely remove the VSTi's as I go in the new copy. It makes mixing later on simpler.

Dan

 

[msmcleod]

On 6/5/2019 at 2:04 PM, dcumpian said:

The answer is...maybe.  Using multiple instruments per VTSi instance should reduce the overhead required by each separate instance. However, depending on how efficiently the VSTi is coded, that may not be a significant savings.

I do use multiple instruments per instance and I wait to add envelopes until after I bounce each track to a separate audio track.  I normally make a new copy of the project before I start bouncing to audio so I can safely remove the VSTi's as I go in the new copy. It makes mixing later on simpler.

Dan

 

@Craig Anderton did a tip on multiple instruments per VSTi vs multiple instances: http://www.craiganderton.com/tips.html (180713 Multi-Timbral Instrument vs. Multiple Instances—Which is Better?)

The TLDR version is, that unless the VSTi is coded to support multiple cores, then using multiple instances can spread the load more evenly, making CPU use more efficient at the expense of using slightly more memory. Using a single VSTi as a multi-timbral instrument saves memory, but can put more load on a the CPU core its using.