Plugin/DSP Load Management

I have an hour long session (film mix) with multiple tracks (vocal). most of them have one or two plugins (eq/comp/rev/delay) on the track. Just wanted to check that if any of these tracks has no audio (empty, but has plugins) will it take any DSP/RAM/CPU resource?

If there are plugins there it will take resources as long as the track is active. If you don’t need the track, deactivate it.


so, what would be the best way here… i have most of the tracks which is active only for few minutes at various point of time… how about bypassing plugins? will that free resources?

Bypassing the plugin will cause it to be fed silence, it will still be processing silence IIRC. Deactivating the track is the only way. You can always bounce with plugins to another track and deactivate the original in case you need to update it at any time. This is a trick I use often, especially on video projects, and is one of the only places I use a certain part of Ardour that is very confusing for many folks(So am intentionally not saying it right now, I encourage you to see what you can figure out first)

how about bypassing plugins? will that free resources?

That depends on the plugin - I’ve always tried to design plugins which when bypassed conusme only minimal DSP, however I would expect that if you disable the plugin via Ardour (rather than using the plugin’s own bypass switch) then its DSP will be completely inactive, although in some cases you might get clicks and pops when the plugin is switched in and out this way, whereas you shouldn’t if you use the plugin’s own bypass.

bypassing a plugin does NOT cause it to be executed (or fed silence). bypassing a plugin means that it will no tbe executed until it is no longer bypassed.

@paul Ok thanks for the clarification, I stand corrected.


How about muting tracks? Would a track that is muted that has plugins still be processed or would this deactivate anything. If muting a track stops the plugins on that track running, then you could use mute automation

@Paul: thanks for the clarification. So, we could save some cpu cycles by bypassing the fx chains if required right?

@veda_sticks: I think muting is happening at just after the fader and it just opens the signal path to not to flow anymore signal and hence
it won’t cause any track to be deactivated.

@abhayadevs: Deactivating tracks is a very useful feature you can use in a very smart way, as an example i’ve been working on a 26 track (about 2 - 3 plugins per track) and 7 mixbuses (around 3 plugins per mixbus) session on Mixbus, my PC is capable of handling that however mixbus uses one or two cores of the four my CPU has, this limits greatly the handling power of the machine and when over 40% DSP for some reason mixbus won’t export so this is a workaround meanwhile mixbus hopefully sometime moves to A3 core.

  • I choose one type of sound to “bounce” for example Drums
  • Deactivate all tracks (select them and Alt+A to activate/deactivate them in groups rather than one by one) i don’t need, just mantain Drums, Drums mixbus, reverb and room, deactivate all other plugins in master or other mixbuses… and Solo the needed mixbuses.
  • I record the master output to a stereo track, and do the same for all other types of sounds

At the end i have maybe 4 o 5 Stereo tracks with all processing applied to mix back, it is possible you have to realign them since you won’t be using the same DSP processing and Plugin related latency, but once done that, you can deactive all other tracks, all mixbus plugins and just mix by Fader these tracks and use a few plugins on the master bus that won’t go up to 40% DSP, in my case it’s been reduced to 20% (FROM 90%) and mastering is a charm them… however it is not so funny to re bounce tracks when changing the mix of a mixbus… so better try when confident on the sound of every mixbus.

I hope that helps you.

great… nice workaround and really helped to have some options in hand.

For me, the new film mix i am planing (for the first time !) i am preparing a workflow. so it has many (not less than 50) tracks altogether and i really need to plan well to mix them all with my core2 duo cpu :frowning:

well, as long as you do prepare a workflow everything should be fine, thinking about the result and the workflow even before you start is very smart, however consider using high latencies and A3 for all the project since it works much more easy than Mixbus, also editing is more intuitive and easier and with the new meters is perfect for a film mix, i wouln’t use Mixbus until its about mixing a few tracks with that CPU, my laptop is a Core2Duo 2.0 and works fine most of the time with mixbus even with more than 20 tracks using a usb 2.0 interface but im pretty sure its not happy doing it.

@fernesto: yeah, getting smarter with so many constraints and low/0 budget;) what I am planning now is to have a session/project with all the tracks in the region/track list and will create multiple/duplicate sessions for each scene for Voice/Dialogue mix which can be bounced to a mono track and then some more sessions for FX/Foleys/Music which all could be bounced to one or more tracks and then mix them all together to have the final stereo mix. I am also planning for producing this in 5.1 but this is not yet finalized.

Its been some time and before me getting in to the actual film post i thought of doing a short-film… please see the film ( and let me know your feedback on the audio-post - foley, sound design etc…

abhayadev s