Solution to missing Plugin Delay Compensation for Automation

[orangeman22]

Hey everyone,

First post here, interested in PDC and automation. Please correct me if I'm wrong, but isn't a potential solution to resolving this issue just simply bouncing an automated clip DRY first and then sequentially rendering with each effect added on?

That is to say, if you have a lead synth chain like this:

NI Massive > Reverb > Compressor > XYZ

and you'd first like an envelope filter to open up in Massive, then automate the reverb, then compression, couldn't you just:

1) Automate the envelope filter, render the lead
2) Use the rendered audio clip, apply your reverb automation to it, render again
3) Use the re-rendered audio clip, apply your compression, render
...and so on.

With each step, you'd only have one plugin (technically, as long as the others aren't in the FX chain) applying minimal latency. Seems a bit tedious, but until we get any sort of real automation support for PDC, isn't this the next best thing?

Please correct me if I'm wrong, and share your thoughts/more optimized strategies if you have any!

[fx23]

indded thats one way to go, but note you would have to DELETE the plugins, the latency is not recalculated and remain if they are just OFF.

then once you rendered x times and you say 'oh shit, that filter is a bit to open' u just kill urserlf :p

a workaround, but not really worklfow fiendly and modular...
my prefered method is to keep all to be able to go back if i need., and so i prefer personnally to manually offset the automation/modulation, not by offseting it itself, but by offseting the start play header.
(by setting it to a loop and moving the start header to the end of loop and slowly move it backward till i hear the result is correctly offseted), works nice even if boring for automs/mods, but won't fix timed plugins problems btwn...
the cool thing with this tech is you A) keep all if need to go back B)automation/mods are still snap on grid, as only the
offset of playing header acts as 'MANUAL PDC'

[orangeman22]

indded thats one way to go, but note you would have to DELETE the plugins, the latency is not recalculated and remain if they are just OFF.

- Yeah, that's what I meant; sorry for not specifying.

Your method is interesting too (and I've tried it in the past) but like you mentioned, maybe not the easiest to implement.
That's one of the few gripes I have with Live ATM TBH, everything else has been pretty solid. Just a shame something this simple has been overlooked.

[fx23]

i think so . imo live is quasi perfect and mature now, they deserve lot of respect for having atlast implementing the
automation to session, and transparent moves session arrange are just so much more coool.
i also find that's the only 'big' problem that reamain,.. to get it rock solid. other stuff is not drastic imo,
so i do hope they'll focus now hard on that one..and would be pretty logic, now we got shinny automations, they just have to be accurate

tightness is very important, groove/attack/sync effects is a subblte but very important thing, and often neglected aspect of 'sound engine quality' debates, but once we ll get it , probably live 10?? the beast will be perfect.

[dolomick]

I've never fully understood the following so just to clarify... if I avoid using automation in Live, shouldn't Live's PDC work reasonably well at keeping my music in sync? My understanding is there is a 1ms latency to using sends, but other than that, everything else (including 3rd party vsts) is compensated when just used as plugins, correct?

Thanks for clarifying.

[Stromkraft]

With each step, you'd only have one plugin (technically, as long as the others aren't in the FX chain) applying minimal latency. Seems a bit tedious, but until we get any sort of real automation support for PDC, isn't this the next best thing?

I've yet to be bitten by this, but one alternative method to investigate would be to "group" single tracks and split instrument automation on the subtrack from effects automation in the group bus track. I guess this is only worthwhile for the simpler cases.

I think I read this in one of the PDC discussions. Anyone more familiar with this method can tell more, I'm sure.