A question about drums, and compression

[Grime Syndicate]

So, I've got this break all chopped up, and i've got a kick bus playing a kick under the break, same thing with a snare bus, playing two different snares under the break (all layered up and effected and reverbed) then I've routed the kick bus to a return with a compressor squashing the shit out of it, same thing with the snares. I.E. NY compression or parallel compression. All this effecting has done a number on my cpu tho, and i'm sitting at like 50% cpu usage with just the drum tracks playing. SO, here's my question, if i just resampled what's coming out of the, lets say for instance, the snare bus and the compression return, simultaneously, would this translate properly, and do away with the need for a real-time compressor/reverb/distortion running on my snares? It seems to make sense in theory that i could then plug the resampled hit sound back into a drum rack (I'm on ableton live and theoretically, all those things should be heard, but not be processing in real time. Is this logical? or am I just hoping against hope? how do ya'll run your super punchy drums? Punchy is what i'm on about, cuz if they're some ambient/deep drums, i'm guessing all this wouldn't be necessary... Really curious about this one for SOME time now, so yeah, any insight would be delightful...

[Mike145]

i have the same problem except with all the parts of my tracks . I just keep adding effects and midi until my CPU lags so much that you cant even really listen to the song... I would assume this would work though, of course the only problem is you wouldnt really be able to go back and change it afterwards, so youd have to be pretty sure thats exactly how you want to do it. You should bounce it to audio and let us know how it goes

[nowaysj]

Honestly, you shouldn't be that high on cpu unless you have a completely shit computer. Like completely.

If that is the case, f'ing resample your beat, f'ing just resample the whole thing, lock stock and barrel, and move on. When you are a super producer, you can have your assistant do this for you.

[Wrigzilla]

SO, here's my question, if i just resampled what's coming out of the, lets say for instance, the snare bus and the compression return, simultaneously, would this translate properly, and do away with the need for a real-time compressor/reverb/distortion running on my snares? It seems to make sense in theory that i could then plug the resampled hit sound back into a drum rack (I'm on ableton live and theoretically, all those things should be heard, but not be processing in real time. Is this logical?

Short answer yes.

As long as it bounces properly you'll get what you're hearing now in an audio file. Hell I've done similar things with breaks before, I've made an EXS24 preset which has this sliced up amen break that I've already processed loaded up in it with all the clinical EQing, distortion, compression, transient shaping etc it needs done..

[Grime Syndicate]

Honestly, you shouldn't be that high on cpu unless you have a completely shit computer. Like completely.

If that is the case, f'ing resample your beat, f'ing just resample the whole thing, lock stock and barrel, and move on. When you are a super producer, you can have your assistant do this for you.

Ok, well, checkit out, I DON'T have a shit computer, not a little, not completely. Its an AMD turion duo-core with 3 gigs of ram and many other similarly decent specs... so, ummm, nope. Also, I have absolutely NO SPYWARE, ADWARE, MALWARE OR VIRUSES anywhere on my system, multiple scans by the best companies for doing so have all come back negative, (avast, ad-aware, TDSSKiller by Kaspersky, Emerisoft emergency kit scanner, etc, etc.... I've got a LOT going on. reverb as a return in the first drumrack, effecting all the combined layers in different amounts, plus, izotope trash and alloy on both drum channels. SO, truly, i don't think THAT's the issue. jus saying

[Fbac]

Are your virus scanners running in real time? That could be the issue. but i would agree with norwaysj that you shoulnt be having such high cpu usage if thats all your doing.

back to the main question. (In my opinion) Yes, you are correct, that by exporting your effected drum tracks out to audio then create a new project and re import them (Or re-save your project,, so you can go back if you think you need to change something)

I would suggest, exporting a Dry and Wet version, Kicks without the effects on,, personally i might suggest taking the reverb off then you can still compress and change dynamics but usually you dont want a compressor after the reverb.

[nowaysj]

I could run that on my intel quad core 2.6 ghz and not go any where near 10%. Just saying. Sounds like there is something awry in your computer. Could be hardware/software or a combination. Maybe you are using a stock audio device, or using a usb device?

[Grime Syndicate]

usb audio interface, yes....

[Lichee]

dude no need to get touchy

[CyMoRrR]

A duo core is already old... Could be it's not so fast as you think it is
3 gb of ram is not much so i advise to boost things up (if you run 64 bit)

Also think of the freeze function in your daw. this way you can freeze a track and while you cannot edit it
at once you can always unfreeze and edit. This saves the trouble of bouncing and not beeing able to fix some errors!

[staticcast]

ableton's CPU metering is a little funny. as far as i understand it, the CPU meter in live reads the hardest-worked CPU of all of your cores. therefore, if you've got one track that's using a huge amount of CPU time, you'll get a high reading even though only one of your cores is working at high capacity.

the trend in modern computing seems to be to add more cores rather than increase the speed of each one. therefore, design of efficient software largely hinges on how optimized it is for multicore processing. DAWs do this by spreading the tracks over all your cores, but in the example you've cited, you've basically got one massive track with loads of effects, made worse by the fact that any return track has to be processed AFTER all its feeder tracks have been processed. on this example, there's no reason why an 8-core, 2 ghz machine should perform any better than a single core 2 ghz machine.

my guess is that even though your CPU reading is high, you could then go and add a bunch of other tracks without it increasing significantly, because the load would be spread over the remaining cores, provided that you didn't do something complicated like route that track into the first one.

(this doesn't answer your question, just saying...)

[zerbaman]

Bounce to .WAV
Then start a new project file using the stems.

this way, if you decide later that you want to change something, you can load up the old project and re-bounce the new/changed element