MIDI vs. Audio

I was wondering at what point people bounce their MIDI tracks down to audio. And how you organise the MIDI tracks/soft synths the audio came from.

I'm using Cubase SX and VSTIs, and the problem I have is that as I use more and more VSTIs pretty quickly I have to start bouncing them down. Fair enough but even if they're not playing Cubase is using shedloads of RAM in no time (like 800MB) and I have to start removing the VSTIs entirely from the project. Problem then is it starts to get confusing which audio tracks came from which instrument/MIDI track etc.

How does everyone manage a tradeoff between keeping the original MIDI/soft synths somewhere safe, and having a system that's not on the verge of crashing?

I bounce em as soon as I've got a decent sound or pattern going. I just mute the midi-track and when I want to alter the sound I simply bounce it again. I think this is how most of us work

I build all of my tracks from audio.

I make all the sounds in synths, bounce them down and process them from there.

You will be suprised how much tighter your tracks will be if you ditch midi this early on in the process

Quote:

---

Originally Posted by benzene

How does everyone manage a tradeoff between keeping the original MIDI/soft synths somewhere safe, and having a system that's not on the verge of crashing?

---

Labelling and placement. But, I use Sonar. When I need to do this, I often just use the freeze track option.

i tend to get them bounced down to audio asap. if theres a lot of effects automation need, i'll usually record the whole file, bounce it down and re-import it as a full wav. if theres only minmal automation needed i'll bounce down to rex loops immediatly and then draw in automation. working with rex/audio gives you a lot more scope as far as plug-ins and editing are concerned.

Thanks for the answers - makes sense to me to bounce em down as early as possible.

Yes I guess audio is more powerful to work with - that and it will prevent me from going back and tweaking MIDI patterns too much ...

once i bounce to audio i save the cubase project under different name, so i can go back if i want to change something on VSTs.

i also bounce to audio as soon as i've got decent sequence done. i mostly leave it without any effects and doing it then after i reimport the audio file.

was also wondering if muting VSTi is helping to CPu,. i think not. because even it's not playing and it's muted, it's still calculated on fly, so itl takes some cPU power, or?

it does, but not as much. same as when no midi signal is present. but still, it drains some cpu power.

sometimes it drains pretty much CPU,,with synts like Oddity, Rhino or such hungry beasts, even it's muted.

What about if the MIDI is muted, and you power off the VSTi? (Keeping settings, but disabling the plugin).

it helps, but its not the same as getting rid of them completely. check for yourself, load VST performance and test it.

if the vst is powered off then there is no processing, so cpu drain is reduced. but then you dont hear anything... (unless you've bounced)

Quote:

---

Originally Posted by rhythmtech

if the vst is powered off then there is no processing, so cpu drain is reduced. but then you dont hear anything... (unless you've bounced)

---

Yeah, that's what I figured. But presumably it still uses some RAM for storing the setting etc.

I guess the best option is to try and be conservative with the way in which you're using FX to start with.

MIDI is for girls! :)

^^lmao..

i dont bounce down until i have to .


i allways have loads of different controller information on my synths and i like to change it all on a bar by bar basis as the tune progresses also i often have a few chords and the odd key change , i like to be able to change key or try out different chord versions , inverted , diminished etc in different sections of the song . once its bounced its just a loop and as such limiting . Audio is a hell of a lot tighter though .

I write all of my patterns, bounce them down to wave files....re-import them to the project...do my edit work....do a bit more arrangement, then do more edits to my wave files from there as needed within the project. Works best for me.

I virtually never bounce down until the final render of the whole tune.
Unless I want to do some audio specific edits to a certian synth, I might bounce it down and then edit the audio track, but very rarely.
Lots of vsti, lots of automation.

Quote:

---

Originally Posted by dirty_bass

I virtually never bounce down until the final render of the whole tune.
Unless I want to do some audio specific edits to a certian synth, I might bounce it down and then edit the audio track, but very rarely.
Lots of vsti, lots of automation.

---

nazi

Quote:

---

Originally Posted by stjohn

nazi

---

I am the serendipity nazi

i bounce down from midi to audio once i like what i hear and have a basic idea of the direction of the track.

For me, i bounce down to audio nott out of choice but a necessity. I have noticed tthat if i dont force myself to bounce down, i end up fiddling with midi notes for DAYS and in the end get nothing done.

Sometimes u just gotta force yourself to work with what you have and print down to audio and deal with it later.

When you guys say tighters. What are you talking about? Timing errors in the MIDI? Is this the case even with an all software setup?

dunno about other sequencers but in live 6 you can track freeze midi tracks which temporarily renders it as audio. if you want to adjust any of the plugins settings you just unfreeze the track. snorted :)