Previously while performing I would simply arrange pre-produced tracks, with about 50% being bounced stems from those tracks and the other 50 being the actual rack instruments and automation from those tracks that I'd manipulate from a controller (or two). I didn't really produce with playing live in mind though. Over the past few years, that's all changed and most of my producing is now improvised, and in general the nature of my song files and Reason configuration has changed quite a bit:
- Firstly, I use Reason exclusively, nothing else (this has always been the case).
- I never use the sequencer if I can help it. If it can be done in the rack, it is getting done in the rack.
- Even when composing at home, if I can help it, I never hit stop on the sequencer (after all, I wouldn't hit stop if I was playing live!)
- Since the sequencer's use is almost 100% limited to Rex Loops and raw audio files, as well as any ad-hoc automation that can't be CV-automated, like certain mixer channel settings, there's a LOT of step sequencer use in the rack and a LOT of CV routing, gating, merging, and splitting. I've gone so far as to create CV mux devices to automate pattern selection via other patterns or sequences.
- My typical rack is usually 6-7 columns wide, just to give an idea of the scope. Also, I don't use a lot of synths other than drum synths. Mostly samplers/rex loops, and there'll be on average 2 synths for non-rhythmic elements that I'll simply manipulate through various effects and parameter automation. I'd say no more than 10-15% of the devices in my rack actually make sound.
- I use combinators for almost everything. Most of them have between 10 and 50 RU worth of devices in them. By FAR the most common devices in my rack (other than Combinators) are CV mergers/splitters, Euclid Rhythm Generators, Select CV Switches, various Chronologists devices (mostly Truth logic gates and Now counting gates), the Quadelectra CV suite devices, and Rex players - all fairly lightweight devices AFAIK.Other than I guess the Dr. Octo Rex none of these are more than 2 RU but there's a LOT of them.
Finally, here's some stuff I'm already doing. Some of it I know for a fact can help, and some of it is out of necessity, but some of it just me guessing and may actually be doing more harm than good:
- Obviously, I close everything else but Reason (which I usually open my song file in well before playing), even widgets and small stuff like that. If its dark enough wherever I'm playing I turn the brightness way down.
- As far as power being drawn from the laptop, I have two midi controllers and a 1TB HDD used as a swap/scratch disk.
- I abuse "save and optimize", remove unused recordings, and usually self-contain all samples immediately before playing.
- Settings: Mouse knob range and automation cleanup levels are set to their minimum values. Trigger notes while editing and Cable animation are off. Reduce cable clutter is on and set to show cables for selected devices only.
- Sample rate: I use 88200hz. The reason for this is that I do a lot creative BPM manipulation and I'll jump between 1 and 500 bpm regularly (if Reason would add 256th notes the need for this would be lessened...) once you get up past 200-ish or so 44100hz and 48000hz are insufficient and I start getting artifacts in some cases.
- Buffer: 2048 - this is to help with audio files which get stretched and un-stretched all over the place while I'm playing.
- I'll usually set a loop in the sequencer and then never touch it again, just for the sake of resets on pattern devices. This is typically between 64 and 256 bars.
- I try my best to keep everything in the rack I don't need to see collapsed, especially stuff like oscilloscopes.
- Everything I've made as far as song files, loops, patches, and audio files, as well as all refills, are all under the same folder.
Reason itself as well as all VSTs are in different locations (not sure if it would help as far as I/O to keep literally everything in the same spot).
- Is a scratch disk helping in this context?
- For live performance is it better to self-contain all samples or not?
- Not counting the Rex/drum machines, I rarely use samplers - how much benefit will I see from loading any audio files I don't need to stretch or cut up into an NNXT or something similar?
- Can rearranging file locations improve performance?
- As far as my rack workflow: I'll typically use CV splitters with the maximum number of inputs/outputs (I rarely use the Spiders). I'm under the impression that sequencers/pattern devices hit your CPU harder than simple CV processing devices and therefore I'll limit the number I use and instead use splitters to route them to various combinators stuffed full of CV gates and routers - the thing is, I'm pretty sure after CV data has traversed a certain number of devices that I'm going to run into latency. Is there a good balance to try and strike here?