even before trig conditions came to the Octatrack, it has been possible to create a variety of probabilistic and ‘conditional’ emulations through clever application of simple LFOs. After exploring a few ways to implement probabilistic and conditional behavior without using trigger conditions we will also discuss WHY you may still want to utilize this technique even though trigger conditions are readily available - namely modulation of probability and masking patterns as well as crossfader control of these ‘trigger conditions’
5 Likes
Good tutorial! Main complain : screen is too blue! 
Sorry, but I need to add something maybe more efficient : randomize lfo designer! 
No need for audio editor in that case. 
I also used slices randomization but recording a pattern will all variations states. Example :
Flex with “Open Hihat” sample, every 2 step, 2 bars pattern. Plock variations : vol, filter, release (short for closed hh), etc…
Resample, 16 slices grid, random lfo on slices.
2 Likes
haha! I went for high contrast screen instead of an accurate representation of Octatrack reality 
to know what % you are deriving the easiest way is to use slices and audio editor, otherwise you are guessing or need to do fancier maths to figure it out. Also you can throw in lots of slightly varying slices in (softer hits, repitched hits, etc.) and save some LFOs you might otherwise route to vol or pitch
also for ‘masking patterns’ we have 3 things to work with to derive new patterns - slices, slice selection LFO (and modulation of that LFO ofc but I’m leaving that out), and trigs - if you neglect slices you lose a valuable source of abstraction for generating a wider variety of patterns more easily (edit: using sample locks you can have different versions of your slices and different slice patterns to select from for individual trigs for example - lots of other things to be done with this technique and slices - in this video I only showed the most simplistic thing possible)
Me neglecting slices! 
I use start / slices most of the time, and sometimes I’d like to slice my slices. 
I edited my above post before your answer. I see your point concerning perfect %.
I’d rather record a loop with right variations, slice it, eventually delete some slice for right %… 
Anyway thanks! It is really interesting and motivating to do this again and go further. 
1 Like
since I am modular guy I think of things in terms of ‘modules’ - the pattern of slices in a file is basically a sequencer module, in this case what amounts to a trigger stream since all the hits are the exact same - we have the trigs on the track as another sequencer that can interact with the ‘slice sequence’ - and then we have LFOs to dynamically change the relationships between these two patterns and modulate their internal states (slice selection window via starting slice and LFO depth)
the more separate ‘modules’ and layers we can access then the more we can get in terms of different results as changing things on one layer of the ‘network’ has cascade effects to other ‘nodes’ - each extra ‘module’ adds exponentially to the possibilities - the way to really explode this technique would be MIDI loopback - but I’m not sure I want to take it that far yet! 
edit: I think my favorite part of this technique currently is being able to control probability with crossfader and add probability to conditional trigs like fills
2 Likes
Explode is intended?
I explored midi loopback quite a lot, endless possibilities with additional lfos, plocks, TRCs, arp, rdm designer lfos on arp speed…
With my midi processors I can trigger slices with midi notes, so I let you imagine with arp…
1 Like
yes! an explosion of possibilities! 
I am always wary of MIDI latency/jitter producing inconsistent results when I do fancy things like that though, I want to make sure everything is nice and tight and performing in the way I expected. If I am working on making music and all of a sudden the MIDI glitches a bit I am suddenly no longer making music and thrust into troubleshooting mode - so often I keep things purposefully simple to make sure they are working properly 
1 Like
very well made tutorial
3 Likes
thanks for the compliment! sezare might be right though… maybe I did make the screen too blue! 
this concept seemed like a good way to efficiently demonstrate various basic Octatrack functionality and possibilities but still provide something to gnaw on for more advanced users
2 Likes
Good stuff. Using modulation esp crossfade really gets one thinking.
I didn’t mind the blue. Reminded me of:
1 Like
thanks! definitely lot of cool possibilities here - to go into everything in detail would be a very long video, but I do have in mind a more specific and detailed implementation of similar techniques that may be worthy of a video - have to sit down and try some things out
hah, chicago flag blue is more like real-life mk1 screen than my processed version!
Out of curiosity why don’t you normalize in the editor?
1 Like
the main reason is that if I normalize something it generally just means it’s going to clip more easily and I’ll have to turn down the gain later in the chain anyway - I aim for the hottest tracks in a given project to be at a track level amount of around 100-110 - lower than that and the dynamic/volume resolution feels dodgy (I’ll have to take a look at the value scaling sometime to see what’s going on there or if I’m just imagining things), and higher than that there’s not any headroom left on the control
4 Likes
Right. I try to keep my levels there too but at the same time ensure its at 0 gain in the sample attributes (for samples I want to work with later…live I leave the gain as is, Im not trying to normalize live, no time for that). Without the gain or normalize they’re way too quiet.
I was curious if you had discovered any quality issue. Thanks!
yeah, normalization is just gain, and a purely mathematical operation - it has no effect on quality in and of itself. If I want gain at the sample level I prefer to use the sample attribute gain as it’s not destructive and is more flexible than normalization
4 Likes