Hi,
I'll try to give some example for why it is important to have the choice of controller being sent after or before the according notes. Please accept my apologies in advance for this being such a long post.
EXAMPLE 1
Imagine the OT controls a MIDI synthesizer producing a bass sound with short attack time. The sound uses CC1 (ModWhl) to control filter cutoff frequency. The OT's MIDI track has 3 trigs with MIDI notes, the first trig also sends a high CC1 value, the second a low CC1 value, the third a high CC1 value.
If you playback this short sequence, the audible result with the current behavior (send notes first) will be that the second note will have a short "blip" at the beginning (because cutoff is high at the beginning of this trig due to the high CC1 value of the previous trig, then very quickly drops to the low value of the current trig), while the third note has a short "whop" at the beginning (because cutoff is low at the beginning of this trig due to the low CC1 value from the previous trig, then very quickly it raises to the high value of the current trig).
This is not what the user expects: He simply expects that the second note has a low cutoff (because he set CC1 to control cutoff in the synth and set the CC1 value of the second trig to give a low cutoff), and he expects that the third note has a high cutoff, both without any audible glitches.
The only way to achieve this is by sending the controller immediately before the note event.
EXAMPLE 2
Now imagine a synthesizer percussion sound (a "zap") in which CC1 increases the decay time of the VCA envelope (from a "zap" to a "piuuuu"). With low CC1 values, the sound decays very quickly, with high CC1 values it decays slowly. With alternating low/high CC1 values on adjacent trigs, the resulting sounds will actually have unexpected 2 decay phases: Either the sound decays very quickly to a very soft level and then decays slowly from this very soft level to zero, or the sound seems to stay at around full level for a very short time and then decays quickly to zero.
But what the user expected were 2 sounds with predictable decay times: one very fast "zap", the other a very slow "piuuu".
The whole thing becomes even more complicated (and gives even more unexpected results) if you use a MIDI controller to control multiple synthesizer parameters at once (e.g. using Clavia's morphing capabilities).
Or to put it short: First the sound should be defined, then a sound should be triggered. The current behavior reverses this: First you trigger the sound, then very shortly after the triggering you modify it. Unless you use sounds with a considerable (= audible) attack time, this will lead to glitches.
If you put this parameter on the MIDI machine's playback setup page, users could decide based on their needs which behavior (notes first or controller first) they want.
Please reconsider this for a future update.
Thanks and best regards
P.S. If you are concerned that MIDI note timing might be affected by this, please consider that the SND SAM-16, a hybrid MIDI/CV stepsequencer (first built in 1997 & still in production, see
http://www.schneidersladen.de/en/snd-st ... r-sam-16-2) gives the user the choice of what to send first (notes or controller) and is still considered as one of the sequencers with the most rock-solid timing. I use one of these and have to say that this what-comes-first-setting really makes a difference.