›

Random Velocity in the Step Edit Page

widdlywiddly
edited October 2013 in Feature requests and bug reports
The probability parameter is great for adding variety to a pattern but it would be cool if random velocity was available as well.

I imagine it would work as follows.. 

A random parameter is added to the step edit screen (with the velocity, note and probability settings) so that each step (and substep) can have a random setting. 

When a step is played, the actual velocity is calculated like a multiply.  So a random setting of zero means the output velocity is the same as the velocity setting.  A random setting at 127 means the velocity is completely random between 0 and the velocity setting.

It would allow adding a little random velocity to patterns which sounds more natural.  And added benefit is that it would allow some cool random modulation when using velocity for other modulation destinations.

Comments

  • tonebursttoneburst
    +1 for this.

    a|x
  • julianjulian
    Good idea. But I have to think about the implementation. Each parameter added on a per step basis makes the pattern memory consumtion much bigger. 128 steps * 7 tracks * 8 patterns = 7.168 new values per parameter.
    So i'm not sure if adding 4 (random, note, vel, prob) new parameters to each step is feasible.

  • loderbastloderbast
    i think at least for the random one, 4 bit /step is plenty. if you implement like widdly suggested, the resulting value can still have a more fine resolution because the randome parameter only sets the upper boundary for the randome generation.

    the other values are 7 bit if i am correct (because of midi cc range) so maybe it might be a idea to have the data coded somehow to save memory?

    i think oliviere wrote something about that topic recently over at mi forum, but i cant find the thread right now.
  • julianjulian
    what about simpler approachs:

    - a random parameter set on a per track instead of a per step basis
    - reduce probability resolution to 4 bit, add 4 bit random parameter = 16 steps each + a global vol,note multiplier set per track

  • julianjulian
    on the other hand random velocity per track can also be achieved using an LFO and a retriggerd S'n'H waveform
  • widdlywiddly
    I think 4 bits for probability and random would work great.  I certainly can't hear the difference between all 127 values for the current probability.

    Doing with an LFO certainly works but to my thinking, the randomness should be part of the pattern not the voice.  
  • julianjulian
    the 2nd argument is a good point! ;)
Sign In or Register to comment.