# BRAIDS : Quantize the poly modes inputs?

Would it be doable to quantize the poly modes’ intervals ?

Do you mean the triple modes? They are already quantised. here is the intervals array used by the timbre and colour parameters to set the pitch offset for the second and third voices in these modes:

#define SEMI * 128

const int16_t intervals[65] = {
-24 SEMI, -24 SEMI, -24 SEMI + 4,
-23 SEMI, -22 SEMI, -21 SEMI, -20 SEMI, -19 SEMI, -18 SEMI,
17 SEMI 4, -17 SEMI,
-16 SEMI, -15 SEMI, -14 SEMI, -13 SEMI,
12 SEMI 4, -12 SEMI,
-11 SEMI, -10 SEMI, -9 SEMI, -8 SEMI,
7 SEMI 4, -7 SEMI,
-6 SEMI, -5 SEMI, -4 SEMI, -3 SEMI, -2 SEMI, -1 SEMI,
-24, -8, -4, 0, 4, 8, 24,
1 SEMI, 2 SEMI, 3 SEMI, 4 SEMI, 5 SEMI, 6 SEMI,
7 SEMI, 7 SEMI + 4,
8 SEMI, 9 SEMI, 10 SEMI, 11 SEMI,
12 SEMI, 12 SEMI + 4,
13 SEMI, 14 SEMI, 15 SEMI, 16 SEMI,
17 SEMI, 17 SEMI + 4,
18 SEMI, 19 SEMI, 20 SEMI, 21 SEMI, 22 SEMI, 23 SEMI,
24 SEMI - 4, 24 SEMI, 24 SEMI
};

Actually, they are not quantized, but the pot response has a little plateau around these values.

You can quantize (without hysteresis) by setting this to 0

Ah yes! But added some hysteresis is probably essential, otherwise it will jump around between intervals with an audio frequency spectrum in some circumstances, and will sound horrible (as I found when adding harmonic series quantisation to Bees-in-the-Trees - but adding just a little hysteresis as used when quantisation is enabled for the 1v/oct input solved that nicely. Adding some hysteresis is not hard though. I’m happy to try this out and if it sounds OK, add it to Bees-in-the-Trees if there is sufficient room. So the answer to the OP’s question is yes, it seems like it is quite doable.

Are there different sets of intervals that would be better, rather than 2 octaves of chromatic scale up and down? In other words, a smaller number of intervals to make it easier to form specific trichords?

Pichenettes, you mean, 0 instead of the 8 value ?
I realize that if it was hardly quantized, it would be a bit sad… some “detuning” in chords is great.
i wouldn’t know how to change the code and make a soundfile out of it…

And semitones are ok to me !

@chapelierfou, no, set xfade to zero so there is no interpolation between detune1 and detune 2. But without hysteresis, if the voltage from the timbre or colour controls is just at the point where it would cause a jump from one semitone to the next, then it may jump back-and-forwards due to very slight noise in the voltage and/or noise in the ADC chip. This jumping back and forward may occur at up to 4000 times a second, so can cause an unpleasant audible screech. Hysteresis means that the change in value from the ADC is made to be very slightly “sticky”, so that these tiny changes due to noise are suppressed. It will be essential IMHO.

Anyway, I’ll add this to the list of modifications to be tried out in the next version of Bees-in-the-Trees, which I am working on now - but no promises on timeframes, and it may not make the cut because space for extra code is very, very tight. But I’ll make the changes in a branch in git called chapelierfou, which I’ll push to GitHub so you can see what needs to be done, even if it doesn’t end up in Bees-in-the-Trees. I’ll report results of the experiment back in this thread.

1000 thanks, Mr Bicycle !

Well, I tried it without hysteresis, but frankly it doesn’t make much difference, so it won’t be going into bees-in-the_trees v3.5, sorry. However, if you have a Frames, you can set the frame voltage to be at the semitone point in any case.

I can compile a version of the plain Braids firmware with the change if you like. Why do you want the change? How do you intend to use it? I am asking because there may be other approaches. Do you need 24 semitones up and down on the timbre and colour controls for these triple modes? Maybe just a smaller number of steps and broader plateaux or landings for each step would be better?

great! didn’t know that ! noticed the plateaus, thought “wow”

Didn’t see you replies, sorry !
Did you compile it in the end ?

No, but I need to improve a few other little things in Bees-in-the-Trees this weekend, so I will look at it again then.

Nice, thanks.