Page 3 of 3

Re: HardSync FM problem

PostPosted: Thu Aug 08, 2019 8:15 am
by Spogg
I do hope there comes a solution which satisfies you Adam, and I would indeed love to see it.

One thing that bothers me about Martin’s idea is that it would surely involve a small delay while the simulation takes place. This would manifest as latency and could be up to 50mS, depending on frequency and phase adjustment. If this would be acceptable (which I doubt) then tulamide’s approach of blanking the first part-cycle would be a solution. Simply wait for the first sync pulse to arrive before allowing the output through.
The big issue would be with playing chords, since each note would start at a slightly different time I think.

Anyway, good luck!

Spogg

Re: HardSync FM problem

PostPosted: Thu Aug 08, 2019 8:21 am
by martinvicanek
1. Start simulation at a sync point
2. simulate master and slave including FM up to the point where your real oscillator starts (no need to simulate a full cycle)
3. use the osc states at the end of the simulation as initial values for the real oscs

In blue stream you can do that in stage0. I don't think that's possible in poly, though. :(

Re: HardSync FM problem

PostPosted: Thu Aug 08, 2019 8:26 am
by adamszabo
martinvicanek wrote:In blue stream you can do that in stage0. I don't think that's possible in poly, though. :(


Hmm, in a poly stream you can do a stage0 hack with bitmasks I think?

So you do declare:

int true = -1; //true bitmask
int false = 0; //false bitmask

then do some calculation and get values with: val & true;

then at the end of the code:

true = false;

from, there the "val" will not be active in the code anymore?

Re: HardSync FM problem

PostPosted: Sun Aug 11, 2019 7:35 pm
by adamszabo
I have played around with it some more and I think I might have something! If I modulate the synced saw with a static number (like 0.5 for example) I can get the phase1 to align correctly with any fm or semi1 or semi2 combination, but when I switch it to an osc wave (which is the whole point of FM) then it gets messed up again. The 3.0.8 FS version does not have some functions as in the Beta version, so I had to use assembly to simulate the stage 0 for poly signals. Martin (or anyone else) do you have any ideas why the static number works but the not the osc?

Re: HardSync FM problem

PostPosted: Thu Sep 05, 2019 11:20 pm
by martinvicanek
OK I have implemented my proposal. It does work in the demo, however I am not sure it can be applied in poly?