Pfm2 : unison discussion

[Xavier]

@matrix12
Thanks for the feedback 
There are still some UI bugs to fix.
Unison menu should no show up among the modulator pages.

Yes i think you figured out how it works, but the preenfm2 UI can be a bit confusing.
So i confirm :
. Glide only shows up for mono sound (voice = 1 OR voice >1 and Unisons).
. The engine2 ("Play mode" menu) only shows up when voice > 1.

Will try to upload fixes tonight (france time).

@vurt
Yest that can be usefull for you as you mainly use mono sound in your Eurorack.

[Xavier]

@matrix12x
Please If you have time, let me know if you find some bugs with the latest Unison branch.
Thanks

[matrix12x]

Xavier,
I compiled it and played with it for an hour trying to find bugs and did not find any.
I read through the changes to the code as well. I see you fixed the items in common that caused the Engine2 page to also appear as the last page in MOD.

It works perfect. Great work. 

Ive been making all sorts of new sounds with unison since yesterday. With unison, the PFM2 sounds soooo fat, with such a beautiful beating sound.

[Xavier]

Great !
Thanks a lot for testing and for the feedback

I'll include a couple of fix that i found while developping the pfm3 (poly note choser, small midi note (<4ms) are not played), and then create the zip.

[Xavier]

Here is a link of a compiled version for those who wants to give a try :
https://drive.google.com/file/d/1IGHz_fjZTfmmGBk1qzq7rgYY6UQ9ADMG/view?usp=sharing

It's called 2.20t, not sure yet if that will be 2.21 or not.
CV/Eurorack version is not tested at all, as i don't have any Eurorack where i am for the week.

[matrix12x]

I tried out the 2.20t that you compiled and found no issues. Again, thank you for all your work on this great synth.

[vurt]

Tried the CV version with my eurorack. Playing from a euro sequencer it sounds like unison mode (set to 8 voices) is 1 fixed pitch for all notes, just getting a short clicky sound. If i change to poly and press a key on the sequencer then first press is fixed tone, second press (same key) is the real pitch.

Tried it with MIDI and it sounds pretty good I do agree with Solipsvs that randomness to pitch could do wonders.

[Xavier]

Thanks for the feedback.
I need to try with the Eurorack version but that won't be before this week end.

I don't want to add any new parameters in the UI.
I'll see if i can add randomness somewhere to mitigate the phasing effect.

[vurt]

Cool

Here's a video of the bug i'm getting, if you need it.
https://www.youtube.com/watch?v=4VZKEjcxP7U&feature=youtu.be&ab_channel=Vurt_NexusMods

i can get some notes if i use Slide Output from my sequencer. For some reason.

[solipsvs]

one way to mitigate the phasing might be to stagger each voice, even randomly, (by less than one phase of the root)  all they need is to be out of phase i think.  actually manipulating the pitch gives a very interesting knotty warbling.

[matrix12x]

@solipsvs I believe this just sounds like typical unison behavior. Also, we may be making different types of sounds.
The analog synths that I have that can do unison have this phasing sound to it. I also think my Access Virus (digital synth) has this effect when using unison. Personally I like it. What is an example of a synth where you don't hear the phasing when using unison? Would you mind posting a sample? I would love to compare. I'm very much into new sounds.

However, even if the unison voices are out of phase, there will still be phasing because there is a phase difference between the voices. The reason unison sounds stick out is because of the beating.

Some suggestions for reducing the phasing during sound design are:
1) use more unison voices
2) don't do unison on low frequency sounds
3) try different waveforms. for example unison on a saw sounds is less likely to have this phasing than with a sine wave

[solipsvs]

ok heres an example from serum. one clip is where each layer starts at the same phase and then diverge creating a phasing sound and the next is an example of each layer starting with a random phase, creating no phasing sound.

[matrix12x]

Thanks. Good samples for the explanation.
 
The random start phase sample you posted does sound better.

Looking at the difference in waveforms in an editor, one can see that the in-phase sample has distinct peaks, whereas the random phase one has no discernible peaks.

I'm assuming each unison voice in serum starts at a random phase, e.g., not the pairs of unison voices (assuming an odd number of unison voices).

[solipsvs]

that was 16 layers and im not sure if they are actually random as much as just deliberately out of phase with each other.

[Xavier]

Here is the next beta Unison firmware for the preenfm2  2.20u
https://drive.google.com/file/d/104UNZ-R6j8ukwVbELv4xVnOZ2ii47ixH/view?usp=sharing

. CV/Eurorack should work much better
  I can still hear some weird random glitch in the sound. Not sure what it is (maybe the CV i send ?). Still have to investigate.

. Now the Unison Detune is between -1.0 and +1.0.
  With positive detune there is a different starting phase for each voices of the unison
  With negative number phases are all the same (as in 2.20t).
  The difference is not huge but i had the feeling in my quick tests that different phases sound a bit better specially with bigger detune value.