LinuxMusicians

For now I've pushed a ADSHR envelope to XPolyMonk.

@khz I'm unsure if FM synthesis will be suit into it.

Sounds pleasantly musical.
Now an (volume) amplitude envelope (ENV AMT).
Operator / Formant Shaping synthesis would be an enrichment regarding 'SFormantModelBP' speech synthesis, if it is technically feasible.

<EDIT>Would Formant Shaping - FS Synthesis be technically possible?</EDIT>

Maybe possible.
Just for now I'm off for some other stuff I've to do. It may take some weeks before I could go back to play with the XPolyMonk sources.
Thanks so far for your input and suggestions.

okay, I've added the ENV AMP.

Even better to play. The vibrato is interesting. You can elicit beautiful sounds from the instrument!
With less effect (12/19) switched on it sounds almost better than with more effect? Can you set the effects in XPolymonk differently, different presets?
A full ADSR/ADBSSR/... - AMP would be more fine/musical to regulate. This gives you more points of intervention for the sound creation.
In general both envelopes could be accessed faster and thus intervene more in the musical process. When the envelopes are closed there should be no sound.

Then still 5 - x Formant Shaping Presets. . && ...

Okay, it seems now we talking about the delay. In Xmonk I use a fixed delay to compare a tad better with DelayLama. In PolyMonk, now that we've a more fine graduated ADSHR envelope, the delay is more disappointing then helpful. I'll remove it to give you more control over the sound generation.

Yeah, it was the delay. Much better now without.
# Better (several) sliders for finer fading of the individual SFFormantModelBP "some Formant Shaping".

I've removed the delay now, and implement a better algorithm for the release. Now, when release is zero, the gate will close immediately.

WOW
Without delay the effects are musically useful in XPolymonk again.
The envelope is much better. But you can still hear a sound when the envelopes are completely closed. The ENV AMP would be musically more finely controllable with an - e.g. - ADBSSR. This way the instrument can be made to speak even more differentiated.

# I hope I'm not annoying you, if I am, put me on your ignore list.

tramp wrote:I've removed the delay now, and implement a better algorithm for the release. Now, when release is zero, the gate will close immediately.

Thx. That's the way it should be.

khz wrote:WOW
Without delay the effects are musically useful in XPolymonk again.
The envelope is much better. But you can still hear a sound when the envelopes are completely closed. The ENV AMP would be musically more finely controllable with an - e.g. - ADBSSR. This way the instrument can be made to speak even more differentiated.

# I hope I'm not annoying you, if I am, put me on your ignore list.

Nope, I appreciate your input. This is the first synth I made, before that I'm more into the effects then into synthesis. So a lot stuff is new to me, and your input helps me to get a deeper understand of what is needed to make a synth useful.

A ADBSSR wouldn't be so hard to implement, now, than the basics been in. Just as I said some posts ago, I'll be busy with other stuff for the next weeks, so properly that wouldn't happen in the next time. (except someone fork and make it)

bluebell wrote:Thx. That's the way it should be.

thanks for feedback, bluebell.

regards
hermann

An envelope - all areas of the ADSR - should be fast so that it can "smack" (zap). Of course it can also decay for a long time and ... .
The envelopes shape the sound 100%, the sound process, ..., intervenes 100% in the sound.
I have no idea if an ADBSSR envelope is useful for the Formant (!Shaping!) synthesizer. Musically interesting I imagine it in combination with an ADSHR (?)filter envelope(?).

Is it possible to use the vibration frequency of the vibrato - in addition to controlling the modwheel - as ~LFO? The vibrato ~LFO (small +/- range) has only a minimal influence on the modwheel, only as much as it sounds has a minimal influence on it. The position of an (external) software/hardware modwheel command is detected and processed as usual. The vibrato ~LFO orients itself to the current position of an (external) software/hardware modwheel command and influences the modwheel in the minimum range.

<EDIT>Aren't knobs more space-saving?

<2>One voice (of the 12 existing) consists of 2 Formant oscillators (*1). Each oscillator has its own volume control.
The 2nd oscillator is detuned with the Detune control against the 1st oscillator. Possibly: vibrato effect only affects the 2nd oscillator.

(*1.) Would an suboscillator and noise oscillator to the 2 formant oscillators be useful?</2>
<3>Synthesizer Wiki - Sound Synthesis - Synthesizer & Synthese (LFO, VCO, FM, Filter..)</3></EDIT>