Charlatan³ Free VA synth updated to v3.3.2 (Win/Linux)

[markp]

Excited excited, looking for download link.......no macOS. Dead on the inside.

[firebolt41]

Is it possible to add CC2+34 (breath control 14bit) or cc11+43 (expression control 14bit) to the modulation input list?

Depending on the Daw, parameter links can be too slow to do articulations with a breath controller.(Ableton is terrible with this, where CC -> param is either quite high smoothing (latency + destroying articulation) or aliasing galore.) Its not ideal to override cc1/modwheel with Breath because I'm using both modwheel and expression input simultaneously.

[lolilol1975]

Excited excited, looking for download link.......no macOS. Dead on the inside.

If it was open source, you'd get a Mac version.

[Korg Supporter]

Any chance the envelope attack could be made a touch less neutered? As it is, it lacks a proper click even with the fastest curve.

I'm aware of this. Unfortunately, there is a limit on how fast a modulation destination (such as volume) can react to changes of the envelope generator due to the way modulation signals are processed. Currently, modulation signals are generated at one 16th of the audio samplerate. As a consequence, an instant attack stage will result in a ramp of 16 samples or ~0.4ms @ 44.1kHz. While this should be easily fast enough in the majority of use cases, it does make it hard to achive sounds where a bright, clicky attack is desired, like 808-style kick drums. There are however some tricks that can help to get you mostly there:

- Engage oversampling and/or increase host samplerate. These will increase the modulation processing rate and hence sharpen the transients.
- Make sure that the oscillator start phase is set so that the oscillator signal starts at maximum amplitude. E.g., for a sine wave, choose 90°.
- The attack click can be emphasized by activating saturation in the master FX section and quickly increasing the gain during attack phase using a very short envelope.

Anyway, I'm thinking of doubling the modulation processing rate in the final 3.0 release. Initial measurements have shown a ~10% increase in CPU load, which may be affordable.

Maybe that would be an optional feature.

[wikter]

On purpose, bypassing the Attack stage it could be possible. There are more than ADSR EG's out there:
AHDSR
D
AHDSD'S'R
HDSR
AD
HR
DADSR
LADSR
ADDSR
MSEG

But since there's a logaritmic curve for attack and u can use an lfo as envelope by bypassing the ADSR to amp...

[wikter]

About fast attack:
Use standard ADSR, set speed multiplier to 16x, set attack to 0, envelope shape to LOG(?) (the faster shape), enable oversampling. At 44.1khz, the complete attack time is around 100microseconds, 0.1 milisecond, it just takes 2 samples from 0 to top level, there are 5 low level samples previous to the gain step.

Comparing with NI Massive that takes up to 16 samples, I think Charlatan V3 is a serious tool for such job.

[gentleclockdivider]

Bug or not ?
As it is right now the lfo can go pretty damn fast when ' selecting 10Hz fro the drop down menu and set rate to max .
But it can go faster : set lfo to max rate , and modulate the rate with an evelope set to full sustain .
This will push the lfo rate even further .
This seems a bit inconsistent

[brok landers]

Bug or not ?
As it is right now the lfo can go pretty damn fast when ' selecting 10Hz fro the drop down menu and set rate to max .
But it can go faster : set lfo to max rate , and modulate the rate with an evelope set to full sustain .
This will push the lfo rate even further .
This seems a bit inconsistent

I use that feature quite often - also for the amp of the osc's (which then drives the filters more - really necessary with max resonance). So i'd stronly vote for NOT talking away such features. Being able to enhance parameter ranges often is very welcomed here.

[gentleclockdivider]

OH I don't want the feature to be taken away at all .

[Odd Fella]

About fast attack:
Use standard ADSR, set speed multiplier to 16x, set attack to 0, envelope shape to LOG(?) (the faster shape), enable oversampling. At 44.1khz, the complete attack time is around 100microseconds, 0.1 milisecond, it just takes 2 samples from 0 to top level, there are 5 low level samples previous to the gain step.

Comparing with NI Massive that takes up to 16 samples, I think Charlatan V3 is a serious tool for such job.

Does the human ear even "go so fast"? I read some time ago that the human ear has a "resolution" of no less than 5ms, so what is the point of such crazy attacks?!

[FrogsInPants]

Does the human ear even "go so fast"? I read some time ago that the human ear has a "resolution" of no less than 5ms, so what is the point of such crazy attacks?!

You won't hear pulses separated by less than that as separate events, but that doesn't mean faster modulations don't have audible effects. Such fast modulations necessarily introduce frequencies well into the audible range, heard as a click or snap on a fast attack.

[wikter]

Does the human ear even "go so fast"? I read some time ago that the human ear has a "resolution" of no less than 5ms, so what is the point of such crazy attacks?!

5 ms is slow enough to feel while playing a fast rythm or musical phrase, I mean that's timing. But on the timbre side 5 ms is a very long time: it's like 200Hz, that said, the difference between a sinewave & a squarewave with a 5kHz tone is still easy to be discerned, so in 0.2ms you have some small pieces that make the difference. Ear is supposed to be sensible up to 20kHz. That means 50 microseconds, 0.05ms --- But of course, it's hard to cope with frequencies higher than 10k and identify waveshapes at that rates is very hard even for trained ears.

[wikter]

This will push the lfo rate even further .
This seems a bit inconsistent

This is also hapening with Filter cutoff. Setting it to its highest frecuency you can get a more open sound using the envelope. In fact I've seen that althoug in Low & mid frequencies, the resonating frequency is the same as the filter shows, with frequencies higher than 8k that starts to drop.
Filter cutoff frequency displayed/Resonating frequency shown in analyzer:
1050/1050
1775/1750
3430/3430
10200/9700Hz
12050/11510Hz
14600/13580Hz
15900/14600Hz
22000/18790Hz

[brok landers]

OH I don't want the feature to be taken away at all .

Hahaa, good then.
To me it kinda sounded like that.

[Koldunya]

a YouTuber ('thatpluginguy' or something) posted something about this, and while I absolutely do not need another synthesizer, I liked the UI and took it for a spin, and damn, I love this synth, now. It's pleasing to look at, easy to get around and figure things out, and sounds great

Thanks for sharing!