MPowerSynth improvements & feature requests

[MeldaProduction]

Voice limit - I did not. Why would you want that? To me it seems like really useless these days. Last thing I would want is for the notes to start disappearing...

Filter frequency by note - well, it won't be possible for sake of backward compatibility, but anyway why on earth??? The whole hearing range is about 10 octaves! 16?????????

[CableChannel]

Voice limit - I did not. Why would you want that? To me it seems like really useless these days. Last thing I would want is for the notes to start disappearing...

EVERY advanced synth in existence has this. Complex patches with long release tails will kill any CPU on earth when voices build up. Cutting down release tail is not an option because it just sounds different. So for pad sounds, one just needs this. You just can't deny this one, come on

Filter frequency by note - well, it won't be possible for sake of backward compatibility, but anyway why on earth??? The whole hearing range is about 10 octaves! 16?????????

As I mentioned, to program a filter cutoff that does not go up with higher notes but which goes down. If you have a sound that has lots of high frequencies, and these high freqs sound good in the bass note range to give a pristine sound, but they are too harsh when playing high notes, then it would be highly useful. And to get this you need at least more than -8, because -8 just compensates the normal filter which always goes up with the note in MPowerSynth. -9 or -10 would overcompensate and actually mean the cutoff frequency is going down for higher notes.
Since this is a pretty new feature, maybe backwards compatibility is not so important?

[MeldaProduction]

Ok I'll add the voices to the to-do list.

But about the filter range, sorry but that won't happen. Backward compatibility is more important than this. Since the hearing range is about 10 octaves it just doesn't make sense sorry. You'd need to play the lowest lowest keys and have some really extreme filters or I just don't get what you are up to. Now there's the "Constant frequency" thing, that should help you too (from what I understand).

[CableChannel]

Ok I'll add the voices to the to-do list.

Thanks! Much appreciated!

But about the filter range, sorry but that won't happen. Backward compatibility is more important than this. Since the hearing range is about 10 octaves it just doesn't make sense sorry. You'd need to play the lowest lowest keys and have some really extreme filters or I just don't get what you are up to. Now there's the "Constant frequency" thing, that should help you too (from what I understand).

It doesn't have to do anything with the hearing range, I tried to explain I think I didn't explain well enough. I need a charboard I'll make a small sketch and scan it

[dmbaer]

Now there's the "Constant frequency" thing

OK, sorry but I need to ask what is maybe a dumb question: what do you mean by the "constant frequency" thing?

I just want to make sure I'm not missing something important.

[CableChannel]

Now there's the "Constant frequency" thing

OK, sorry but I need to ask what is maybe a dumb question: what do you mean by the "constant frequency" thing?

I just want to make sure I'm not missing something important.

Resolved by email. It's the constant frequency thing

For a filter you can enable constant frequency. This disables the default behaviour where the cutoff frequency is linked to the note. Then you can dial in any slope you like for the filter/note dependency.

[CableChannel]

Will MAmp be added to the FX matrix of MPowerSynth? That would be great.

[MeldaProduction]

Will MAmp be added to the FX matrix of MPowerSynth? That would be great.

Yep it will .

[CableChannel]

If LFOs and velocity controlled MIDI assignments were polyphonic (i.e. per voice), then the fun really would start! Is that at all conceivable? Maybe some time in 2015 MPowerSynth 2.0?

Oh, and the tremolo for the filter seems to have a pretty odd shape. Can't believe it is sine? It sounds more like sine^4 or something like that. Any chance to make the shape configurable?

[Breeze]

If LFOs and velocity controlled MIDI assignments were polyphonic (i.e. per voice), then the fun really would start! Is that at all conceivable? Maybe some time in 2015 MPowerSynth 2.0? ...

Not to criticize our esteemed dev, but yes: these are kind of bread and butter items. Personally considering all the modulation that is already possible, I can understand the weight of enabling poly LFOs (sorry CableChannel!) but MIDI velocity is a real goto for adjusting playability. Categorical denial seems a little odd. Give me BIG PRESETS, I don't care!

[CableChannel]

but yes: these are kind of bread and butter items.

I agree. Currently MPowerSynth has a lot of advantages over other synths but also this big disadvantage of exclusively global LFOs/ENVs/velocity-assignment, and the non-routable filters. If the LFOs would be per voice and filters would be serial/parallel then this would really unlock everything that is limited now. There would be just advantages and no more disadvantages. The competition would be crushed

[CableChannel]

I am noticing that I throw way too many requests at Vojtech which he all parries in a very experienced way Anyhow, while building patches and comparing synths and so forth I kind of collected the biggest shortcomings of MPS which I already shared, the missing parallel filter routing and the non-polyphonic LFOs. Now two other rather major things keep striking me:

- The monophonic mode often stops playing when I play a quick sequence on my keyboard. Don't know whether this is related to my midi system or DAW host, though. Could be a latency thing.

- The quality of the reverb. Its rather grainy and metallic. Don't know whether that could be improved. Even the convolution reverb has a grainy metallic tendency. Putting a little hishelf damp helps, but it's always a little difficult. Modulation should help, but it doesn't really.

When I try to rebuild Virus patches on MPowerSynth, most of the time the parallel filtering or the reverb is the stumbling block. Also, the unison of the Virus is after the oscillators, so you can mix 3 osc's and then apply unison voicing. This sometimes makes a big difference, but I would accept this as different design decision.

[robertszalapski]

The reverb needs tweaking to get right, the default settings are rarely what you want. Try multiple reverbs seperate from the main signal, in series, with a smaller size on the first one. It gets it better anyway.

[MeldaProduction]

Well, I understand the limitation of the global modulators, but I'm afraid it has to stay like this. There's just no way, to make it a different way, sorry. For this synth it has to stay like this. There are at least the filter parameters available per-note/per-velocity and after all there's not so much to modulate otherwise, pitch probably, and that seems to me like more for drums, which this is not designed for really. And most sections have a dedicated ADSR so it should cover most needs.

[CableChannel]

@Robert: that is a very good idea!! I will try that. I think the random generator once created such a scenario and I remember the reverb sounded really smooth.

@Melda: You are right. So it might not be neccessary to rebuild the whole LFO architecture but just add a few more functionalities. As you say filters are the most important per-voice modulation targets. Maybe it suffices to add a dedicated ENV/LFO for two more targets: pitch and pulsewidth. Pitch is also important for pads and leads, a pitch env just adds the neccessary spice to many sounds. If you could make the filter tremolo a little more flexible (i.e. a few shapes to select from, or even a full blown shape editor like in the LFOs) that would already help a lot. Then add the same to pitch and pulsewidth and 95% of the usual cases are taken care of.