controller for DIVA

[PAK]

Gosh. What's your problem?

That automation doesn't work as expected. If you regard this as a Novation problem then at least do your (Novation using) customers the courtesy of contacting Novation to explain. Perhaps you'll have better luck as a developer, perhaps not. The least you can do is try and, if you do that, then a genuine "thanks".

If we split this parameter up into two indepoendent parameters (octave/semi) then there's no way to do such sweeps in automation tracks. There's also no way to use a modmatrix to sweep the full range.

I don't see why it couldn't be implemented via ways, already suggested previously, to maintain full range. If you're saying otherwise then "ok". No point in going over that again anyway, since you've made your position clear.

end of story.

Indeed.

[Howard]

Wow, do you seriously think that Urs is an asshole?

No. I'm just struggling to understand his reasoning.

Hmmm. Then don't be so quick to attack the one person who can give you that information, and even accuse him of being discourteous to his users afterwards.

[Echoes in the Attic]

Woah, lets back up a second here. I need to chime in.

Firslty, no need for name calling, but we need to stop pretending that this is only a Novation issue. All Novation automap does is assign controls based on exposed parameters, the same way many other hardware mapping solutions work. This is how NI Kore works, or Maschine and how Abletons instant mapping works for things like the APC40 or ipad controllers, and how DAWs mapping work (Logic's plug-in mapping, or Cubase or Studio one and other host based hardware mapping). This is the most common way to map controls to hardware besides midi. This is absolutely not unique to Novation.

I have to say that I am disappointed by how this works and haven't come across this problem with other plug-ins and I rarely say a bad thing about an u-he product.

But, it almost works in Novation! It is actually worse in other mapping types like in Ableton or DAW based mapping unless the mapping system lets you assign the parameter more than once and set the increments of change. In automap, you can come close. You can set one control to move in full octaves and another to move in a fine resolution and it does keep the relative offset of the fine control when you move the coarse control. The problem is when you get towards the end, it resets the fine, which is a huge pain.

So for the record, this affects all mapping systems with any kind of intelligent control. And these are far superior to midi mapping because they allow feedback. That is, the value is sent both directions so the APC40 or Maschine controller or Kore update when the value change in the plug-in. But this tune problem is worse in other controllers.

If there was some way that the fine tune wouldn't reset when a full octave change is sent and it reaches the end, that would solve. Urs, you must already have some trick in place that prevents the fine tune from changing when an octave change is sent, no? Could this trick not be improved to work when the end of the range is reached?

[PAK]

don't be so quick to attack the one person who can give you that information

I haven't been given any information which solves the issue, nor any information which changes my opinion that it's an odd way to implement the control. But I do respect that it's his software and his right to do things as he sees fit. All I can do is push for a solution. That avenue is now at an end, and I have asked Urs contact Novation on behalf of Diva users.

So, as far as I'm concerned the issue is closed..

we need to stop pretending that this is only a Novation issue.

I was aware this was likely to be an issue with other controllers - I can only speak of what affects me personally. If you feel you can get any further help on the issue then go for it, but I'm done discussing the topic on this forum..

[Urs]

I don't see why it couldn't be implemented via ways, already suggested previously, to maintain full range.

Then please tell me what to write on lines 1450-1573 of AM_AudioManager.cpp in order to make it work.

[PAK]

I don't see why it couldn't be implemented via ways, already suggested previously, to maintain full range.

Then please tell me what to write on lines 1450-1573 of AM_AudioManager.cpp in order to make it work.

Maybe Novation will help?

Ok that really was the last words.. cheers all and no hard feelings, even if it comes across that way it's not meant like that.

[Urs]

I haven't been given any information which solves the issue, nor any information which changes my opinion that it's an odd way to implement the control.

Ok. It's an odd way to implement the control. There you have it.

We've done something new. 2 controls for 1 parameter. Better usability for most users. I can't see how that is wrong.

[Bronto Scorpio]

I like the current solution
Reminds me of a software version of the Jupiter 6 "Wide" pitch mode or however it was called

Cheers
Dennis

[Urs]

I like the current solution
Reminds me of a software version of the Jupiter 6 "Wide" pitch mode or however it was called

Cheers
Dennis

That's possible too. Keeping the tuning a single parameter allows for different ways of control: 1 knob full range, 2 knobs semitone/cents, 2 knobs octave/semi, 3 knobs octave/semi/cents whatsoever. Or even 5 switches for octaves.

Splitting the parameter up into dedicated roles makes things less flexible.

[Echoes in the Attic]

I agree that the current method makes it flexible, but it would be nice if it would work with all automation based mapping systems.

Urs, would you mind addressing my question at the end of my last post? I fell that you must already have something in place to work around the issue, but it doesn't work when it hits the ends of the range.

[urosh]

Ok. It's an odd way to implement the control.

Not for me. Maybe put in manual "Hey people, it's ye good ol' vernier dial just like on your dads VCS-3. But this time it's actually readable and you can easily set octaves".

[Urs]

I agree that the current method makes it flexible, but it would be nice if it would work with all automation based mapping systems.

Of course it would.

Urs, would you mind addressing my question at the end of my last post? I fell that you must already have something in place to work around the issue, but it doesn't work when it hits the ends of the range.

Well, I can't distinguish host automation of drawn curves / intended values from automation data coming in from a control surface.

What's in place is a MIDI control mechanism where I can easily distinguish between different controllers that are assigned to the same parameter. This does not require any "knowledge" on the side of the controller.

So here's the difference. Even if multiple knobs could be assigned to the same parameter in the plugin automation world, it would be impossible for our plugins to distinguish them. Hence any separation in range and control can onbly be addressed on the side of the controller.

The only workaround I can see would be offering a range of "dummy parameters" that allows the user to individually map whatever he wants to control in whatever way. This however is quite a stunt, just like using wrappers to tamper with automation data inbetween host and plugin. We'll think about that next year maybe.

[Echoes in the Attic]

Thanks for the explanation.

Urs, now that I think about it again, it might be something that Novation could fix. I think I've identified the problem. I'm just reporting this in case you can think of any work around you can do or can suggest to Novation (If you wouldn't mind):
When you assign one control to two knobs, you must set a start and end range and increments for each. The fine tune is set as min -30 to max 30 with whatever increment you want. The Coarse tune though, in order to send octaves, has to be -24 to 24 with an increment of 12. This works perfect and leaves the fine tune offset alone as long as the overall tune value is between -24 and 24. So for example if your fine tune is +1 semi-tone and you move your coarse down 2 octaves, it leaves the fine alone because overall it goes down to -23 (still in range of the coarse knob). However if your fine tune offset was -1 and you go down two octaves, that wants to send it to -25 overall. BUT, since the coarse knob has a min value of -24, it actually stops it at -24, thus resetting the fine.

Unfortunately you can't set the min and max to be -30 to 30 for the coarse tune because then increments of 12 makes the steps be -30/-18/-6/6/18/30 which moves the fine all over the place instead of -24/-12/0/12/24, which leaves the fine alone.

The best workaround I've found is to set your coarse knob to be increments of 6 with a range of -30 to 30. Interestingly, when you send 2 steps in a row this way, it moves the fine on the first step and then moves it back to where it was after the second step. This lets you use the whole range, but you have to be careful not to go to -30 or 30 with the coarse or it will reset.

The idea of a dummy parameter which controls the coarse sounds interesting. I guess the other option would be to have a range of -36 to 36 with the 6 on each end be the same as -30 and 30, if that makes sense? So the values for 30 to 36 for example would not change the tune. It would not affect backwards compatibility, right? Just throwing things out there. Thanks for listening.

[xh3rv]

Those who use Automap with Zebra, Diva or Ace - (Diva coarse/fine tune aside ) is it worthwhile?

I would be afraid to try mapping Zebra. Diva and Ace are doable but it needs to be streamlined, it takes a lot of pages to get everything that's worth mapping in. Uhbiks are pretty nice candidates for Automap - most only need 2-3 pages.

It's only situationally useful and not a replacement for a mouse. It can add a lot of life to things that aren't alive otherwise. Some general cases are very precise balance on a single parameter or triangulating a balance between multiple knobs. In any case where the tactile nature of knobs can merge in my brain with the sound coming out of the synth.

Particularly on oscillator tuning, the mouse is generally a quicker choice for coarse tuning. When there's an inter-oscillator routing like sync or FM, however, harmonic relationships are altered. Coarse tuning ratios are no longer ideal, and fine tuning can be used for enharmonics. In this context - having to jump octaves is an obstacle; knobs are great because it's really tactile for exploring balance in the sound. So my opinion is that two hardware knobs might seem right on the surface but it's precisely the cases where knobs are nice to use that one parameter, with multiple knob resolutions, is the better design choice. Just IMO.

[e] The dream solution for me is Reason Rack Extension Diva here anyway

[Gonga]

PAK, I have a lot of respect for your knowledge, your contributions to the forum and to my knowledge. After reading through this thread it appears you were feeling frustrated that nobody was responding to your issue with a solution to your satisfaction. Maybe you felt frustrated, maybe you wanted to draw attention to your issue. Well, you did that, but I don't think your rudeness was called for, nor the best way to go about getting people's attention. There are other avenues such as PM or e-mail to the company, etc.

If we split this parameter up into two indepoendent parameters (octave/semi) then there's no way to do such sweeps in automation tracks. There's also no way to use a modmatrix to sweep the full range.

I'm no power user, but I'm now using Diva for my most important lead sounds, and the ability to automate sweeping the osc freq full range without any stepping is critically important for me. So from what I can tell it looks like I should be happy Diva is set up the way she is.