Printable Version of Topic

Click here to view this topic in its original format

Plugivery Forums _ GRM Tools _ 2d Joystick Bug Across All Plugins

Posted by: awelex Mar 5 2017, 03:08 AM

Hi,

I've been having a persistent issue for years now with all GRM plugins that use the 2D joystick as part of their interface. The problem is that the joystick often simply doesn't remember the last setting and jumps to a random position once I hit play. This happens when there isn't any automation on the track and the track is set to read (in Pro Tools). If, for example, I set the upper and lower frequency bound in Reson with the joystick and then hit play, about 70% of the time the joystick will jump to another, random position. This also often happens right when I open a session -- the parameters are already different from what they were when I saved the session. Very frustrating to say the least -- I have to resort to writing down all settings in the comment field of the track and printing the track once I feel I dialed in the right settings.

I'm on PT HD 12.5.1 on OS X Mavericks, but I've experienced this behavior on multiple workstations with different OS and PT versions.

Anybody else experiencing this?

Thanks,
Alex

Posted by: awelex Mar 5 2017, 04:14 AM

To follow up on this: I just wrote some automation with the joystick and when I played it back, it still jumped to another position about half of the time! I was hoping as a workaround I could at least write some automation to lock the parameters in.

I also inspected the written automation, and it appears that instead of writing the two parameters that the joystick controls (e.g. first/range in Delays, low/high in Reson) it writes two controls called "Center" and "Width". I wonder if the issue is related to that.

Posted by: Anatolyj Mar 5 2017, 08:55 PM

Hi awelex, i perhaps can not precisely understand all described due to my english, but i tried just now write automation in 2d joysticks (in realtime via computer mouse) in: bandpass, delays, reson, (tested both versions, stereo and mono), and it works propely for me. Well, i mean, automation written, and then i save project and reload it, and during play 2d joystick moves exactly along the path that I moved via mouse, without any random or unexpected movements.

Next

QUOTE (awelex @ Mar 5 2017, 05:14 AM) *
I also inspected the written automation, and it appears that instead of writing the two parameters that the joystick controls (e.g. first/range in Delays, low/high in Reson) it writes two controls called "Center" and "Width".

Hm, yes, really there is it. Well, I think the existence of an extra pair of parameters should not be a problem in this case. But the combination of these two pairs in recorded state can interact with each other, that obviously (since it impact to the same, but may be with different measures, i not sure). In any case, as i said, it works(automate) properly for me anyway, because my mouse movements affect to record of just one pair (ie "center" and "width").
Maybe really in your case somehow occur unexpected touches to the extra parameters? I think it is not difficult to detect through automation line.

(Added: the more parameters - the better smile.gif )

Tested on: Win7 64bit, Stidio One v3, plugins v370 vst 64it.

Posted by: awelex Mar 5 2017, 09:15 PM

QUOTE (Anatolyj @ Mar 5 2017, 11:55 AM) *
Hi awelex, i perhaps can not precisely understand all described due to my english, but i tried just now write automation in 2d joysticks (in realtime via computer mouse) in: bandpass, delays, reson, (tested both versions, stereo and mono), and it works propely for me. Well, i mean, automation written, and then i save project and reload it, and during play 2d joystick moves exactly along the path that I moved via mouse, without any random or unexpected movements.

Next

Hm, yes, really there is it. Well, I think the existence of an extra pair of parameters should not be a problem in this case. But the combination of these two pairs in recorded state can interact with each other, that obviously (since it impact to the same, but may be with different measures, i not sure). In any case, as i said, it works(automate) properly for me anyway, because my mouse movements affect to record of just one pair (ie "center" and "width").
Maybe really in your case somehow occur unexpected touches to the extra parameters? I think it is not difficult to detect through automation line.

(Added: the more parameters - the better smile.gif )

Tested on: Win7 64bit, Stidio One v3, plugins v370 vst 64it.


Anatoly,

That's the weird thing: the weird jumps don't register on any automation lane. Moreover, they also often occur when there isn't any automation on the track and the track is set to read. Strange... I've recorded a screencap and sent it to the GRM guys. If there is a way to upload it here, I can do that as well.

Posted by: Anatolyj Mar 5 2017, 09:28 PM

QUOTE (awelex @ Mar 5 2017, 10:15 PM) *
Anatoly,

That's the weird thing: the weird jumps don't register on any automation lane. Moreover, they also often occur when there isn't any automation on the track and the track is set to read. Strange... I've recorded a screencap and sent it to the GRM guys. If there is a way to upload it here, I can do that as well.

Hmm, ok i understand.
Really, would be good if you attach also here quick video for demonstrate this behavior, interesting to see, and i can make too.
About upload - very good service Mediafire, just upload video to it and share link.

Posted by: awelex Mar 5 2017, 10:10 PM

QUOTE (Anatolyj @ Mar 5 2017, 12:28 PM) *
Hmm, ok i understand.
Really, would be good if you attach also here quick video for demonstrate this behavior, interesting to see, and i can make too.
About upload - very good service Mediafire, just upload video to it and share link.


I've uploaded two videos. First one:

http://www.mediafire.com/file/8d57np6jtjmadu5/grm_bug_2.mov
This one demonstrates parameters jumping to new positions right after I open a session when I hit play.

http://www.mediafire.com/file/y6u6ycy7ob4m7i7/reson.mov
Here you can see the parameter jump to a new position every time I hit play. There is no automation on the track.


Posted by: Anatolyj Mar 6 2017, 12:49 AM

awelex Oh damn, catched, you right, there is some issue with saved state. video here: http://www.mediafire.com/file/gmekb4ghk1nbdqp/Reson_grm1.mp4
As can see, in Reson case it good reproduce for me only when i save joystick in upper right corner.
Also, can note that when we move joystick in this region, the "High" value not changes. Well, it seems some wrong offset in internal 2d positioning. (Will later test other plugs with 2d)

But what about automation (and unexpected joystick movements during play) hmm, still can't reproduce it. See in another video: http://www.mediafire.com/file/c5xq9y1jdyb8dnh/Reson_grm2.mp4

Perhaps you right, maybe this second behavior somehow related with first, but reproduced only in PT maybe.

In any case thanks for info.
I hope support will pay attention to it.

Posted by: grmtools Mar 6 2017, 01:53 PM

Hi Awelex and Anatolyj,

Effectively, there is a strange behavior with ProTools and 2D controller, with or without automation.
I will try to understand and fix the problem.

But, some other behaviors are normal :

Some parameters can be controlled by several controls. For exemple, low value in Reson is controlled by :
- low parameter
- center & width parameters
- preset and superslider
If all those parameters are automated simultaneously, results may be unpredictable.

In Reson, the 2D slider controls Center and Width parameters. If it is placed on the right, or on the left, different values of center and width can correspond to the same values of low and High.
So the displayed position is not the same, but the values of low and high are the same.

All the best,


Emmanuel
GRM Tools Support

Posted by: Anatolyj Mar 6 2017, 03:10 PM

QUOTE (grmtools @ Mar 6 2017, 01:53 PM) *
So the displayed position is not the same, but the values of low and high are the same.
GRM Tools Support

Hm ok thanks for info, then good. (Was such thought also, but was unsure in audible difference ).

Emmanuel why latency slightly different (slightly less) now compared with previous versions ? What was changed in fft engine ? Accidentally nothing was degraded nor simplified? laugh.gif dry.gif Just need to make sure.

Posted by: grmtools Mar 6 2017, 04:55 PM

no, nothing was changed. It is exactly the same fft code.
But latency depends also on the block size in your host application. Maybe you have changed something in the preferences.

Emmanuel

Posted by: Anatolyj Mar 6 2017, 05:07 PM

Damn, i forgot, really i increased card buffer. Then all ok, sorry.

Posted by: awelex Mar 6 2017, 10:15 PM

Thanks for the reply, Emmanuel. Were you able to reproduce it on your end?

Powered by Invision Power Board (http://www.invisionboard.com)
© Invision Power Services (http://www.invisionpower.com)