Regression in 4.11 - dimmer control on RGB Matrix patterns seems to be broken

[Majik]

I was using version 4.10.5b until recently, with a LEDJ Pixelbar (my own fixture definition).

This worked fine when using RGBMatrix definitions on the function manager, such as "Fill From Centre", etc.

I've recently installed 4.11 and this no longer seems to work correctly if the "Dimmer control" option is selected

Some more details:

The Pixelbar has 8 x RGB panels (heads) in a row. There are various modes available on this fixture offering varying degree of control, but I am using it in the 27-channel mode. In this mode the channels are laid out as follows:
CH1: Mode control & presets
CH2: Master level/sensitivity
CH3: Strobe control
CH4,5,6: Panel 1 RBG
CH7,8,9: Panel 2 RGB
CH10,11,12: Panel 3 RGB
CH13,14,15: Panel 4 RBG
CH16,16,18: Panel 5 RGB
CH19,20,21: Panel 6 RGB
CH22,23,24: Panel 7 RGB
CH25,26,27: Panel 8 RGB

The Fixture definition I have created (attached) has these channels and modes defined, along with definitions for the heads in each mode. For the 27 channel mode the heads are defined as above (e.g. Head 1:4,5,6, Head 2: 7,8,9 etc.)

In 4.10.5b, when I created an RGBMatrix Pattern, it worked correctly, putting channel 2 (Master dimmer) up to 100% and then only altering channels 4-27

In 4.11.0 if I select the pattern option "Dimmer control" in "Other controls", then when the pattern runs instead of putting the master dimmer up to 100% and leaving it alone, it continually switches the dimmer channel between 0% and 100% creating a flashing effect. The pattern in which it does this seems to be related in some way to the pattern.

I have attached a workspace definition which can be used to observe this.

cheers,

Keith

[janosvitok]

Hi,

the idea was that dimmer is 0% when the color is black and 100% otherwise. This was requested by some user here to maintain backward compatibility.
The problem here is that you have 8 heads (colors) controlled by single dimmer and the result now seems indeed random.

Correct behavior would be that dimmer is 0% only if all dependent heads are black. I can't promise any time frame for the fix though.

Thanks for reporting.

TL;DR: confirmed bug.

Jano

[Majik]

OK, thanks. I can probably work around it for the moment.

Cheers,

Keith

[janosvitok]

Workarond for now: create scene with all Dimmer channels @full, start it before the matrix, and uncheck dimmer control. That way dimmer and RGB channels will be controlled separately.
You may assign the dimmer scene either to a button or a slider in virtual console.

If you can compile from sources, change line https://github.com/mcallegari/qlcplus/b ... x.cpp#L790

from

Code: Select all

fc.setTarget(col == 0 ? 0 : 255);

to

Code: Select all

fc.setTarget(255);

[mcallegari]

To be honest I would like to remove the dimmer control from the RGB Matrix function. Its purpose is to control color channels, not dimmer channels.
It's been a bad idea to add it and now we have to face all these kind of issues and workaround them.

In fact, I have already removed it in the QLC+ 5 UI.

[mgweber]

The only time I check the dimmer box in RGB functions is when I am testing them for a look. I almost never run them at 100% for a show. It is just a way to get a feel for how they look with the speed settings as I am configuring them. Then it is uncheck the box and bundle them in a collection with a scene with an appropriate intensity level for the various fixtures.

[Majik]

Great, thanks for this. I'm not using this setting now.

If it's not working as intended and is causing issues, then I would agree it's best to remove it.

Cheers,

Keith