Author Topic: Testing of "Linear" brightness of my P10 panels  (Read 3674 times)

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,190
  • Kudos: 124
    • Granbury Christmas Lights
Testing of "Linear" brightness of my P10 panels
« on: June 11, 2015, 11:59:38 AM »
I was curious. Created a Nutcracker sequence with the same 4 color spiral, just dialing down the Intensity value.
I then used a Text Layer to display that value on screen. I know the camera compensates some, but it is pretty accurate.
The Nutcracker preview window is "accurate" per the displayed percentages, but not when compared to Real World.

I may run my P10s at 60% because it will reduce my current draw with no apparent change in brightness to the viewer.

https://youtu.be/UI-A341qdBA

Offline jnealand

  • Developer
  • ******
  • Join Date: Aug 2013
  • Location: Kennesaw, GA
  • Posts: 2,903
  • Kudos: 68
Re: Testing of "Linear" brightness of my P10 panels
« Reply #1 on: June 11, 2015, 02:59:07 PM »
That was a great test.
Jim Nealand
Kennesaw, GA all Falcon controllers, all 12v Master Remote Multisync with Pi and BBB P10 and P5

Offline wirekat

  • Newbie
  • *
  • Join Date: Apr 2014
  • Location: East Kent, WA
  • Posts: 18
  • Kudos: 0
    • Christmas In Kent
Re: Testing of "Linear" brightness of my P10 panels
« Reply #2 on: June 11, 2015, 03:23:26 PM »
Excellent.
This will help me a lot!
Kevin Thomas
LOR + Lynx + DMX + Falcon = Win

Offline JerryPlak

  • Supporting Member
  • ******
  • Join Date: Jul 2013
  • Location: Sunbury, OH
  • Posts: 599
  • Kudos: 2
Re: Testing of "Linear" brightness of my P10 panels
« Reply #3 on: June 11, 2015, 04:02:10 PM »
Excellent.
This will help me a lot!


Same for me  ;)
Jerry Plak

Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,190
  • Kudos: 124
    • Granbury Christmas Lights
Re: Testing of "Linear" brightness of my P10 panels
« Reply #4 on: June 11, 2015, 06:13:15 PM »
Here is the XML file for the test sequence.
It only has one Model called Matrix. When you open it, just let it Re-map to your matrix (of whatever type or size)


Offline JonB256

  • Supporting Member
  • ******
  • Join Date: Mar 2013
  • Location: Granbury, Texas
  • Posts: 5,190
  • Kudos: 124
    • Granbury Christmas Lights
Re: Testing of "Linear" brightness of my P10 panels
« Reply #5 on: June 11, 2015, 07:04:38 PM »
And - this might be good for the Wiki

If you are using a P10 Matrix, and everything is working nicely, you should consider going to Status / FPP Settings and turn off Logging or reduce it heavily. I was getting an occasional and non-repetitive "pause" during playback.

Turned of the logging - pauses are gone and the Matrix is smoother than ever.

Offline dkulp

  • Moderator
  • *****
  • Join Date: Sep 2013
  • Location: Framingham, MA
  • Posts: 1,505
  • Kudos: 87
    • KulpLights.com
Re: Testing of "Linear" brightness of my P10 panels
« Reply #6 on: June 11, 2015, 07:42:57 PM »
I was curious. Created a Nutcracker sequence with the same 4 color spiral, just dialing down the Intensity value.
I then used a Text Layer to display that value on screen. I know the camera compensates some, but it is pretty accurate.
The Nutcracker preview window is "accurate" per the displayed percentages, but not when compared to Real World.

I may run my P10s at 60% because it will reduce my current draw with no apparent change in brightness to the viewer.

https://youtu.be/UI-A341qdBA

I *just* got one of my panels working and can definitely see what you mean.   Above 60% there is no way to tell the difference.   

The problem in xLIghts is if you set the "brightness" of the model to 50% or something, then even on screen while sequencing, it's dim in the previews.   I'm thinking of add a "Display Brightness" to the Appearance dialog to allow you to specify and adjust that is made when doing the previews.    Thus, the model can be set to 60%, but then set the Display Brightness to 150% or so and it should then be nearly full brightness for the preview, but the values stored in the fseq file would be the 60% dimmer value.   

Need to think about this some more. ....   Now that I have a working panel, something to play with...  :-)


Offline AussiePhil

  • Sr. Member
  • ****
  • Join Date: Dec 2014
  • Location:
  • Posts: 495
  • Kudos: 11
Re: Testing of "Linear" brightness of my P10 panels
« Reply #7 on: June 11, 2015, 08:01:54 PM »
the same applies for all the pixels we use as well by the way..... the fact that none of the software we use allows us to change the levels in real time to see this makes it hard to identify.
I have run my show starting at 60% in 2012/2013 and ramping down to 40% late at night

HOWEVER all power supplies and cables should always be calculated and provisioned for 100% brightness

Offline gadgetsmith

  • Hero Member
  • *****
  • Join Date: Apr 2014
  • Location: Charlton, NY
  • Posts: 988
  • Kudos: 14
Re: Testing of "Linear" brightness of my P10 panels
« Reply #8 on: June 12, 2015, 05:54:06 AM »

I'm thinking of add a "Display Brightness" to the Appearance dialog to allow you to specify and adjust that is made when doing the previews.    Thus, the model can be set to 60%, but then set the Display Brightness to 150% or so and it should then be nearly full brightness for the preview...

How about keeping the preview at 100% regardless of the effect brightness setting? It would be handy to have the model preview window display the actual brightness setting in the corner of the display pane to know what the actual brightness is set to.

What algorithm is being used for brightness control on the preview window?  Perhaps it could be changed to better reflect what is being seen with the actual brightness of the display?

 Just thinking out loud here.

I had this issue last year when making some sequences for after 9pm.  I wanted things dimmed to 30-40%, but just can't see it in the preview window, so I had to sequence at 100, then the last thing I did was to set model brightness.

Offline dkulp

  • Moderator
  • *****
  • Join Date: Sep 2013
  • Location: Framingham, MA
  • Posts: 1,505
  • Kudos: 87
    • KulpLights.com
Re: Testing of "Linear" brightness of my P10 panels
« Reply #9 on: June 12, 2015, 08:13:49 AM »
How about keeping the preview at 100% regardless of the effect brightness setting? It would be handy to have the model preview window display the actual brightness setting in the corner of the display pane to know what the actual brightness is set to.

The preview displays the actual data that is found in the fseq file.   Thus, the model brightness and the effect brightness has already been  applied.  Those are taken into account at render time.

I just pushed a new setting to the the Appearance for the models to allow the preview to "brighten" the model while displaying.  Thus, the model can be at 40% which would put 0-102 as the channel values in the fseq, but you can set the preview brightness to 200 or something to make it brighter on the preview.

Offline dkulp

  • Moderator
  • *****
  • Join Date: Sep 2013
  • Location: Framingham, MA
  • Posts: 1,505
  • Kudos: 87
    • KulpLights.com
Re: Testing of "Linear" brightness of my P10 panels
« Reply #10 on: June 12, 2015, 08:23:59 AM »

That all said, I'd greatly prefer if FPP had channel dimming curves built in.  :-)

Last year, when I had to dim one of my models to be able to video record it (the LED's were overwhelming the sensor), I hacked the FPP source to dim that one channel range.   That was easier than going through all the sequences in HLS, setting the dimming curves, re-compiling for show, converting in xLights, overlay the pixels, re-convert to fseq, then re-upload.   Lots of steps.   10 lines of code or so in FPP handled it as a temporary measure.

Offline CaptainMurdoch

  • Administrator
  • *****
  • Join Date: Sep 2013
  • Location: Washington
  • Posts: 9,837
  • Kudos: 214
Re: Testing of "Linear" brightness of my P10 panels
« Reply #11 on: June 12, 2015, 10:46:35 AM »
That all said, I'd greatly prefer if FPP had channel dimming curves built in.  :-)

It's on the TODO list.  I also want to allow you to control this in the schedule/playlist so you can run the same playlist from 5-10 but have it get dimmer at various times as the night progresses.  My idea is that FPP would cache the dimmed values so it's only a performance hit the first time through a sequence, but that optimization would come later once we had the ability to dim.
-
Chris

Offline tbone321

  • Hero Member
  • *****
  • Join Date: Oct 2014
  • Location:
  • Posts: 1,584
  • Kudos: 50
Re: Testing of "Linear" brightness of my P10 panels
« Reply #12 on: June 12, 2015, 01:37:24 PM »
Is it possible to identify the light or node type and set the appropriate dimming curve in the preview window?  Some of us still have icans and standard LED strings connected to an LE in our displays along with the RGB stuff.

 

Back to top