PDA

View Full Version : Color order and e681 tested.



angus40
08-18-2012, 03:01 PM
A lesson I have learned . While testing some Lpd 6803 strip usin the e681 controller .

Test mode 1-3 will show the strips true color order

test 1 expected color Red>>>> actual color Blue

test 2 expected color Green >>> actual color Red

test 3 expected color Blue >>> actual color Green

Color order confirms as BRG


Changing the color order in hls does not change the output color order of the e681

this must be done in the controller config

Command >>>> RGB *(cluster) * ((color order) 0=RGB, 1=RBG, 2=GRB, 3=GBR, 4=BRG, and 5=BGR )

Changing the color order in hls is not required leaving it as rgb with the e681 ( in my case ste to brg ) was all that was needed for correct color order output.


I hope this may help others save some time i wrestled with this for several hours before the light went on :)

Richard

mschell
08-19-2012, 10:39 AM
Richard,
I would only slightly disagree with you.

The E681 will output the channels you tell it to. However, in the latest firmware, Jim has done some things to try to get more colors out of the 6803, I believe, by having the LEDs be on one color for part of the time, and another color the rest. I don't remember the actual number of colors a 680x pixel can do, but it's like 5-bit color instead of 8-bit color that the 2801 and others can do. By changing what color is output quickly, you can fool the human eye to see a different color, somewhere between the two.

So that will complicate things if the color order is not set correctly in the E68x, but only on lights that use the 6803 chip.

So HLS could change the color order itself, and it would look OK, IF the E68x wasn't trying to fool you.

Just wanted to make this point, since if someone is NOT using the 6803 pixels, changing the color order in HLS should do what you want.

JHinkle
08-19-2012, 12:13 PM
Richard,
I would only slightly disagree with you.

The E681 will output the channels you tell it to. However, in the latest firmware, Jim has done some things to try to get more colors out of the 6803, I believe, by having the LEDs be on one color for part of the time, and another color the rest. I don't remember the actual number of colors a 680x pixel can do, but it's like 5-bit color instead of 8-bit color that the 2801 and others can do. By changing what color is output quickly, you can fool the human eye to see a different color, somewhere between the two.

So that will complicate things if the color order is not set correctly in the E68x, but only on lights that use the 6803 chip.

So HLS could change the color order itself, and it would look OK, IF the E68x wasn't trying to fool you.

Just wanted to make this point, since if someone is NOT using the 6803 pixels, changing the color order in HLS should do what you want.

Mark:

I think what Rich is saying is that he got some pixel strings that are BRG - not the typical RGB.

HLS is always delivering RGB as channels 1,2,3 - so if the pixel string expects the channels differently (different order) - then the controller needs to align the correct color channel with the proper string channel.

If I'm not interpreting Rich's post properly - please explain.

Thanks.

Joe

angus40
08-19-2012, 12:22 PM
You're correct Joe .:)

The color order as defined by the element manufacture and type must be configured in the controller .

hls is left at rgb .

this was a well learned experience yesterday .

in my case the 6803 color order is brg set in the e681 and output is correct.

Cheers

mschell
08-19-2012, 01:59 PM
No issues with the final result.

I guess my comment was built on the phrase "Changing the color order in hls does not change the output color order of the e681"

If HLS always outputs RGB, then you can't change the color order in the software, and therefore, of course, you need to change it in the hardware.

In other sequencing software that I have used, you CAN change the color order in the software, and so it's not absolutely mandatory to change it in hardware.

Back to your regularly scheduled program...

angus40
08-19-2012, 09:33 PM
hls does have the ability to set the rgb channel color order. however while testing with the e681 the color order visible on the pixels would never match my sequence

until i set the color order correctly on the specific cluster wintin the web config.



i will need to test the color order on some dumb rgb to see if this occurs also .

stay tuned , unless some one has tested and can verify .

mschell
08-19-2012, 11:17 PM
hls does have the ability to set the rgb channel color order. however while testing with the e681 the color order visible on the pixels would never match my sequence

until i set the color order correctly on the specific cluster wintin the web config.



i will need to test the color order on some dumb rgb to see if this occurs also .

stay tuned , unless some one has tested and can verify .

If you decide to test it with pixels, try it on something other than the 6803s. As I mentioned in my original reply, the E68x does color interpolation on the 6803s to get more colors than the IC inside the pixel/strip can generate.

I believe that if the color order is different from RGB on the pixel, and you change the color order in HLS, and DON'T change it in the E68x, then it will come out right on pixels/strips that don't use the 6803.