More strange test results

[jnealand]

I have a pixel tree made from 8 TM1809 flexstrips made into 2 strands each or 16 strands of 60 pixels.  I have them saved as a test config and can load the PT config for testing easily.  I am on XL/NC 3.6.2 and using an ETD, an active hub (Univ 1) and a passive hub (Univ 2) for the PT  The PT is configured for channels 4097 thru 6976.  XL is configured for 3 pixelnet universes.

I am using the RGB test with background only and moving the 3 sliders at the top.  From the left to the right I consider them to be channels 123 for each pixel and to be RGB in order.

If I do a select all on the channels for testing the PT works as a BRG

If I select just the channels for the PT it works as RGB

Has anyone else seen this issue or can anyone else verify this issue.
Or maybe when xlights is in select all mode it just starts with channel 1 as red and then every 3rd channel is red, likewise with green and blue.  In that case 4097/3 is a whole number and would be blue or the 3rd part of RGB with 4098 being the part of every 3rd pixel starting with 1, 4, 7 etc.

I tried manually selecting all channels instead of using the button and still get BRG BOO
I tried manually selecting channels 3 thru the end but still get BRG BOO
I tried manually selecting channels 4097 -8192 and get RGB YEAH

Something is wrong in the test function.

[flyinverted]

It would be cool if you could select a set of channels in the test function to display on the screen.

Those channels should light up accordingly from the test output so you could then have it match your physical channels.

When I first started, I had to change some lights to BRG from RGB but it was via trial and error to get it right.

[Steve Gase]

xlights does not know (or respect) your pixel definitions.

if you had 2 identical strings with the same controller (RGB) definition, and one started at channel 100, and the other started at channel 200, then their first pixel starts at different offsets with respect to the test patterns.

the 3-color pattern will do...

1,4,7,10,13,16,...   then 2,5,8,11,14,17,...  then 3,6,9,12,15,18,... then cycle to 1,4,7,10,13,16,...


if your first string is starts in the first group, it might turn red...   
if your second string starts in the third group, it would be red 2 timings later...  and would be blue when the first group is on.

make sense?

this year i've been moving my pixels around... looking for consistent RGB during the test pattern.  When you start crossing universes it gets even more interesting because the universes are slightly offset from each other since universes are not allocated channels in multiples of 3)

[CaptainMurdoch]

I think your calculations are off.

4097 / 3 is not a whole number, it is 1365.6666.

4098 / 3 = 1366

[jnealand]

That's what I get for doing it in my head, but the principle is the same.

[JonB256]

Old grade school trick is handy.

Is a number divisible by 3?  Add the digits. If their sum is divisible by 3, then the original number is, also.

4097   4 + 0 + 9 +7 = 20   (nope)
4098   4 + 0 + 9 +8 = 21   (yes)

[jnealand]

And that was the trick that failed when I added in my head and came up with 18.  LOL  It's hell to get old.

[JerryPlak]

It call brain fart Lol

Sent from my QMV7A using Tapatalk