Odd Colors Showing Up Where They Are Not Supposed To

[rayster]

What is the voltage at both ends? What is the length of cable from first pixel to controller? Is sound like a voltage issue. Do you happen to have any F-Amps that you can add in the string?

[MasterTJ]

Yes, sorry.  Didn't mean to detract...

I can share my personal experience...

I had an issue just like this and found out it wasn't a power issue but a DATA issue:

so I had a dedicated MW 350 30a 12v for 405 pixels...

MW>5 ft extension>Tune To Sign (99 pixels)>5 ft Extension>10 ft Extension>Singing Tree (252 Pixels)>2 ft extension>50 C9 PixaBulb.  

I had 6 pixels on the tree working but not changing...but they weren't fading, I mean I was injected into the tree, 2 times! had a color issue on the other strand as well but just color, I didnt lose any pixels.

After various troublshooting avenues I found out that the data was getting corrupted if the extensions were over 10' in length.  So F amps will probably do you right if you have long extension runs.  If ~10 ft runs, nulls should work fine. 

The pros (thanks Ron!) over at the xLights Zoom Room advised that Nulls will be needed ever 8-10 ft of lead, but f-amps will go to 30 ft plus-ish. 

From what I gather, voltage drop tends to make you lose brightness and causes lag, data loss (corruption) keeps the brightness and response but turns your colors all sideways.  The data signal pixels use is digital not analog.  It's either there or its not but it can and will show up to the party drunk too. 

I hope this helps.