Find where it defines values for White, Magenta, Cyan, Blue, Yellow, Red, Green, and Black. That should tell you how many bits are set aside for each primary color.
Well I do appreciate the information you have provided thus far that is way more then even I was able to figure out. I even tried adding new colors to the list but got pretty weird results.
So if I understand you correctly it's not possible to add new colors based on the way it presently functions or have I misunderstood?
If I could just get purple that would mean the world to me. lol
Whatever 0x1fff is must have something to do with it right?
That value preserves the part of each output word that is NOT the background color. Those other parts of the word seem to control which digits are illuminated in the nixie tube and which of the two colon dots are illuminated. If you want more control over the background color you can connect the RGB LEDs to PWM output pins and control them directly rather than driving them from the shift registers that drive the Nixie tubes. There are plenty of examples of controlling RGB LEDs with PWM pins.
The unit in the little bad is the NIXI TUBE Module.
I wrote an RGB Library today that drives a similar RGB LED behind the tubes but I did not think about trying to make it work behind a shift register, I assume it's not possible? Or can it somehow be done?
The developer sent me this chart and said the following "Now the library is only provide simple colors, if you want an RGB color, you need a PWM control base on serial data." (was in Chinese so it is a little odd sounding)
There are chips that can do PWM dimming on a bunch of LEDs:
I believe these will drive Common Anode RGB LEDs. The 24-channel model will allow full color control of eight RGB LEDs and the 12-channel model will control four RGB LEDs.
If your LEDs are Common Cathode you may be out of luck.