Addressable Light Algorithm
Brian Erickson
Hi everyone,
In yesterday’s meeting, we were talking about how addressable pixel LEDs work, and I realized I gave some incorrect information about counters and decrementers. That was just my guess — the real algorithm is actually much simpler and more elegant.
Here’s how it really works:
After a short period where the data line is held low, the controller sends out a stream of 24-bit values - 8 bits each for red, green, and blue.
The first pixel detects the low period as a reset signal, then “gobbles up” the first 24 bits and uses them to set its own RGB value. It doesn’t forward those first 24 bits; instead, it holds its output line low while reading them. Once it’s done, it begins relaying the rest of the data down the line to the next pixel - minus the bits it already consumed.
The next pixel wakes up when data starts arriving, captures the next 24 bits, and passes the rest along. This continues until the last pixel receives its 24 bits.
Here’s a good video explanation (from Dave’s Garage):
🔗 https://www.youtube.com/watch?v=gHsJP5br_ok&t=588s
I think it’s a really clever system - no addresses, no special math, just a simple “grab the first 24 bits and pass the rest” approach.
