Solved!
I encountered WS2812-compatible LEDs which have pre-programmed position addresses! These have been described sparingly during the past 2 years on the internet but appear still to be a kind of novelty.
First I have to say that I know how the WS2812 LEDs work, how data is sent and how to control them. I like the user interface of WLED and therefore I want to understand what happens.
I have a string of 50 WS2812-compatible LEDs on a Wemos D1 mini, WLED v0.13.0-b7 “Toki” build 2202222 installed from the Chrome-installer.
I cannot control anything from the user interface. No diode reacts.
However, for every LED above the real number of 50 I get one more LED under control starting from the end of the whole string. So with 51 LEDs exactly the last LED is under control.
Setting Length=97 gives me control over all but the very first 3 LEDs in my string.
I would go back to V13B6 and try it. B7 broke my configuration and I went back to B6 and all good now. B7 is the first beta I have had a problem with since I started using WLED last Summer.
Actually I tried this already yesterday, also using the Chrome-uploader. But it didn’t change the problem. The Chrome-uploader does not support any older version than b6.
For now my project works with the workaround, but all dynamic effects are of course hindered by the weird counting…
It was my bad - I actually encountered a string of weird neopixel-like LEDs which appear to have internal addressing implemented. This means that independent on their position in the string, the LEDs pass through all bits in the stream and react to a certain position which must have been programmed into them during production or more likely assembly.
and my own analysis confirms this function. So if you want to repurpose these types of neopixels which are otherwise fully WS2812-compatible you need to know the original length of the delivered strip and the start position of a substring…
