* setPixelColor: ensure that 0/0 is used
* getPixelColor: accuracy improvements
unfortunately, "scrolling" audioreactive effects are still not working properly - they end after 1/4 of the circle. Could be due to limited resolution of getPixelColor.
fixing holes that appeared during testing
* at 52x52 (big 296 -> 304)
* at 24x32 (medium 192 -> 208)
* at 12x16 (small 72 -> 82)
... there is still one hole at 14x16 ... well wtf
- increased outputs to 17
- increased max possible color order overrides
- use WLED_USE_PARALLEL_I2S during compile
WARNING: Do not set up more than 256 LEDs per output when using parallel I2S with NeoPixelBus less than 2.9.0
The BME280 usermod uses a multiply-round-divide approach to cap the temperature/humidity/pressure values to some number of decimals. But the divide-part was missing in a few instances.
Issue:
When taking the initial voltage reading after first powering on, voltage hasn't had chance to stabilize so the reading can be inaccurate, which in turn may incorrectly trigger the low-power preset. (Manifests when the user has selected a low read interval and/or is using a capacitor).
Resolution:
A non-blocking, fixed 10 second delay has been added to the initial voltage reading to give the voltage time to stabilize.
This is a reworked version of the (now closed) PR here:
https://github.com/Aircoookie/WLED/pull/3959
- Rebased the update for 0_15.
- Added a constant so the delay can be modified via my_config.h.
- Small adjustments to make the PR compatible again after the recent restructuring in this PR: (https://github.com/Aircoookie/WLED/pull/3003).
Thankyou!
Script update based on latest version from Tasmota
* add support for all esp32 variants
* add "-C" : Decode (demangle) low-level symbol names into user-level C++ names.
The build script was not looking into the right place, so there was never a .map file dropped into build_output/map/
Builds with the newer arduino-esp32 v2.0.x framework actually generate a .map file that is placed directly next to firmware.bin