All channels in Rgb48Color initialised from R channel in RgbColor

[blazoncek]

https://github.com/Makuna/NeoPixelBus/blob/d212d3ac5cad3d33785aea838160d11e17111eb2/src/internal/colors/Rgb48Color.h#L60

I may be wrong, but this does not seem correct to me.

[Makuna]

Run the calculation with some sample data:

R = 255 (max uint8_t) = 255 257 = 65535 (max uint16_t) - correct R = 0 (min uint8_t) = 0 257 = 0 (min uint16_t) - correct R = 1 = 1 257 = 257 - close, expect 256 to be precise R = 127 = 127 257 = 32639 - close, expect 32767 to be precise

So, look at the arduino map() method and you will see that it factors down to the same...

long map(long x, long in_min, long in_max, long out_min, long out_max)
{
  return (x - in_min) * (out_max - out_min) / (in_max - in_min) + out_min;
}

call like

uint16_t map (127, 0, 255, 0, 65535);

out = (127 - 0) (65535 - 0) / (255 - 0) + 0; out = 127 65535 / 255; // factor our zeros to get this out = 127 * 257; // factor out consts (other than 127 input) to get this out = 32639!

It's a fast approximation without branching or additional calculations.

[Makuna]

Ok, some days I need to get my coffee before reading/answering. YES, you are correct.

[Makuna]

https://github.com/Makuna/NeoPixelBus/pull/687

[blazoncek]

Ok, some days I need to get my coffee before reading/answering. YES, you are correct.

I know. I am the same. 🤣

[Makuna]

v2.7.5