Discrepancy in setting dotstar brightness byte

[dunkmann00]

The dotstar brightness byte is exposed to the user as a float, but then internally converted into an integer, so it needs to be rounded.

In pypixelbuf the brightness is always rounded up using an approach equivalent to math.ceil:

Adafruit_CircuitPython_Pypixelbuf/adafruit_pypixelbuf.py

Lines 255 to 257 in e52c672

	# same as math.ceil(brightness * 31) & 0b00011111
	# Idea from https://www.codeproject.com/Tips/700780/Fast-floor-ceiling-functions
	w = (32 - int(32 - w * 31) & 0b00011111) | DOTSTAR_LED_START

However, in pixelbuf the brightness is rounded down by truncating:
https://github.com/adafruit/circuitpython/blob/41fe62929f268bfc1f2b06be8b649409234d9d1a/shared-module/_pixelbuf/PixelBuf.c#L181-L185

If rounding down is desired I can make a pull request to correct this? Or if not, I can file an issue in CP and update pixelbuf to round up!

[tannewt]

Consistency is the most important to me. Does rounding up prevent setting brightness to zero?

[dunkmann00]

No, if you set the brightness to 0.0 it will be set to 0.

The rounding, is just happening because of the float needing to become a whole number (a 5 bit int for the dotstar). So in the pypixelbuf code, it really is just equivalent to a math.ceil, so 0 still yields a 0.

As an example:

User enters: 0.0  -->  Multiply by 31 (5 bits): 0     -->  Set brightness with: 0 (same with pixelbuf)

User enters: 0.4  -->  Multiply by 31 (5 bits): 12.4  -->  Set brightness with: 13 (pixelbuf would set to 12)

Either way is equally useful, really just a matter of preference!

[tannewt]

I'd round down to be consistent with _pixelbuf because it is harder to change.