RGB is being ignored in sample()

[eleweek]

Thanks for the library, it works great for computing a dominant color in the image. The results are great for our purposes.

However, I noticed that RGB components are being ignored and only H, S and M are taking into account.

 ((img.data[i] & mask) << shiftR) +
 ((img.data[i + 1] & mask) << shiftG) +
((img.data[i + 2] & mask << shiftB));

This evaluates to 0, because shiftR. shiftG and shiftB are all negative:

var shiftR:number = -lowBits + 2 * topBits;
var shiftG:number = -lowBits + topBits;
var shiftB:number = -lowBits;

Just check 42 << -1 or 42 >> -1

There are two ways to fix this:

1. Either get rid of negative bitshifts. This leads to different palettes with more similar colors.
2. Drop RGB from the space entirely. This should bring the memory consumption to 1KB, while keeping the results the same.

I'll let you decide what do with this.

[darosh]

@eleweek Thanks a lot! I dropped the RGB completely.