Chocobo dyeing R&D Thread - Post your results here!

[Giantbane]

You tell me, just exactly how you can even suggest the fruits ALWAYS behaves the same when you have the follow values??

Xelphatol Red Green Blue Red Δ Green Δ Blue Δ
Desert Yellow (D) 223 183 87 0 0 0
Millioncorn Yellow 231 159 55 -8 24 32
Pumpkin Orange 199 119 39 32 40 16
Sunset Orange 183 95 55 16 24 -16
Blood Red 143 63 39 40 32 16

O'Ghomoro (sum)
1 Desert Yellow (D) 223 183 87 0 0 0
2 Cork Brown 2x) 207 151 87 16 32 0
3 Ul Brown (3x) 183 167 119 24 -16 -32
4 Aldgoat Brown (2x) 167 135 103 16 32 16
5 Gobbiebag Brown (2x) 191 167 135 -24 -32 -32
6 Shale Brown (3x) 151 135 111 40 32 24
7 Goobbue Grey (2x) 135 135 135 16 0 -24
8 Lavender Blue (2x) 143 135 175 -8 0 -40
9 Lilac Purple 135 111 111 8 24 64
10 Plum Purple 127 87 111 8 24 0
11 Gloom Purple 79 71 103 48 16 8
12 Raptor Blue 95 135 199 -16 -64 -96
13 Othard Blue (2x) 55 95 143 40 40 56

I assume you're talking about the fact that feeding only berries or apples causes wild fluctuations in the RGB result values?

This threw me for a loop at first . But after seeing people have some luck with the +5/-5 theory, I decided to try it out. And it worked very well. I took what I had already fed it (ignoring color), calculated what more it would take and was able to get from gloom purple to soot black in two feedings (one mass feed, 1 small corrective feed). While this isn't proof, obviously, it was accurate enough that I'm inclined to believe that there is indeed an exact internal RGB value stored for the chocobo and that the fruits do work very closely to what was suggested with +5/-5.

As to how that fits into the path that the color takes when you feed the chocobo only blue, my best guess is that this is due to how the system selects the color based on the internal RGB value. It makes sense, if you assume an exact internal RGB value, that there must be some algorithm that best *fits* the internal RGB to one of the available dyes. The result is that each dye doesn't represent a single color, but a range of colors. And some of those colors can vary wildly.

Take blue for instance. If you keep feeding the bird berries, it will end up at Royal Blue and will go no further (39,47,103). If we assume the RGB +5/-5 theory, then continued feeding will result in 0,0,255, even though the dye remains at Royal Blue. There's a massive difference in actual values between 39,47,103 and 0,0,255, but it's still outwardly displayed as the same color. So each color can be representative of some pretty big color ranges.

I believe what we're seeing when you feed the chocobo only berries (or any other single fruit) is that you are adjusting an exact internal RGB in a linear path. But the result of what we see is where that linear path intersects the various color regions created by the dye picker. For instance, your first step in the berries.

Desert yellow (223,183,87) +2 berries = Cork Brown (207, 151, 87).

I believe that the berries are adding +5 b, -5 r & g for a result of

Desert yellow (223,183,87) +2 berries (-10,-10,10) = (213, 173, 97).

And then the internal color picker takes that (213, 173, 97) and pick cork brown as the *best* fit.

This then continues through the various color changes you listed above.

As far as the RNG is concerned, I believe that the ratio of all colors is preserved in a change. So berries *may* be +4, -4 OR +5, -5 OR +6, -6 (not exact values, i'm just saying that the adjustment to R, G & B all match), but berries would not cause a change of -4, -6, +5. My reason for thinking this is that berries always follow the same color path. If the fruits were random in a non-linear manner, then we would see the path diverge from what's listed above. However, everyone I know that has tested this sees the same transition you listed above.