Chocobo Dye Experimentation

[Rorichan]

*snip*

Hi, I don't think you read my log post. That's EXACTLY what I did and it skips EVERY time. I give him 1 berry, get one feathers message, he goes to Regal Purple so then I give him 1 apple, get one feathers message, he skips to Dalamud Red. If I have to reset him and try again, it's going to be way to expensive.


I said in the beginning that the way they have come up with for dyeing our Chocobos is just silly and frustrating. I'm all for taking steps to get things done but throw RNG into this system is just downright stupid. *sigh*

[Anonymoose]

I give him 1 berry, get one feathers message, he goes to Regal Purple so then I give him 1 apple, get one feathers message, he skips to Dalamud Red.

I'm wondering if the fact that it took 18 berries to get there the first time messed with the other two values in a way that you'll skip over Rolanberry most times, if not every time. If you're open to a little gambling, I'd try starting at Regal Purple and using berries instead of apples, pushing you into something around Abyssal or Royal Blue-ish territory and breaking your loop. That way, you can use apples from there to push back into Rolanberry territory from a less volatile position. Berries from purple will inflate the blue value, but it'll drop again when you switch back to apples to bringing the red back up - and both steps will just plummet the green, which is theoretically what you want.

Dalamud Red (113 22 22) +Berries
Regal Purple (95 42 71) +Berries
Royal Blue (34 42 96) +Apples
Abyssal Blue (43 39 80) +Apples
Rolanberry Red (84 19 35) Profit?

Worst case scenario, you don't end up at Rolanberry, yet, but still end up well outside of your loop, yeah?

[Rorichan]

Great suggestion here.

I was thinking that the berries pushed him too far that time.. I may try your suggestion once my new fruit crop comes up and see if that doesn't get me where I want to go. Thank you though!

[Rorichan]

[/B]Worst case scenario, you don't end up at Rolanberry, yet, but still end up well outside of your loop, yeah?

So.. my chicken is really stubborn.. as soon as I fed him apples POOF! Regal Purple.. *facepalm* I put him back to Dalamud Red and regrowing more fruit. I think I'm going to feed him a lemon and start over again in hopes that he gets reset and I can do a smooth curve to Rolanberry Red.

[Oniguro]

According to this chart:

It might be better to refer people to the newer chart, which you can find here.
My own progress can be seen on this page (on the ChocoboDye tab).
I'm linking to google docs so that I can just give the same link each time without having to convert my chart for the forum.

[Seradra]

So I followed the steps to get Abyssal Blue from the default Desert Yellow, and instead landed on Plum Purple.

Fed him 3 plums, ended with Slate Gray.
Fed him 2 Berries, went back to Plum Purple.
Fed him 3 more berries, turned Ceruleum Blue.
Fed him 2 more berries, now I'm GLOOM Purple.

Fed him 2 more berries, and he's changing again. According to the chart, I should finally get Abyssal Blue. We'll see.

[LyricalMelody]

I'm taking the gathered data and moving to my next step - Chocobo Color Map. It's not complete, but I'm getting there. I'm not sure how the google draw thing works with sharing but you are welcome to copy it to your own drive to play with it. Initially I think it looks small, but you have to zoom in for full size.

[Freakindon]

So this has probably been addressed, but it feels to me as though the "only the first feather proc" counts deal is false. I feel as though every fruit counts, but the message is only indicative that the fruit you just fed puts the RGB at a value that will be a different color from what your chocobo is. Which is why it is possible to get the "feather proc" message, and then not get it on some subsequent feed. Sometimes it takes a good deal of fruit because there are many RGB values that are not valid as colors, so it might take a while until you get to a valid one.

[AttacKat]

but it feels to me as though the "only the first feather proc" counts deal is false. I feel as though every fruit counts, but the message is only indicative that the fruit you just fed puts the RGB at a value that will be a different color from what your chocobo is. Which is why it is possible to get the "feather proc" message, and then not get it on some subsequent feed.

Right now, most of us aren't saying that only one-fruit one-proc counts, but that the results of not doing it that way are harder to understand or replicate. By doing it this way, we get a clearer picture of the full color-to-color map, as well as a better understanding of the change mechanics. I bet there are lots of ways to feed more than one fruit and proc at a time to leap ahead on (and even cut diagonally across) the map - but we're not sure yet how to do it with confidence or replication.

After receiving so much input HERE. The RNG factor is pretty damn obvious. RNG causes you to; skip a step or two ahead, get a bust, land you on an 'uncommon' colour for that particular fruit.

It seems like most of the single-fruit-feeding results are becoming quite clear. Still, not every colour + fruit is mapped. By that I mean, colour x + all the results of the 6 fruits.

NOTE: The conclusions derived is from applying the RGB code defined within my own spreadsheet.

The Givens:
1. The color mixing must follow the law of nature. IOW, if white+black = grey, SE cannot make it green in game. Hence if we mixing blue and red, expect purple.
2. Opposite mixes will do just that, opposite. Deepen red mixed with lighten red = wash, back to original tone.

My my own observations, I suggest the following on RNG:

To keep it fun, SE kept this in game, however, my personal observation is this is limited to the absolute value of 48. So, starting with a base R{103), we will only with colors that falls within the range of R(55) to R(151). Anything falls outside of that range will be considered invalid values. I will defined this RNG|48| from here on.

For an example, for Regal Purple R(103, 55 87), the possible RGB range is between (151, 103, 135) and (55, 7, 39)

Then from this (151, 103, 135) and (55, 7, 39) range, existing colors are choose to be the next possible color, based on the law of color mixing. This MAY help explain why some colors has more then one end result.

So, to answer some of the questions quoted:

1. The feather proc: The feather proc is always true. The question here isn't about if it is a valid indicator for the next color change. When you receive this message, your bird will change to a new color. What is unknown ATM is if we need feeding the bird more fruit after receiving this proc, what, if any, effects the fruits will have.

2+3: My personal believe here is due to the RNG|48| factor. Therefore, all the color calculation spreadsheets is inclusive because of it. As of now, we have only done controlled test of ONE fruit, and even that is giving us a lot of open cans of worms that we do not have clear answers on.

We do not know the behavior of the RNG|48|, is it doing it to every fruit, or the total value of all the fruits? Until we figure that out, we will always be in the dark. Different people feeding different combos of multiple fruits only makes matter worst, especially those recommending feeding multiple fruits in given quantities. There are no solid date to say a given fruit will give a fixed amount of effect, therefore, how can any quantity be given?

One experiment we need to start is feeding the birds same fruit in large qualities, and see if we can always achieve the same color scale jump.

Color scale jump experiment: It is known that we can go from Desert yellow -> millioncorn -> pumkin -> blood red. This was repeated via single feather proc in 4 stages and X amount of fruits. Can we attempt this via large feedings and conclude that by giving X-apples + RNG|48|, we can always result in blood red. (Say X = 30, result = X +/- 3)

My current experiment is feeding 2+1 all the time, and see if the RGB value will be close to my desired goal. If this comes out conclusive, it should show that the color mixing pattern is valid.

IMHO, until we move forward to the next step of color experiments, this can of worms will always be inclusive.

[Anonymoose]

LMAO Nice! Congrats.

Thanks much!

Trying to find Wine Red myself.

I'll keep a close eye on the thread, then. The other member of the Moose clan is going for the same. In fact, I think I'll do a diet version of my changelog with her progress. In the event of a loop, know she's more likely to overfeed on one color and try to go around than reset, so if RNGesus gets involved again, we're like to get a long list of "Whassis dooo? Whassat dooo?"

it feels to me as though the "only the first feather proc" counts deal is false.

To be more articulate about the "first feather proc" thing, we're pretty sure that every fruit can count, and most of them likely do. It's just that feeding a bunch of random fruit to see where you land doesn't give you a good idea about the what-affects-what-and-how, or help determine how much of a role RNG plays.

Some people report that once you get the first proc, feeding the same fruit seems to give you more even if it doesn't color-shift again. Feed it apples to one proc and you're almost[?] definitely headed for Millioncorn, but people have fed apples to four procs and still got to Millioncorn instead of skipping to blood red. Others report better results by switching fruits every time there's a proc, which they say increases confidence that it wasn't a "random repeat" kind of thing.

Right now, most of us aren't saying that only one-fruit one-proc counts, but that the results of not doing it that way are harder to understand or replicate. By doing it this way, we get a clearer picture of the full color-to-color map, as well as a better understanding of the change mechanics. I bet there are lots of ways to feed more than one fruit and proc at a time to leap ahead on (and even cut diagonally across) the map - but we're not sure yet how to do it with confidence or replication.