I tested it in photoshop, it simply made it more yellow. Basically you did a combination who's result was already what you had, which is why there was no changes at all.
Printable View
Hello, I'm that friend. thoughd I'd add a few more details:
Apologies if this has been posted already. This is only a THEORY, so it might be wrong, but it's been working for me so far.
Theory 1:
I "think" that each fruit basically alters +/- 1 of the first digit of the RGB hex value, with a random chance of +/-
a few of the 2nd digit. This means, each fruit basically adds or subtracts around 16 points.
Theory 2:
If you go over 255 or under 0, it loops around. This part is definately THEORY, I haven't 100% comfirmed this, but
it seems legit so far. If someone wants to confirm this, they could test it by giving a Snow White +1RGB to get Soot Black, or -1RGB for the opposite.
So does this mean?
First you have to understand how RGB hex works. There's a table floating out there with the RGB values in integer format (which may or may not be accurate)
meaning theres a number from 0 to 255 representing Red, Green, and Blue. Each of these numbers can actually be represented
with a 2 digit Hex value, so then it would actually range from 00 to FF.
Colors are represented as six digits then, the first 2 being the red channel, 2nd being green channel, and the 3rd being the blue
channel. HEX is used in computer graphics because its far more efficient to a computer in terms of memory usage and something
to do with they way 32 bit processors work (someone feel free to correct me on the this, I'm an artist not a programmer
so I don't know the exact details for this)
So, 000000 is Black, and FFFFFF is pure white.
16 in red, green, and blue would 101010. 32 would be 202020, and so on and so forth.
So lets put this into practice. You start with Desert Yellow (216, 175, 80). Using this calculator,
http://www.javascripter.net/faq/rgbtohex.htm
it converts to D8AF50. For the purpose of this, we only care about the odd number digits, so we'll do some rounding. So it goes to
D0B050.
lets say we want to get Ochu Green (57,92,51) or in hex (395C33) and rounded Hex (406030)
so we put these to numbers next to each other, Source over Target
D0B050
406030
And we calculate the difference
-9 Red
-5 Green
-2 Blue
So, in theory, 9 doman plums, 5 valfruit, and 2 cieldales pineapples would get you to the target color, or very close that
the 2nd feeding would be much easier to figure out.
But wait, you should never have to use more then 8 of one fruit. So, if you look at red, you realize you can go the OTHER direction.
If you go the other direction its +8 Red. So now it becomes 8 apples, 5 valfruit, and 2 pineapples.
This all speculation, but again, this has been working well for me. Feel free to play around with this idea and let me know.
My target was Dalamud Red. From the default yellow, I went off the "instructions side" of that spreadsheet going on and it
was totally off, but using that result and a few other results, I was able to extrapolate the above math and my 2nd attempt using this
math got me to rolanberry red (which is super close to dalamud). From rolanberry red, it was only 2 more fruit to get dalamud
using the same math. SO, either my math just happens to be a big coincidence, or it actually works. I hope some people can test this
cause I got the color I want so I'm not gonna mess around anymore >.> But to be clear, if this does work, it means
no target should require more then 16 fruit total, and this is to go from one color to a complement.
So the maximum fruits are required from going from
Red <==> Green
Blue <==> Orange
Yellow <==> Violet
The range would be 1 Fruit < Mono < Analogous < Triatic < Complementary < 16 Fruit
There is one exception to the Max 16 rule under the above math, and that's the Monochromatic grayscale. A jump of 50% in on the monochromatic gray scale would require 24 fruit.
(ie, White or black to 50% gray, 25% to 75% gray, etc.)
However, if Theory 2 proves to be wrong, that means it would take up to a maximum 16 fruit to make a change, in which case then the max numbers turn to 32 and 48 respectively.
If theory 1 is wrong but 2 is right, then it means it really is just a case of plus or minusing in the direction you want to go, but the amount of plussing or minusing that happens from fruit has far more variables then being around 16 (such as diminishing returns, different seed numbers, higher rng range, or just being flat out random).
If theory 1 AND 2 is also wrong then this whole venture of math just gets thrown out the window. LOLOLOLOLOL.
Lastly I should mention my stable was in good condition for this.
*edit* was reading Renik's post about how the fruit actually do multiple things. If this is true (and it probably is), it seriously complicates the above theory for big jumps. I think for small jumps (under 4 fruit, with no more then 2 of any one fruit), you can ignore the secondary effects of a fruit, but for larger jumps the math becomes seriously complicated. Le sigh.
SOI:SOI:SOI:SOI:SOI:SOI:SOI
..........___|___.........
...L..../.......[`\........
..LOL===........[__|.......
...L....\..........|.......
.........\_________/.......
...........|...|...........
..........————————/........
Makes sense. I was half asleep just shoving food down it's throat at this point. Remind me not to do things at 5am.
Anyway, since it wasn't going to change, I took my bird out of the stable, then put it back in and tried again:
Starting Colour: Honey Yellow
1 Doman Plum > "New feathers" (This made me facepalm hard)
2 Doman Plum > "New Feathers"
15 O'Ghomoro Berries > Each one had a new feathers notice.
I will post result when he's ready!
Cearka, I wanna try out your theory to go from Desert Yellow to Lotus Pink and compare it to Canadane's recipie for Lotus Pink. If you're theory is correct, I should be able to achieve it in less steps since she fed over 8 fruits twice.
I have just learned from your post how HEX works so I'm not too sure how the rounding works.
Desert Yellow: D8AF50 rounded D0B050 (as per your post)
Lotus Pink: FECAF4 rounded FFD0F0 (I took the RBG value of Lotus Pink from here and then used your link to convert it to HEX)
Does the rounding makes sense? F0 for red made it look terribly close to Colibri Pink, but on the other hand I can't add another unit to FF, so it' not exactly a rounding either.
Canadane's recipie is -12 r, -12 g, -6 b (only counting new feathers messages)
With your theory it would be +2/3 r, +2 g, +10 b
Either something is wrong with my rounding or I can't seem to get the blue to less than 10 berries. Even going the other way around, it would still be -10.
I offer my Chocobo for SCIENCE if anyone wants to experiment with your own fruits since i don't have any xd send me a /tell if you want :D at Cactuar!
Midnight Blue - Fed it Apple and got Kobold Brown
Kobold Brown - Fed it Berries and got Midnight Blue
Midnight Blue - Fed it pineapple and got Kobold Brown
Kobold Brown - Fed it Plums and waiting on result
So I've been doing mine in steps, counting feathers messages etc, and i was working my way towards Celeste green. This morning I woke up to find my chocobo was snow white, and it was done with a lot less fruits than people are feeding their chocos currently, so I thought I'd hop on and post how :) I started at the base desert yellow.
1) 6x doman plum (2 feather messages) = Cream yellow
2) 3x doman plum (3 feather messages) = Bone white
3) 3x doman plum (2 feather messages) = Snow white.
Snow white: 8 pear procs apple green
i got mines in 2 6hr feeds. 1 6hr - started at desert yellow fed 4x plums (feather messages on 3rd&4th feed) =sylph green. 2nd 6hr - started at sylph green fed 8x plums(feather message on last feed) = snow white..... this is getting interesting. hope we can figure something out soon. I just want a colibri pink choco :)
I did not do it myself, and I cannot say for sure, but from what the original Spreadsheet creator said, I believe it was obtained with screenshots of the Dye Interface and a color picker in some Photoshop variant. It's actually fairly clever to do it that way, the UI element is not affected by game lighting or conditions, and because the color was maintained internally between programs, the HSV of the color should have remained exactly what they actually were from the game's perspective. If I'm correct about his methods, those numbers should be 100% accurate to what the game actually uses.
I love the information we're gathering everyone and how it's helping you all to work on your theories. I updated the OP with links to experiments, more theory posts, requests for volunteers, volunteers who have offered their services, etc. I also condensed some of the information down using the Hide feature so that the post wouldn't be so long.
Could I get my Data Gathering project listed, please?
Link to my post
Just a theory :
When you feed your chocobo you can get a "growing feathers" message. Once you got it you *always* get it in the same session. I think it's just to notify we fed our chocobo enough for him to pass a step, and change color, and that's why it's always displayed after the first one.
In my opinion, all feeds are effective, but only a certain amount will allow to pass to another color.
It also means that the 2nd message and the other ones are just a repetition of the first, so furthest messages are irrelevant to the effectiveness or not of the last given fruits !
I don't think it's relevant to note "growing feathers" messages like that. Either people STOP feeding the chocobo and wait to see each color each time it says "growing feathers" to really note when fruit counted as a big step, either it's no use.
PS: some people have noted the message went away after feeding a lot, I think that's just because they went back to current chocobo color with all their feedings ^^
Your rounding seems fine. In the case of the FF because there so close to flipping over, we Round them back down to F0. Its ok if its close to another color. Under this theory the idea it to try to get in ball park and then go from there, not necessary to hit the target color spot on. As long as you get something numerically close it shouldn't take more then one or two feedings of 1-2 fruit to hit the color you want.
So
D0B050
F0D0F0
D=>F: +2
B=>D: +2
5=>F: +10, so lets go the other way, (F=>5: -6)
2 Apples, 2 Pears, 6 Pineapples. In theory that should get you numerically close if blue actually loops around.
A little more explanation. Your always going to count UP. when I say go the other way I mean swap the target and source.
Remember, the count is in this order:
0 1 2 3 4 5 6 7 8 9 A B C D E F
After F you go back to 0.
It may or may not be as straight forward as that though. I've been reading some other peoples theories
where some of the fruits do multiple things at once, which then really complicates stuff.
Part one confirmed. After I corrected the Chocobo RGB spreadsheet, all of my results are in multiple of 16s, check my post to see my on going findings.
Just did the same and can confirm his method works, but the RGB he reported are off. What I am seeing is pattern of 16s, and RGB hex ending in 7s. In other words, only values of 00h, 07h, 17h, 27h ... E7, F7, (FF) are being used in the colors. I am working on a new spreadsheet to reflect these values.
Just on the deepen red, I am seeing patterns of
-16 | 16 | 32
32 | 48 | 16
16 | 16 | -16
32 | 48 | 16
16 | 16 | -16
A quick grab of someone's result on a different fruit, I am seeing
64 | 48 | -16
16 | 0 | -80
So I will have to gather more results to come up with a more solid pattern.
That's why I messed up the last one, thank you!Quote:
A little more explanation. Your always going to count UP. when I say go the other way I mean swap the target and source.
I have started feeding the Chocobo the apples first, but I did not get a "new feather" message after 2, but 3 (which is still consistent with your theory if we consider the exact numbers, as it's a +38 difference in red).
Tonight I will check what colour it has become, take pics and notes, and feed him pears until I get a new feather message, which I expect in 1-3 feedings (it's +27 from Desert Yellow but it may vary once the green value is adjusted to match whatever dye I will get from the first feeding).
I will stop feeding everytime I get a new feathers message and edit this post with the findings.
Experiment: Following cearka's colour theory I will attempt to go from Desert Yellow to Lotus Pink
According to the theory, I will need around 2 Xelphatol Apples, 2 Mamook Pears, 6 Ciedalaes Pineapples (assuming the blue will loop). I will stop feeding the chocobo every time I get a "new feathers" message.
First Feed
3 Xelphatol Apples ==> "new feathers" message
--- 6 hours wait---
Colour change! From Desert Yellow to Millicorn Yellow
Sharing my results so far in my quest to get Morbol Green.
I had previously seen 2 posters claim they got Turquoise with the following:
Person 1
Feeding 1: 3 Pears
Feeding 2: 20 Berries, 10 Pears
Person 2
Feeding 1: 20 Berries, 20 Pears
Here's what I did:
Attempt 1
Color: Desert Yellow (Default)
Stable Conditions: Good
Food1: 20 Berries (18 feather procs on the final 18)
Food2: 10 Pears (feather procs on all 10)
Result: Seafog Blue
Attempt 2
Color: Seafog Blue
Stable Conditions: Good
Food1: 4 Pears (a single proc on the final fruit)
Result: Turquoise
Turquoise is pretty close to Morbol Green so I'm tempted to not muck with it, but I really want that other color. Will update with more info if I pull the trigger.
So, I goofed a little at the start, but here are the results of my guild so far:
Person 1 (Me) -- Chocobo: Oddard, Goal: Something Bright Green
Attempt 1 -- Fed 1 Pear. No Feathers message. Desert Yellow, no change.
Attempt 2 -- Fed 6 Pears (Darken Green). One Feathers message. Desert Yellow --> Lime Green
Attempt 3 -- Fed 6 Pears (Darken Green). One Feathers message. Lime Green --> Honey Yellow (o.O)
Attempt 4 -- Fed 7 Pears (Darken Green). One Feathers message. Honey Yellow --> Lime Green (Gah! The heck?!)
Attempt 5 -- Fed 5 Pears (Darken Green). One Feathers message. Lime Green --> Apple Green
Person 2 -- Chocobo: Maltear, Goal: Soot Black
Attempt 1 -- Fed 2 Apples (Darken Red). One Feathers message. Desert Yellow --> Millioncorn Yellow
Attempt 2 -- Fed 10 Apples (Darken Red). One Feathers message. Millioncorn Yellow --> Pumpkin Orange
Attempt 3 -- Fed 6 Berries (Darken Blue). One Feathers message. Pumpkin Orange --> Sunset Orange
Attempt 4 -- Fed 3 Pears (Darken Green). One Feathers message. Sunset Orange --> (awaiting results)
Person 3 -- Chocobo: Edge, Goal: Royal Blue
Attempt 1 -- Fed 4 Berries (Darken Blue). One Feathers message. Desert Yellow --> Ul Brown
Attempt 2 -- Fed 4 Berries (Darken Blue). One Feathers message. Ul Brown --> Aldgoat Brown
Attempt 3 -- Fed 1 Berry (Darken blue). One Feathers message. Aldgoat Brown --> Shale Brown
Person 4 -- Chocobo: Enkidu, Goal: Something Dark/Bright Green
Attempt 1 -- Fed 8 Pears (Darken Green). One Feathers message. Desert Yellow -- > (awaiting results)
My findings are consistent with the idea that after the first "growing new feathers" message it will continue to show up on each of the subsequent feedings of that snack type. It might be a bug that will be fixed. Right now, you always get a success message after the first one (on the same snack), but you don't know if it actually proc'd a color change. So saying you got 8 feather messages to go from color A to color B would be like telling someone it will take 8 tries to slot 3 overmelds. (assume that after the first proc, they can no longer see a success or fail and then get to see their results when they are finished). 8 is not a number that will work for everyone.
Evidence: from my data
Dolman Plum
Person 1
Desert yellow: 2 "new feather" messages => Cream Yellow
Cream Yellow: 3 "new feather" messages => Bone White
Bone white: 2 "new feather" messages => Snow White
Person 2
Honey Yellow: 1 "new feather" msg => Cream Yellow
Cream Yellow: 1 "new feather" msg => Bone White
It's possible that both of Person 1's color changes proc'd on the Desert Yellow, skipping over the Honey Yellow color, though I would need to use a Plum on a Desert Yellow (to one feather message) to confirm this.
Person 3
Desert Yellow: 8 "new feather" messages => Snow White
Person 1 shows 7 "new feather" messages in total from Desert Yellow => Snow White, so one of Person 3's proc's failed.
I believe it is possible that certain colors may have an "HQ-type" proc that results in a different color.
But that is why I am trying to gather all the datas :)
Data Gathering Form
I can confirm the Pumpkin Orange to Sunset Orange with one new feather message from apples, and also the Sunset Orange to Blood Red with one from apples.
My other post. http://forum.square-enix.com/ffxiv/t...=1#post2384747
I will be continuing on feeding apples until one "new feathers" message appears. Hopefully I'll hit Dalamud Red next, and then maybe Rust or Wine Red.
Just waiting on my apple plant which should be finished soon.
I'm going to guess that Red Wine is going to contain some blue, so feeding it berries. Wondering if plum (a.k.a. adding cyan) would do it as well, or if that would just take you a step back on the red scale.
Maybe.. we'll see. Trying to find out what we can get just feeding apples.
For the replies in here, please see my post
Result:
Desert Yellow -> Millioncorn Yellow -> Pumpkin Orange -> Sunset Orange -> Blood Red
The next color after Blood Red will most likely be Mesa Red, then Dalamud Red.
Update on my experiment discussed in the OP. If you check out the document linked there, you'll see where I fed my bird a lemon to reset it to the base color and started to duplicate the experiment. I have already seen a difference in that this time it only took 4 plums to get the message about him growing new feathers.
Went back and put in the Millioncorn info after Person 2 woke up, so yeah, it's lining up with that path :)
Not reading all new posts since I'm hunting atm. What I've learned thus far:
Desert Yellow to Millioncorn Yellow - Apples to 1 feathers message
Millioncorn to Pumpkin Orange - Apples to 1 feathers message
Pumpkin to Sunset Orange - Apples to 1 feathers message
Desert Yellow to Sylph Green - Plums to 1 feathers message
What I'm not sure on is how getting multiple "feathers" messages in one sitting works.
back to hunting
Someone proposed a theory that the "new feathers" message is only meaningful at the first display by the game.
Afterwards it is very likely you will receive the message for every fruit feeding, or you may not. In the case of where you do keep getting it, you have no reliable indication on when the "new feather" is stacking on the previous one, and which one may actually go into the next color scale.
IE:
feed fruit
feed fruit
new feathers (new color 1)
feed fruit
new feather (stack 1)
feed fruit
new feather (stack 1)
feed fruit
feed fruit
feed fruit
new feather (stack 1)
feed fruit
new feather (new color 2)
...
I decided to shoot for Sunset Orange as well based off the info here. My chocobo was Apple Green when I started. I fed it a Hans Lemon and then fed it apples
Fed 17 apples. Message of feathers 16 times
Stable conditions- good
weather- overcast
http://i.imgur.com/sRFkeT2.png?1