Template talk:Dye infobox

From Guild Wars 2 Wiki
Jump to navigationJump to search

Different color boxes[edit]

For each piece of armor, there are 2-3 elements that accept color. If all such boxes on a full set of armor are given the same dye, how many different variations will we see? Just three (depending on whether the element is cloth, leather, or metal)? or can there be more?

Similarly, how are we asking people to determine the colors in the boxes? About the only thing I remember from psychology classes is that the same box of color can look different depending on the color of nearby boxes. Are we relying on subjective assessments of "this color is the same/different as that one?" – Tennessee Ernie Ford (TEF) 06:17, 19 June 2012 (UTC)

uh.... I Think anet has mentioned being able to dye 3 of the 4 regions on armor. As to filling out the colors in the info box, ditto.Torrenal 06:59, 19 June 2012 (UTC)
Color picker. —Dr Ishmael User Dr ishmael Diablo the chicken.png 12:26, 19 June 2012 (UTC)
What I'm asking is what determines the RGB of each of those 3-4 boxes. Is it limited to three variations per dye? Or are there far more permutations, depending on the "brand" of armor (e.g. heritage), the armor section (e.g. torso), as well as the material involved (e.g. metal/leather/cloth)?
To whoever used an odd number of tildes: thanks for the GIMP link. I assume that's used on screenshots and delivers consistent results from player to player. – Tennessee Ernie Ford (TEF) 16:56, 19 June 2012 (UTC)
Selection is picked up on the left, and the cloth-metal-leather is picked up on the right (not character model)
http://i.imgur.com/x9q3N.jpg (provided by Dosvidaniya ) the part labeled dye palette is where you pick up the material colors. Depending on the material the color is being applied to that box actually changes color. Ill see about uploading a more clear example of this when i get home from work. --Draygo Korvan 19:17, 19 June 2012 (UTC)
Please tell me you are not pulling colors from the character model Torrenal 19:22, 19 June 2012 (UTC)
I said above not character model--Draygo Korvan 19:23, 19 June 2012 (UTC)
The material colors from from the section labled dye palette. You have to look at the character model and determine what material it is, then use a color picker to get the RGB value from the dye palette. --Draygo Korvan 19:26, 19 June 2012 (UTC)
ah, missed that. Sorry. So you,re pulling the color from the textured patches? Err.... That aside, assuming that color varies by material and not object, I would think us remiss to not assume that there will be other materials. Example: chaos. Ghostly. Possibly celestial, etc...Torrenal 20:05, 19 June 2012 (UTC)
Thanks for adding the image, Draygo. Unfortunately, I haven't phrased my question very well. In your image, there are three pieces of armor shown in the dye palette with a total of 7 places to apply color. My question is: if we apply the same dye to each of those 7 boxes, could we end up with seven different RGB values? Or are the only variations for a given dye color the leather/cloth/metal ones mentioned? For example, someone has a heritage helm and crafted boots, both made of leather — do they end up with the same RGB value?
The reason I ask is that in real life, applying dyes has notably different effects on similar-but-not-identical material. The "Deep Purple" used on one brand of T-shirts sometimes looks different from that applied to another (due to different quality of cotton and/or different blends of other materials). In GW2, each piece of armor isn't monochromatic: it's got variations of color to show lines and creases, shinier bits of buckles, and so on. So I can imagine a system which allows for that type of variation by applying dye slightly differently to different sets or to different components or to different sections of those components. – Tennessee Ernie Ford (TEF) 20:12, 19 June 2012 (UTC)
I've gone through and looked at my own images. I have less than a difference of 4 bits (for the individual colors) when sampling from the dye palette or sampling from the selection palette for any dyes in the gray category. I've looked at medium armor images and am also in a similar range. The only fluctuation I see is based on the precise pixel sampled. When I run a 9 bit average on the center of the image, I see almost no difference at all. Are we certain there is a difference on the palettes when players have different armor? There is an obvious difference on model; but that is dependent on more than just the material. Is there a palette difference? Image links: (Saz-Medium) http://imgur.com/a/WZlUE/#0 (Mine-Light) http://dosvidaniya.imgur.com/ --Dosvidaniya 20:21 19 June 2012 (UTC)
yes there is a difference, Ill be uploading my reference when I get home marked up with the specifics. Also careful with 9px wide grabs, you probably are grabbing part of the overlay image on the color palette doing that, an average of a 3x3 area would probably be the biggest I'd go with. I do think certain colors are closer to each other, and some are farther. I noticed when I was applying dye to my character sometimes the color would have a very stark difference between some of the materials. Anyway, I ask you give me about 4 hours and Ill get a good reference here that explains what exactly i'm looking at. Oh and location on the dye palette had nothing to do with its color. The only thing that affected it is what material that specific spot represented A part that colored a metal spot on the armor would be the same as a spot that colored a different metal spot--Draygo Korvan 21:49, 19 June 2012 (UTC)

(Reset indent) OK here goes. This first one is the dye differences that I marked up to show you what I mean. The lines mark which dye area applies to what part of the armor, and what material it is.

[[:File:User Draygo Korvan dyedifferences.png]]

This second one is clean if you wanted to grab samples yourself to see the difference without the noise in the way. If you use the color range selection feature in photoshop with a sensitivity of 9 you will clearly see that the dye applied to different materials is not the same. I have only identified 3 different types of materials so far that affect the dye color, cloth, leather, and metal.

[[:File:User Draygo Korvan dyedifferences-clean.png]]

--Draygo Korvan 00:32, 20 June 2012 (UTC)

Curious, did you take BMP screenshots or the default JPG? — Rhoot User Rhoot sig.png 00:37, 20 June 2012 (UTC)
Raws are in bmp. --Draygo Korvan 00:41, 20 June 2012 (UTC)
Curious. I get AB6A88 for both, give or take a low bit on the pixel. Let's try something snazzy.... Gives me AF6888 for the top. Gives me AC6A88 for the remaining pink areas. Little or no color intersect between the top pink and the remaining pink panels. I'd quote the numbers I gave as 'exact'. Try for yourself: Select, By Color. Set the threshold to 0. Click around till you get a color that includes large areas, treat that as the base color.
That said, the difference is technical. The top color has 3 points more pink and 2 points less green, on a scale of 0 to 255. Unless players take out their color tools and test it for themselves, they aren't going to notice. Might bear study, not sure if it merits much else at this point. Torrenal 01:16, 20 June 2012 (UTC)
The top pink area is the only pink applied to metal. The bottom are all cloth. So as you said, there is little to no color interset between those two pinks. --Draygo Korvan 01:24, 20 June 2012 (UTC)
For an accurate color selection I was using a threshold of 9, until i got a large area, and took a 3x3 average of that area. The difference between the metal antique gold and the cloth one is visible to me when i stop and look at it. The pink color difference is hardly noticable but the antique gold to me is much more pronounced. I know I ran across colors in the beta that were more than just pronounced like it was completely obvious. Also I was using the original bmp for the color selection, the png may be a bit off on its own. --Draygo Korvan 01:28, 20 June 2012 (UTC)
I have to agree with Torrenal. The difference is in low bits. It's tiny. Now, there are images where it actually changes, but I haven't found a pattern to it. For example: http://i.imgur.com/xWeQ2.jpg That one clearly has a different color in the two sections. However, it's a rare occurrence and I have no correlation between that change and a material type. It only happens with select dyes and I have no idea what "material" is in those slots. Dosvidaniya 01:34, 20 June 2012 (UTC)
Now I have a real problem with listing different color by material, when the same material gets much more different colors than different materials. Warrants research? Yes, but gathering color by material just isn't going to cut it. Torrenal 01:52, 20 June 2012 (UTC)
Those two slots are leather, if i remember correctly. I havent found any inconsistancy. I know for a fact that material affects the color on the dye application part of the UI. This occured to me after I was comparing between multiple peoples notes and noticing the color variances. Some of it is harder to see, especially on the darker colors, some of it is easy to see, like in your screenshot. --Draygo Korvan 02:02, 20 June 2012 (UTC)
Anyway at this point I want to document it. If later on we determine it to be useless, then we can simply remove it from the template. Or maybe someone will come along with an idea of how to display it better. --Draygo Korvan 02:07, 20 June 2012 (UTC)
If you have noticed that it is always consistent to those three shades, then we may as well go that route. I'll try to dye different material types next BWE so I have a good selection. One question: Does the selection palette have its own type or is it one of the three materials? If not, the template should be modified to only have the three options. -- Dosvidaniya 02:29, 20 June 2012 (UTC)
thats the one i'm really not sure of. For example the material type and might be affecting what kind of adjustment it gets. The one leather square i did screenshot matched the selection, but it was also natural leather type. So I need more information. I just needed a way to standardize the color palettes between people taking them from the dye application part, vs the selecter. Especially when we run across an easily visible difference. I noticed it after the fact when I went through enough of my screenshots and some of yours when i was trying to answer the question of why my colors were different (some by a little, some by a noticeable bit) --Draygo Korvan 02:42, 20 June 2012 (UTC)
I've realized a few things. (1) Light armor is shown in the palette. (2) Town armor doesn't adhere to class restrictions at all. (It's leather w/ metal). (1) I wish I had a good screenshot, but I don't. If you look at http://i.imgur.com/WAF3v.jpg , you can see that the selection palette doesn't match the dye at all. I was using that dye on my light armor for pure black; it was darker than everything. I didn't use my normal armor for the dye screenshots because the darkness made it seem glitched. (2) I pieced together these three images to see the difference between metal and leather. http://i.imgur.com/ujFjA.jpg --Dosvidaniya 04:08, 20 June 2012 (UTC)

Icon[edit]

I would much prefer to have the actual dye icon in the icon box. The color blob is really ugly, and the color is already inside the infobox either way. — Rhoot User Rhoot sig.png 00:11, 22 June 2012 (UTC)

I agree that it would probably make it look better and the color is already shown below. Is anyone opposed to the change? Also, do you have access to the icons? I've looked through the list and don't see them. --Dosvidaniya 14:26, 22 June 2012 (UTC)
Dye icons are – for some reason – another size. I just added them to the list. They're at the bottom of BWE2. Also updated the BWE2 zip. — Rhoot User Rhoot sig.png 15:35, 22 June 2012 (UTC)
Some of them are uploaded already as well, such as File:Fine Dye.png and File:Masterwork Dye.png. Someone started doing it and I think classifying icons by rarity could be the way to go, as well. — Rhoot User Rhoot sig.png 15:37, 22 June 2012 (UTC)
Thanks Rhoot. Do you have any preference on the default icon for starter dyes? Since we'll never see them, they won't ever have an icon. I'm defaulting to the square fine dye currently. -- Dosvidaniya 16:58, 22 June 2012 (UTC)
Square fine dye is fine (pun unintended). :) — Rhoot User Rhoot sig.png 17:01, 22 June 2012 (UTC)
Yeah I left for a few days, I was going to make this change considering i moved color down below. I see you went ahead and did it so were good. --Draygo Korvan 17:02, 24 June 2012 (UTC)

modifications[edit]

I was editing this template to add semantic properties, and I ended up making additional changes - too many to describe in the edit summary.

  • Removed the titlecase-enforcing #switch: statements around the hue/set/material parameters. I ran a DPL parm check, and every dye on the wiki already uses titlecase on these parameters anyway, and I think we have 99% of the dyes documented, so it's really no longer necessary.
  • Removed whitespace in the auto-cat code. I'm guessing this was there in order to ensure page content began below the infobox, but that would be better served with the {clear} template (which all dye pages appear to use anyway), so again it wasn't necessary.
  • Removed the auto-cat that stubbed dyes without a {name} parameter. Only 1 dye, Abyss Dye (red), actually needs this parameter; it was misleading to make it look like it was "required" on every dye. I might run a bot to remove it from all other dyes, or I may not feel like bothering, I dunno.
  • Removed the link to Category:Dyes from the "Color" label. Not every label in an infobox needs to be linked, and this link in particular was completely redundant with the links on the other labels.

Dr Ishmael User Dr ishmael Diablo the chicken.png 20:35, 5 February 2013 (UTC)

Categorization capitalization[edit]

So, it looks like there's a new set of dyes ("Flame and Frost") but it's being categorized under "Flame and frost". Right now it looks like it's setting any set to UCfirst. Should we leave it that way, or have it be just the text or what? --JonTheMon 22:06, 15 April 2013 (UTC) Edit: Just having the text should be fine since all categories automatically capitalize the first letter and leave all the others alone and the current sets are all single words. --JonTheMon 22:08, 15 April 2013 (UTC)