The BeadSprites Forum
Welcome to the Beadsprite Forum, a place to show off, share, and comment on pixelized bead sprite creations.
 
Log in to check your private messages

Log in  Register


Profile  Search  Memberlist  FAQ  Usergroups
Indexed Color

 
This forum is locked: you cannot post, reply to, or edit topics.   This topic is locked: you cannot edit posts or make replies.    The BeadSprites Forum Forum Index -> Free Talk
View previous topic :: View next topic  
Author Message
spires



Joined: 09 Jul 2008
Posts: 4
Location: New England

PostPosted: Wed Jul 09, 2008 10:13 pm    Post subject: Indexed Color Reply with quote

A great thing about the indexed color mode, is that it excludes any colors that are not chosen to be part of a given palette, conversely using only those colors specified as included in a palette. Indexing is common; for instance, the gif file format uses indexing.

For instance, say there are 60 solid hues across hama and perler, respectively. With indexing, you can create digital representation of those hues, and define a palette.

Next, you can save that palette and then color covert any sprites you want to bead, to that palette. The computer will match as closely, the colors in the original sprite, to what the bead sprite should be, based on available colors.

If after mapping, you feel as though the computer did not optimally choose colors, you can customize the conversion using hues of your choice, from the palette you defined earlier.

Generally, the computer (photoshop in this case) does a fantastic job finding the appropriate approximation, and converts the colors very satisfactorily.

This process rests on having an exact-as-possible, HEX value defined for each color. I think it would be a significant step forward, to have universally recognized, defined hex values for each of the perler and hama solid colors.

It is hard to use scans and photos to come up with this hex value, because even the best scan and blur can alter a color. Alternatively, eyeball conversions can enter the author's interpretation of a color into a palette, and therefore I think a mix of the two methods would probably create the most accurate sampling.

What do you think?
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address MSN Messenger
Nicholas



Joined: 08 Jul 2008
Posts: 14

PostPosted: Wed Jul 09, 2008 11:54 pm    Post subject: Reply with quote

I've also used Photoshop's indexed color conversion with great success. I had to tweak some of its picks, but I blame that on having incorrect color hex values in the first place.

You're right about getting the colors into a palette being the hardest part. A camera is probably the worst because the lighting conditions have way more to do with the resulting colors than the beads themselves.

A scanner works pretty well, especially if you can switch into an advanced mode and turn off any enhancements they try to help you out with. My current palette is based off making a little 2x2 square of each of my colors and scanning the resulting block. You do have to be a little careful because the lighting on the scanner bed is also still pretty significant. Also, if the colors are the least bit transparent or translucent, the background color on the scanner is also important.

What might work best (even though it relies on subjective eyesight) would be to get a book of Pantone colors or some other documented, physical color reference that you could compare the beads to. Because they both exist in the "real world" it's the one way to discount environmental lighting issues during the comparison.

One final point worth mentioning is that it might not be enough to boil a bead down to a single hex value. Some of the other properties (opacity, reflectivity) might be useful to know for special effects (illusion of depth, etc.)

I think a complete bead taxonomy would treat it less like a color in Photoshop and more like a "material" that you might find in one of the 3D packages out there. It gets a little more ambiguous to "index" a sprite down the palette, but so long as we're working toward some de facto bead reference, at a minimum it would be nice to include an alpha with the RGB hex that represents its percentage of light transmission -- especially since something like a full third of the colors out there seem at least a little semi-transparent.
Back to top
View user's profile Send private message
spires



Joined: 09 Jul 2008
Posts: 4
Location: New England

PostPosted: Thu Jul 10, 2008 12:05 am    Post subject: awesome Reply with quote

Hey Nicholas!

I'm intrigued by your approach.

Which colors have you found, besides the intentionally transparent ones, that have a level of transparency? i don't currently have all the colors.

I mean, the "solid" colors, excluding neon, gitd, glitter, transparent.
Back to top
View user's profile Send private message Send e-mail Visit poster's website AIM Address MSN Messenger
Nicholas



Joined: 08 Jul 2008
Posts: 14

PostPosted: Thu Jul 10, 2008 12:26 am    Post subject: Reply with quote

So, the off-the-cuff "one third" I claimed is mostly made up of the neon, fluorescent, glitter, and those types of colors. (I don't have any idea how many of those colors there are, I guess.)

But, just digging around in my bag of mostly-solids, I do notice some that let a lot of light through. Most of the lighter colors do it to some extent, but that's not really their fault. The colors I really noticed it in, though, were a couple of the reds. The worst offender is Hama Dark Red. Next is Perler Magenta (though that might technically be considered a fluorescent, I'm not sure). You can even faintly detect it in the Hama Reddish Brown.

I guess what I'm seeing is what's usually described as "subsurface scattering". The most opaque colors (dark browns and grays, but even some lights like Perler Cheddar) look like a solid object. But, when I look at Hama Reddish Brown, it's like I can sort of see "into" the bead.

It's subtle, but again, when you melt these things down (especially enough to get rid of the holes) the result is pretty thin. Any amount of light transmission means the backdrop you put a finished piece against plays a pretty big part in how it actually looks.

You could probably get away with faking colors that way: use some of the transparents or something like Perler Pearl, and then print out a matching backdrop that it would be set against with little squares of color the size of each bead. If you got the "material" properties of the beads right and the algorithm to combine it with a background, you could probably match any color you wanted.
Back to top
View user's profile Send private message
Display posts from previous:   
This forum is locked: you cannot post, reply to, or edit topics.   This topic is locked: you cannot edit posts or make replies.    The BeadSprites Forum Forum Index -> Free Talk All times are GMT
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum

phpBB "skin" by DewChugr


© 2007-2008 Informe.com. Get Free Forum Hosting
Powered by phpBB © 2001, 2005 phpBB Group
 ::