Quote Originally Posted by Airget View Post
Simple really. Let's say you have 24 unique colors, and now you are capable of putting 2 colors on 1 piece of gear. So with the ability to use the same color twice you end up with 24x24 which is 576 combinations. So pretty much with the current dye system each piece of gear 1=24, however by increasing the amount of colors you can put on a piece of gear you exponentially increase that number to 1=576 per piece of gear, which of course is 24 times more than before. It adds up. If there were 100 piece of gear in the game, the old dye system would have 2400 variables, now it has 57600 variables, and that number only gets higher and higher the more pieces of gear you introduce that you allow 2 dyes.
That's not how colour variations work in games. The number of possible combinations is of no consequence to the hardware demands. It's really not that simple at all.

Most dye systems simply apply a modifier tint over the base texture with the default colour just being a unique default tint. That tint would be determined via an RGB colour value applied based on the ID of the dye assigned to that channel. Each tint application would be a simple multiplication with the base colour RGB value in the shader. If the artists are being efficient they may have a way to mask each tint application so that the combined dye channels are only a single rendered material, which would mean it would still take up the exact same amount of memory. The only difference that more dye channels would make is that the renderer would have to perform a few extra multiplications for each dye. The number of possible combinations would be irrelevant and there wouldn't necessarily be a significant increase in memory usage.

The only real limiting factor would simply be that making a piece of equipment would suddenly be a bit more work, and over time this would result in either lower quality or lower production overall.