Very nice write up! The one thing that jumped out at me was the picture of the eye in your sample palette images. That blue was stunning in the photograph, but it doesn’t show up in the palette! I wonder if Color Thief might have missed more accent colors like that in your run of covers, leading to the noted skew toward darker colors.
Agree. I did a bit of searching but wasn't able to find a clear definition of how the dominant color is determined. It must have to do with how much area is covered vs. just contrast. The eye is a great example, because the blue takes up a very small space, and yet it's visually striking. So the overall mix from the covers is definitely going to be skewed in that way.
That's a super nice newsletter. I need to do some digging into colorblind-friendly palettes and see how I can use them for the games I publish. I did a few low contrast covers that might not be ideal.
Wonderful article! I'm a scientist loving to explore layout and design for my research communications (slide decks, posters, articles). It's always good practice to have colorblind-friendly design in mind. Thanks for the link to Color Brewer, I'll add it to my list of resources!
Personally, I'm amazed how well the "megapalette palette" turned out, and the image for your sorted megapalette looks beautiful! I was wondering what sorting criteria you used, because it was not what I expected. I'll have to read up on the color sorting problem.
Thanks for your code snippet btw, I'm currently reading through the documentation for Pillow to create the image including color palette.
Thank you for the kind words! I'm so glad you enjoyed this one!
Color sorting can be complex, not because it is "hard" but rather because there are so many different ways to sort colors. Each way is valid; some are more involved than others. I used the easiest method where you sum the R, G, and B values. That gives you a single integer number for each color, and a sort can be done easily. I think a different sorting method would yield a prettier result.
I really should get the code up on Github. I'd be happy to share the code if it helps, or if you want to do it on your own as an exercise/practice that's awesome too!
Very nice write up! The one thing that jumped out at me was the picture of the eye in your sample palette images. That blue was stunning in the photograph, but it doesn’t show up in the palette! I wonder if Color Thief might have missed more accent colors like that in your run of covers, leading to the noted skew toward darker colors.
Agree. I did a bit of searching but wasn't able to find a clear definition of how the dominant color is determined. It must have to do with how much area is covered vs. just contrast. The eye is a great example, because the blue takes up a very small space, and yet it's visually striking. So the overall mix from the covers is definitely going to be skewed in that way.
That's a super nice newsletter. I need to do some digging into colorblind-friendly palettes and see how I can use them for the games I publish. I did a few low contrast covers that might not be ideal.
Wonderful article! I'm a scientist loving to explore layout and design for my research communications (slide decks, posters, articles). It's always good practice to have colorblind-friendly design in mind. Thanks for the link to Color Brewer, I'll add it to my list of resources!
Personally, I'm amazed how well the "megapalette palette" turned out, and the image for your sorted megapalette looks beautiful! I was wondering what sorting criteria you used, because it was not what I expected. I'll have to read up on the color sorting problem.
Thanks for your code snippet btw, I'm currently reading through the documentation for Pillow to create the image including color palette.
Thank you for the kind words! I'm so glad you enjoyed this one!
Color sorting can be complex, not because it is "hard" but rather because there are so many different ways to sort colors. Each way is valid; some are more involved than others. I used the easiest method where you sum the R, G, and B values. That gives you a single integer number for each color, and a sort can be done easily. I think a different sorting method would yield a prettier result.
I really should get the code up on Github. I'd be happy to share the code if it helps, or if you want to do it on your own as an exercise/practice that's awesome too!