To give you Photoshop software users to detailed analysis of the 3 visual information map to share the color matching skills.
While good color schemes are now handy, it's still hard to find the right color scheme for data visualization. Because of the unique attributes of the information map, it is necessary to satisfy the rich and not messy color matching requirements while ensuring the clarity of color recognition. However, even if you are not sensitive to color, with today's dry goods mentioned in the 3 tips, you can easily create a good-looking information map.
In the case of information graphs, things are even trickier, because we have to pass thousands of different sets of data to convey information, they have different visual performance.
Instead of setting up our own color table immediately, we started a survey to study the color schemes that already exist on the network. Surprisingly, only a handful of them are designed for complex graphs and data visualization. We found some reasons why we can't use the existing color matching.
Question 1: Low degree of identification
Many of the color schemes we've seen do not apply to data visualization. Not only because the lightness of color difference is not small, in fact, they are not considered in the creation of identification. The Flat UI color is one of the most widely used colors, for obvious reasons: it's very good. But, as its name describes, this is designed for the interface. Color blindness makes it difficult to recognize data images using flat UI colors.
Flat UI Color Complete color, red blind mode, grayscale mode.
Question 2: Not enough color
Another problem is that many existing color schemes do not have enough colors. To create a data visualization map, we need at least 6 colors of color scheme, and sometimes 8 to 12 colors to meet all the application scenarios. Many of the color schemes we've seen don't have enough color to choose from.
Here are some examples of color hunt:
Although these are great colors, they are not flexible enough to provide a rich color.
Question 3: Difficult to distinguish
But wait a minute, there are some color schemes that look like gradients--theoretically, you can create any number of colors, right?
Unfortunately, their lightness difference is usually small, and many of these colors can easily become indistinguishable, like this group, also from color Hunt:
We try to choose the first group and extend it to level 10 color:
If the normal user can distinguish these colors correctly, and corresponding to the corresponding data items, I will take, especially to distinguish the left 4 kinds of green.
In Graphiq, we take data for life and spend a lot of time looking for a color scheme that can be used for data visualization, not one group, but many groups. We have benefited a lot from this process and intend to share these guidelines to create a flexible color:
1th: The tonal and the brightness span all must be big
To ensure that color matching is very easy to identify and distinguish, their lightness difference must be large enough. The difference in lightness needs to be considered globally. Choose a one-color color, and test its performance in red blind, green blindness and grayscale mode. You will be able to quickly understand this color recognition level.
Google material color with the full color of light blue, red blind mode and grayscale mode.
However, there is a group of lightness span large color matching is not enough. The more colorful the color, the easier it is for users to associate data with the image. If you can make good use of the changes in hue, it will be easier for people who are not color-blind.
For lightness and hue, the larger the span, the more data can be loaded.
2nd: Imitate the natural color
Designers know a little secret, which seems counterintuitive to rational parties: not all colours are equal.
From the point of view of pure mathematics, the transition from pale violet to deep yellow, and the transition from yellowish to deep violet, is probably similar. But as we can see below, the former feels natural and the latter is not.
This is because we are accustomed to the gradients that have long existed in nature. In the gorgeous sunset, we can see the bright yellow to the dark purple gradient, but there is no place to see the light purple to the dark yellow transition.
Similarly, there are light green to Tibetan blue, goose yellow to dark green, brownish red to bluish gray, and so on.
Since I can always see these natural gradients, we feel familiar and happy when we see the corresponding color in the visual chart.
3rd: Use gradients, do not select a series of fixed colors
Gradient colors combine different shades and are best for both. Whether you need 2 colors or 10, the gradients can extract these colors, making the visual chart feel natural, while retaining enough tonal and lightness differences.
Changing to a gradient is not easy, but there is a good way to pull the guides to the breakpoint position in Photoshop, to correspond to the number of data, and then continue to test and adjust the gradient. Here's a screenshot of what we're doing when we fix the gradient.
As you can see, we put the color table next to the grayscale gradient at the top, adjust the gradient overlay (then we get the exact gradient value), and then choose the color from those breakpoints to test the effect of the color matching in practice.
We were excited about the final result. Here are some of the colors we use, they all have a gradient from pure white to black, to maximize the lightness difference.
Cool colors, warm colors and neon colors.
The practical application of color matching
Although there are more and more good color schemes, not all of them apply to charts and data visualization. Our color method is to create a natural gradient that is large enough for tonal and lightness changes. Doing so will make our colors easy to color-blind, more visible to others, and 1 to 12 of data to be met.
In this process, we found some great resources and articles, similar to our conclusions, but they used more precise methods, and even studied the color theory. We feel we should share it for you to read in depth:
How to avoid the linear HSV color, author Gregor Aisch through Chroma.js control of the tonal ratio of colors, author Gregor Aisch Subtle colors, author Robert Simmon Green color scheme, author Bob Rudis, Noam Ross and Simon Garniermatlab color map, author of Steve Eddins
Data Color Collection Tool-a very hand tool that allows you to keep your concentration unchanged while easily selecting a color chroma.js--a JavaScript library that handles colors colorbrewer2--hotspot map and data visualization Color tool with multiple tones and single tone schemes
We've also found some other coloring resources that we can't fondle. Although they are not designed for data visualization, we think it might be helpful.
colorhunt--High quality color scheme, able to quickly preview, if you only need 4 colors, this is the perfect resource colourlovers--great color community, there are many tools to create color schemes, as well as design patterns Colorschemer studio-- Powerful Desktop color application coolors--lightweight random color generator, you can lock the color you want, and then replace the other flat UI colors--great UI color, which is one of the most popular colors material design colors-- Another set of excellent UI colors. Not only does it offer a huge span of color, it also provides a different "color depth" for each color, or a palettab--chrome plugin that presents a new set of color schemes and font inspirations on each tab Swiss Style color picker-- Another excellent set of color schemes.
Well, the above information is small compiled to you photoshop this software users of the detailed 3 visual information map of the color matching techniques to analyze the full content of sharing, you see the users here are very clear now, I hope that the small part of the information shared above can bring useful help.
