Title: Color Rule - evaluation and extension
Author: Alia
Date: 27.01.06
1. Abstract:

This blue note will discuss my proposal on color picking evaluation and improvements to the color picking rules. The end goal is to complete the color picking rules which eventually complements the design rules.

This blue note is also came out from talks with Zeljco about solutions on evaluation of color assignments..

2. Previous Work:

Ref 1. Creating harmonious and legible color schemes in the automated generation of multimedia presentation - Amit's thesis
Ref 2. Color picking - the pecking order of form and function by Frank, Amit and Lynda
Ref 3. Universal Design Guidelines


Ref 1. covers all the basic guidelines: different types of color blindness, theory on color harmony, proposal on implementation of color harmony theory in rules for typography.
Ref 2. is an excellent discussion on the color rule and implementation in Cuypers architecture in more detail. Also discusses color rules priority and decision making. What I find great is that the system already have importance weight factor to allow swap between functional and formal presentation strategy and the attempt to formalizing it with three sets of formula. There are several things that I would change in the architecture that i hope would make it better e.g. communication preference (Form and function should be present in any presentation its not a 1 or 0 discrete value, but rather a continous 100% to 0 % situation. Communication preference should be about choosing a  genre and a template).
Ref 3. discuss the color harmony theory and also the importance of taking space where the color is assigned into consideration.

3. Discussion

There are several points which might improve the color picking system:

1. I am not sure if I agree with the interpretation of importance area in the 3 formulas (page 10).

Where does this formula come from and why is it formulated this way? I notice that the color chosen by the system are very "polarized" and don't seem to work well with each other. Another question is if the HSL system is the best metric to use. I would like to have a color system which allows me to shift gradually between different brightness and saturation and allow me to get to contrast and complementary colors easily. I like the fact that the system takes contrast into consideration, but selecting only colors that contrast each other will create the effect of disintegration on the whole presentation. A suggestion would be only applying contrast rules to text vs. background and not between the backgrounds of different area.   

2. I am also not sure if I agree on the color assignment process. The system chooses a specific area and makes decision on assigning colors only within this area. After completing one area, the system goes to other areas in the presentation and do the same process. In short, there maybe different color schemes for different areas. If the system works the way I think it does, this is where the system fails to keep the presentation united and there is a danger of generating a disintegrated presentation. 

3. There is a need to have and extra color rule based on genre and template. Actually its more important for Genre rather then template. Each genre of presentation wants to convey a certain impression to the reader. Impression can be implied by, among the important things, colors.

It is no supprise that certain color (combination) have inherit certain characteristics which generate certain impression depending on culture and person. This is important when designers are trying to tell their story (for our semantic design system: conveying plot to story) the color chosen acts as a communication device. e.g. black dominant for thriller/horror films, light colors for e-learning to create the calming and safe impression.

Example in a Internet platform: webpage.

Genre Impression Colors that suggest this impression
1. E-business Trust, professional, sophisticated, white, blue, cool colors, medium-to-low saturation or low brightness.
2. E-learning Non-intimidating, friendly, comfort warm and bright colors, pastel colors.
3. Online Games Exciting, engaging, attention colors with high saturations and contrast
4. E-museum Simplicity, Clear, non-intrusive white, grey, colors with low saturation and high brightness.
5. Applications for children Fun colors with high saturation and high brightness.

4. Steps to be taken for improvements

1. have knowledge of the characteristics of the colors (sets).

Note: Experienced designers know which color set to use for different kinds of applications and often keeps records of them (in a digital palette). We can try to do the same by annotating the colors.

The investigation can be done empirically by examining online webpages (categorized by different genres), analyze the color scheme which they use, and evaluate what impression does these website gives to users. The result of the evaluation will be a set of characteristics own by a (set of) colors (in annotated form). e.g. dark blue:professional,trust,sofisticated; red : bravery,romantic,scary.

Analyzing the color scheme which they use can be done semi-automatically by taking screendumps of the website, "scan" them to get the color spectrum (thanks to Zeljco's program) and how big the occupy the presentation space.

The end result will be annotated (sets of) colors which can be used in the process of color picking. The result of this research itself is interesting for the multimedia and HCI community.

2. Refine the color picking system

I would like to reexamine the color assigning process again. Not to rely on the formulas of Ref 2.

The basic theory and guidelines are the same, but what will be different is how to interpret them into rules, how to assign their importance score, and how to manage these important scores to make decisions.

Note: I am still thinking about how to it should be....to be continued.