[navigation]
What is SCA
Seasonal color analysis (SCA) is a method that determines which colors best complement an individual's natural skin tone, hair color, and eye color. It is based on four categories, namely Winter, Summer, Spring, and Autumn. This analysis considers the color temperature, saturation, and brightness of an individual's skin tone to determine which season they belong to.
Seasonal color analysis has its roots in the Impressionist school, as artists used the colors specific for summer, autumn, winter, and spring to express the feeling of the time period in the painting. However, its use for selecting clothing and makeup exploded in the 1970s, when it became possible to publish high-quality color books accurately representing specific tones.
There have been many prominent books on seasonal color analysis, including “The Medically Based No Nonsense Beauty Book” by Deborah Chase, “Color Me a Season” by Bernice Kentner, and “Color Me Beautiful” by Carole Jackson. However, the basic ideas are the same.
This approach is helpful for a number of reasons:
- Helps discover new colors that would fit
- Explains why certain shades look bad on a specific person
- Gives a clear guide on selecting the right palette for each individual
SCA considers three primary factors while assessing an individual's color harmonization. These are skin undertones, skin depth, and bright or soft complexions. By understanding these factors, a trained color consultant or makeup artist can determine which seasonal category an individual falls into. Winter skin tones tend to have cool undertones with a deeper complexion and sharp contrast between their hair, skin, and eyes. Summer skin has cool undertones with a light complexion and low-to-medium contrast between hair, skin, and eyes. Spring skin tones possess warm undertones with a light complexion and high contrast between hair, skin, and eyes. Autumn skin tones have warm undertones with a deeper complexion and low-to-medium contrast between hair, skin, and eyes.
In real life, there are many different approaches to SCA. Some people hold different colored fabrics near the face to observe which shades make the wearer look radiant and healthy. Others just swap sets of makeup for each season and track how many compliments they get wearing each. Things get even more complicated when people don’t fall neatly into one category, so each season can be split into three sub-seasons.
How do you even begin to digitize something like this?
How TINT automated SCA
When developing TINT, Banuba’s team had two priorities in mind: making the most realistic virtual try-on in the world and creating an automated seasonal color analysis system. And it took a lot of thinking outside the box to achieve the latter.
One of the most common issues associated with SCA is that two equally experienced makeup artist could assign different categories to the same person. This means they needed a reliable point of reference to test their algorithms. The team thought about finding such a point and had an idea. They decided to partner up with a group of portrait artists. These people had a keen eye for colors and human faces – exactly what the developers wanted.
The second step was building the initial prototype. Banuba already has a wealth of experience in face detection and tracking, as well as detecting colors. The team had the software conduct seasonal color analysis on several people and showed the results to the artists. They weren’t impressed – according to them, the results were wrong.
It turned out that the hair was the culprit. Not only could different strands have different shades, they also cast shadows that can mess up the color perception of the camera.
Finally, the developers needed to polish the system and make sure it offers good performance and accuracy.
Conclusion
Seasonal color analysis is a complicated process to digitize. There are many factors that need to be considered, including individual’s skin tone, different aspects of each shade (saturation, warmth, hue), and even the shadows cast by the hair strands on each other. It took a lot of time and effort to do well, but the result is the only automated SCA on the market. It doesn’t require lengthy quizzes and works in seconds, offering a better user experience.
