Thorsten Braun is the CTO of ColorLogic GmbH and leads his development team in designing award-winning color management software for the printing industry. Thorsten was an early developer of ICC based color management and focuses on breakthrough color management technology. He is the winner of the 2021 Joe Clarke Innovator Award for Innovation in Software Technology. For 30 years, Thorsten’s experience has landed him on advisory committees to establish international print standards with the latest being Expanded Gamut Printing in a CMYK working space. He has worked at LOGO GmbH & Co. KG and GretagMacbeth AG, where he focused on the development of algorithms for color transformation and developed the well known ProfileMaker. When he’s not applying modern physics to win his next tennis match, you will find him backpacking while thinking about future innovations in color management technology. 

 Methodology of Developing a Color SpaceTimon Braun, CrossXColor, Inc

Thorsten Braun and Timon Braun, ColorXColor, Inc.

In the pre-press and reproduction industry, colorimetric color models play a big role in processing and manipulating colors. However, even the most accepted standards of such models still face issues with correctly representing colors and any changes applied to them. Therefore, this paper presents a methodology of developing a new color model with more consistent perceptual properties. Applications of such a model, like color processing, color management, and color manipulation, should have improved accuracy. A crucial aspect of such a color model is consistent and defined conversions for common data exchange spaces, like CIELAB.

An important part of such a methodology is a method to evaluate models. This will be done by comparing set pairs of colors with given properties, in terms of lightness, chroma, and hue. This method can analyse how well a color space reflects the expected given properties.

We will also discuss different mathematical modules to obtain stability when converting between standard color exchange models. This is necessary to process color values reflecting non-real colors. For example, the CIELAB color (100, 100, 100) is a valid combination in the CIELAB color space, however it cannot physically exist as a real object. Still, we must be able to process such triplets.