How does the color management work?

The following illustration shows the different color management steps which may be involved when printing a proof in GMG ColorProof.

If an input file contains lineworks or vector graphics, it needs to be rasterized first to translate the vector information into discrete pixels for printing.
The next step is only required for files which have not been converted to the target color space yet, for example a PDF containing multiple elements in different color spaces which need to be normalized to one color space first. This color conversion uses ICC profiles for the color transformation from one working space to another (Conversion CT/LW, 1).
The main color management starts in step 2, optimizing the color data for the printer color space by means of a proof standard. Each MX DeviceLink profile linked within a GMG proof standard has been calculated to perfectly simulate a target printing condition on a specific proof printer equipped with a certain print medium and ink set.

Step 3 brings in the printer calibration that separates the color data into the inks used by the printer and hands the data to the printer driver. The printer driver processes the color data and generates a Yes/No pattern for the printer, defining at which places each nozzle releases ink onto the print medium.

The quality optimization achieved by DeviceLink profiles only makes sense if the used printer-medium combination is regularly calibrated. Our key concepts—standardization and repeatability—can only be achieved if the printer is kept in tight tolerances by the calibration and recalibration capabilities of the software.

Simulation features:
All advanced color management settings such as noise or missing dots used in proof standards are applied both to the image and the control strip (except printer calibration control strip). This ensures a high quality, because the control strip is printed under the same conditions as the image. Thus, you can make sure that the target Lab values are still inside the tolerances and not significantly affected by noise or missing dots simulation.