Objectively predicting the discernibility of color differences is a common requirement when assessing the performance of a display device and is the domain of color difference metrics. Metrics commonly used for this task assess the color discriminability of two stimuli based directly or indirectly on the viewing environment with a known, constant adaptation luminance level L_A. These metrics were originally derived for color assessment of reflective and transmissive media as well as low dynamic range displays, where L_A can both be maintained and estimated with reasonable certainty for any likely stimulus pair. With High Dynamic Range (HDR) and Wide Color Gamut (WCG) displays, using Steady State metrics is becoming increasingly challenging when assessing the discernibility of two similar stimuli over the display's full range of capability. This is especially true if spatially or temporally varying HDR content is displayed, causing the Human Visual System (HVS) to frequently and unpredictably change L_A, and with that visual sensitivity. To overcome these challenges, we present the concept of an "Adaptation Hull" color difference metric. Rather than using a specified adaptation luminance that is in most cases substantially different than the stimuli under test, an Adaptation Hull metric instead considers an optimal adaptation state where the HVS has the highest sensitivity to color differences.