Cameras offer new possibilities in assessing human states like stress. This work addresses the measurement of respiration rate with several approaches. Data from a mental stress study (> 40 h recording time, 54 participants) were evaluated and color channel combinations were examined using a hemispherical surface grid search with mean absolute error M AE to optimize the approaches. The grid search converged towards the green channel in the baseline modulation approach (MAE = 2.53 rpm). However, best results were achieved with the frequency modulation approach (MAE = 2.13 rpm) with the color channel combination optimal for heart rate measurement. Respiration rate increased highly significant during stress (p < 0.001, Mann-Whitney U-test). Deliberate selection of the color channel combination is crucial for respiration rate measurement with cameras in order to assess human states or pathophysiological processes.