A visual experiment was performed to evaluate the effect of black separation on spatial image quality. Two types of black printer algorithms were evaluated. One closely simulated conventional printing with the exception of yielding colorimetric matching. The other employed a very simple algorithm based on minimum and maximum amounts of black and also employed a colorimetric paradigm. This second type is significantly easier to implement. Using pictorial and synthetically generated gradation images, two visual experiments were performed to understand the tradeoffs between processing complexity and spatial image quality. Two groups of observers from either academia or the printing industry evaluated six images. For pictorial images, the simple algorithm using the minimum amount of black resulted in the highest spatial image quality. However, this algorithm produced color contours, visible in the gradation images. Thus for general purposes, the conventional separation algorithm or the simple algorithm using nearly the minimum black amount is recommended.