Cognitive attention conditions of multimedia system users were represented by simple operations of two vector sets: stimulus and human information processing channels. A vector-set model illustrated how humans use cognitive attention resources. A comprehensive psychophysical scaling system (MMES) has been designed to measure subjective multidimensional human perception. Unlike paper-based magnitude estimation systems, the MMES has an additional auditory peripheral cue that varies with corresponding visual magnitude. As the simplest, purely psychological case, bimodal dividedattention conditions were simulated to establish the superiority of the MMES. Subjects were given brief presentations of pairs of simultaneous stimuli consisting of visual line-lengths and auditory white-noise levels. In the visual or auditory focused-attention conditions, only the line-lengths or the noise levels perceived should be reported respectively; on the other hand, in the divided-attention conditions, both the line-lengths and the noise levels should be reported. There were no significant differences among the different attention conditions (no divided-attention decrement). Human performance was better when the proportion of magnitude in stimulus pairs were identically presented (cross-modal congruity effect). The additional auditory cues in the MMES improved the correlations between the magnitude of stimuli and MMES values in the divided-attention conditions.