During the saccade a burst of flicker will be seen irrespective of the field scan direction, but because the effective frequency is below the RFTF for L-R, R-L, and B-T, field scanning the flicker will have a directional component, whereas the flicker will be spatially uniform when the screen is scanned T-B because of the lower RFTF in this direction.
A pilot study showed that experimentally naive participants had considerable difficulty in establishing a reliable criterion for the RFTF, and so all were given several practice sessions (with the display scanned L-R and R-L) before the main experiment.
The FFF and RFTF represent the upper frequency threshold of these two mechanisms, respectively.
This was confirmed by the demonstration that the FFF is independent of field scan direction (as long as the eyes are stationary), but the RFTF is significantly lower for rasters scanned T-B than for rasters scanned in other directions.
This results in a burst of flicker with or without a directional component, depending on whether or not the frequency is reduced below the RFTF. Therefore a higher refresh rate is required to eliminate screen flicker when voluntary eye movements are permitted, particularly when the field is scanned L-R, B-T, or R-L, because eye movements are more likely to produce a burst of flicker with a directional component.
It was shown that viewing a TV scanned B-T for 1 h is sufficient to reduce the RFTF for rasters scanned B-T but is insufficient to reverse the FSDP.
After a short break, we determined RFTFs by the same procedure except that the participants were instructed to ignore the flicker and respond according to whether or not a directional component was perceived.
The RFTFs for rasters scanned L-R, R-L, and B-T were between 1 and 5 Hz lower than the corresponding FFF and increased in parallel to the FFFs with retinal illuminance (Figure 8).
If the FSDP arises because of adaptation, some correlation between TV/VDT viewing history and the asymmetry in RFTFs might be expected.
The RFTFs for rasters scanned T-B and B-T were measured as follows.
Table 4 shows the results in terms of the difference between RFTFs (Hz) for rasters scanned T-B and B-T.
The RFTFs for rasters scanned B-T were similar to those measured under the same conditions in Experiment 4 (i.e., between 45 and 55 Hz).