Lawful kinematics link eye movements to the limits of high-speed perception

> In the extreme, the very limits of a sensory system’s access to the physical world might be defined not just by biophysical constraints, but further curtailed by the kinematic bounds of the motor actions that acquire sensory information. Conclusive demonstrations of such action-dependence of the limits of perception are missing, but a key prediction is that perceptual processes should be tuned to an action’s typical sensory correlates, even in the absence of the accompanying action11,12,13. Here we confirm this prediction for a fundamental perceptual process in human vision: We demonstrate that a shared law links the limits of perceiving stimuli moving at high speed to the sensory consequences of rapid eye movements.

I've read countless papers in fields where I have no expertise, but I'm stuck on this section.

> the very limits of a sensory system’s access to the physical world might be defined not just by biophysical constraints, but further curtailed by the kinematic bounds of the motor actions that acquire sensory information.

What is the difference between biophysical constraints and kinematic bounds of the motor actions? Isn't the latter a biophysical constraint? What other biophysical constraints are they talking about? I don't see them mentioned previously and there would seem to be very many.

> perceptual processes should be tuned to an action’s typical sensory correlates

What sensory correlates? That could mean many things. Is it a term of art in this field?

Someone in the know, or field, does this have anything to do with the phenomenon of looking at the second hand on clocks and feeling like the initial second is always longer??

I guess I'm looking for the target of this finding.

This might be interesting for foveated rendering techniques, and certainly it's interesting to know how the eye responds to "real" visual input, but it doesn't seem particularly relevant to standard displays. These images were presented at 1400Hz updates. That's necessary because saccades are 1000deg/sec events over 10s of ms up to >100deg, while the high resolution fovea is only ~1deg wide.

For standard raster scanned displays (CRT, LCD, OLED) upto 240Hz there are other effects such as "tearing", "flicker", and phosphor lag that are much more visible and distracting effects for "rapidly moving" objects. That's even ignoring video compression artifacts.