Exploring the frame effect

Abstract

Probes flashed within a moving frame are dramatically displaced (Özkan, Anstis, ‘t Hart, Wexler, & Cavanagh, 2021; Wong & Mack, 1981). The effect is much larger than that seen on static or moving probes (induced motion, Duncker, 1929; Wallach, Bacon, & Schulman, 1978). These flashed probes are often perceived with the separation they have in frame coordinates—a 100% effect (Özkan et al., 2021). Here, we explore this frame effect on flashed tests with several versions of the standard stimulus. We find that the frame effect holds for smoothly or abruptly displacing frames, even when the frame changed shape or orientation between the end points of its travel. The path could be nonlinear, even circular. The effect was driven by perceived not physical motion. When there were competing overlapping frames, the effect was determined by which frame was attended. There were a number of constraints that limited the effect. A static anchor near the flashes suppressed the effect but an extended static texture did not. If the probes were continuous rather than flashed, the effect was abolished. The observational reports of 30 online participants suggest that the frame effect is robust to many variations in its shape and path and leads to a perception of flashed tests in their locations relative to the frame as if the frame were stationary. Our results highlight the role of frame continuity and of the grouping of the flashes with the frame in generating the frame effect.

Influence of frame and probe paths on the frame effect

“…We first compare the frame's effect on flashed tests (as in Özkan et al., 2021 ; Cavanagh et al., 2022 ) against its effect on a continuously moving test (as in Wallach et al., 1978 ). The frame is always in repetitive left to right motion ( Figure 1 , top; Movie 1 ).…”

“…This earlier literature on frame effects typically examined static ( Duncker, 1929 ) or continuously moving probes ( Wallach, Bacon, & Schulman, 1978 ). Recently, we have shown that moving frames give far larger effects for flashed than for continuous probes ( Özkan, Anstis,’t Hart, Wexler, & Cavanagh, 2021 ; Cavanagh et al., 2022 ). These large effects for flashed tests had previously been reported by Duncker (1929) and Wong and Mack (1981) .…”

Unveiling the time course of visual stabilization through human electrophysiology

“… 5 , 6 This perceived position shift occurs robustly across different visual environments such as apparent motion of a frame (instead of actual motion), whole background shifts instead of a moving frame (i.e., background pattern moving left to right), or 3D frame flipping motion. 7 In all cases, this frame-induced position shift yields a drastic perceptual effect, in some conditions resulting in 100% stabilization: The locations of dots are perceived in the frame’s coordinates as if the frame was not in motion at all. Such dramatic effects of perceived position shifts were taken as evidence that surrounding context (such as a moving frame) may act to stabilize the perception of relative locations during eye movements.…”

Unveiling the time course of visual stabilization through human electrophysiology

“…This perceived position shift occurs robustly across different visual environments such as apparent motion of a frame (instead of actual motion), whole background shifts instead of a moving frame (i.e., background pattern moving left to right), or 3D frame flipping motion (Cavanagh et al, 2022). In all cases, this frame-induced position shift yields a drastic perceptual effect, in some conditions resulting in 100% stabilization: The locations of dots were perceived in the frame’s coordinates as if the frame was not in motion at all.…”

“…This so-called Frame Induced Position Shift (FIPS) illusion elicits a paradoxical stabilization of a moving object: When a frame is moved left to right and two probe stimuli are flashed inside the frame at the exact same physical location shortly before and after the frame moves, participants report these probes to appear at separate locations that match the distance that the frame traveled (Özkan et al, 2021;Cavanagh 2020). This perceived position shift occurs robustly across different visual environments such as apparent motion of a frame (instead of actual motion), whole background shifts instead of a moving frame (i.e., background pattern moving left to right), or 3D frame flipping motion (Cavanagh et al, 2022). In all cases, this frame-induced position shift yields a drastic perceptual effect, in some conditions resulting in 100% stabilization: The locations of dots were perceived in the frame's coordinates as if the frame was not in motion at all.…”