**Participants.** 20 participants will be recruited through word of mouth and campus mailing lists. Given the Covid-19-related situation, a large proportion of participants will most likely be member of Martin Rolfs' lab. Each participant will complete two sessions of approximately 1 hour. Participants will receive 8.5 Euros per session, plus 2 Euros for every 15 minutes of overtime and a bonus of 1 Euro for completing both sessions. We will obtain written informed consent from all subjects prior to inclusion in the study. **Apparatus & Stimuli.** Same as Experiments 1 and 2, see [https://osf.io/94785/][1] and [https://osf.io/2t9nw/][2]. ![Figure 1][3] *Figure 1.* Illustration of experimental paradigm. **Procedure.** Each trial will start with the presentation of the fixation marker (left side of screen at 8 dva horizontal eccentricity relative to screen center) and the saccade target marker (right side of screen with a distance of 16 dva from the fixation marker) displayed on a black background. Instructed saccade directions may vary: While in previous experiments fixation and saccade markers were horizontally aligned, directions now are uniformly sampled from +22 to -22 degrees (0=horizontal, rightward). After 450 ms of successful fixation within a circular boundary with a radius of 2 dva around the center of the fixation marker (trial abort after 2 seconds of not fixating or 20 refixations), the backlight of the monitor will be turned off, which constitutes the cue to make a saccade to the previously presented location of the saccade target. In the meantime, while the backlight is turned off, the image of the scene is presented, but invisible to the observer. As soon as the saccade is detected (see Online saccade detection), the tachistoscopic presentation schedule will be initiated, i.e., the backlight of the screen will be turned on for the defined presentation duration and then turned back off. If no saccade is detected within a time window of 2.5 seconds after cue onset, then the trial will be aborted. The same scene will reappear after an ISI of 500 ms, starting from when the backlight is turned off after the gaze-contingently triggered presentation (with durations ranging from 8.3 to 66.7 ms). Right upon reappearance of the scene, the observer will be able to adjust a linear motion filter to adjust the perceived amount and direction of intra-saccadic motion blur (Figure 1). They may adjust the filter using ArrowRight and ArrowLeft to increase and decrease the length of the motion filter, as well as ArrowUp and ArrowDown to increase and decrease the angle (angle increases will lead to increasing upward components). To speed up convolutions and thus give observers the possibility to see changes in motion filter adjustment more fluently, motion filters will be run on the graphics card and on an image matrix downscaled to 50% of its initial resolution (for presentation, the image matrix is again upscales to its initial resolution, using bicubic interpolation). Motion-filter angle can be adjusted at steps of 1 degree, whereas length can be adjusted at steps of 10 pixels, amounting to ~0.27 dva. Feedback will be given if gaze position did not fall within a 5 dva circular boundary around the saccade target ('Saccade did not reach the target area'), if 2 or more saccades were made to reach the boundary ('Please make one saccade (you made XX)'), or if a wrong response key was pressed ('Wrong response key pressed.'). Trials, in which any of these events were detected, will be appended to the list of trials, causing a repetition at the end of all regular trials. Each session will be subdivided in 8 blocks. ![Figure 2][4] *Figure 2.* Distribution of presentation offsets relative to saccade offset for the eight presentation durations in two pilot sessions. **Design.** Each session will consist of 352 trials (22 trials per experimental cell), resulting in 704 trials in total. The number of trials is determined by the following experimental factors: - Presentation duration (8 levels). Intra-saccadically triggered presentations will have one of eight possible durations, i.e., 8.3, 16.7, 25, 33.3, 41.7, 50, 58.3, or 66.7 ms. By varying presentation durations, we will cause variable presentation offsets relative to saccade offset, depending on each observer's saccade duration distribution (Figure 2). - Availability of color (2 levels). Stimuli will be either presented as color or grayscale images. - Full-field or Gaussian aperture (2 levels). Full-field images will cover the entire screen, whereas in the Gaussian-aperture condition random subsets of the scene stimulus will be presented through a Gaussian aperture which can have standard deviations of 0.5, 1, 1.5, 2, or 2.5 dva and are randomly assigned in each aperture trial. Aperture positions were also chosen randomly in a range of +/- 8.7 dva horizontal and +/- 5.8 dva vertical eccentricity relative to screen/image center (see examples in Figure 3). ![Figure 3][5] *Figure 3.* Examples of full-field images and aperture stimuli. Shown apertures sizes are 1 dva (top row), 2 dva (middle row), and 2.5 dva (bottom row). **Online saccade detection, and tachistoscopic display.** Same as Experiments 1 and 2, see [https://osf.io/94785/][6]. [1]: https://osf.io/94785/ [2]: https://osf.io/2t9nw/ [3]: https://osf.io/ba9mt/download [4]: https://osf.io/tke3p/download [5]: https://osf.io/dycbt/download [6]: https://osf.io/94785/