Inconsistent Duration using the image component

Online experiments
1

OS: Win11
PsychoPy version: (V. 2025.1.0 Py 3.11)
URL of experiment: Till Kempf / t_incon · GitLab
(A simplified version of the experiment with the essential parts)

Description of the problem:
In an experiment, I would like to display an image for 0.1s in each trial. I have programmed the experiment accordingly, but when piloting I noticed that the images are not displayed constantly (too long or too short). I then logged the duration of the image component and found a large variance in presentation time. The deviations are sometimes so large that I don’t believe they are caused by the refresh rate (40ms+/-, sometimes more, see attached files).

What did you try to make it work?:

I have tried to set the duration in the following ways:

  • In seconds
  • In frames

In both cases the deviation is present (see attached files: timing_duration_s (for seconds) and timing_duration_frames (for frames), in both files the column “Prime_pres”)

Then I thought that the inaccuracy might be due to the fact that the images cannot be loaded fast enough, because some pictures are relatively large, around 2MB (the component was set to “set every repeat”, so the images should actually be loaded at the beginning of the routine, right?) I therefore used a static component at the beginning of the trial to load the images, but that didn’t help either (see timing_duration_static_component)

The only solutions I have found so far are:

  • Reduce the size of the image files, but I have to reduce them so much that I lose some details in the images

  • More of a work around than a solution: I set the image for the duration of the entire trial and present it by putting two rectangles in the background color in front of the image and showing these rectangles according to the presentation time of the image

If anyone knows how I can get better timing for the images or knows what is causing the inaccuracy (apart from the image size), I would be very grateful for an answer.
Thanks in advance :slight_smile:

timing_duration_frames.csv (6.4 KB)
timing_duration_s.csv (5.9 KB)
timing_static_component.csv (5.8 KB)

2

Hi Till,

(not a psychopy pro) but I always preoload my stimuli in an extra loop - preceeding the rountine I want to use it in. That solved a lot of issues for me. I found Wake Carters indepedent randomisation demo super helpful to set this up for my experiment. Hope this helps.

Or maybe this Wakefield's Daily Tips - #54 by wakecarter