Online image duration timings are poor

Online experiments
,
21

Thanks for sharing, interesting to see.
Can you confirm whether these plots are with data from an experiment that used the original (large) image files? If so then I’d really encourage you to reduce the file size down so they are no bigger than they need to be. (just to stress, I’m not talking about making the image smaller on the monitor but instead in making it so they are no bigger in picel resolution (x & Y) than they need to be for their funal diusplay size)
I’d also suggest making the y axis scale the same for all plots so that it’s easier to compare them by eye.
cheers,
John

1 Like
22

thanks for the feedback!
and yes, the plots refers to images with original size. I’ll reduce that in case I’ll opt for implementing the experiment online.

Cheers

23

Hi wakecarter and John,

I reduced the size of the images to 25 kb and it has more or less fixed the problem of the presentation duration: now they are presented for ~1.47-1.49 s, which varies from trial to trial. This was achieved by presenting a static component for 1 s, immediately followed by the presentation of the trial image (programmed to be 1.5 s).

The problem now is that the static component introduces a gap between the end of the static component and the image presentation: the static component should be presented for 1 s simultaneously with a text fixation cross (‘fix_static_exptrials’) for 1 s, immediately followed by the presentation of an image (‘trial_image’) for 1.5 s. I’m adding a couple of lines of the output log file so you can see what I mean:

224.3228 EXP fix_static_exptrials: autoDraw = null
224.3228 EXP trial_image: image = {“_SBA4B_eventListenerList”:{“load”:[{“useCapture”:false}]
224.3488 EXP trial_image: autoDraw = true

I’ve also noticed in the output that the autoDraw = null of the static component occurs at the same time as the autoDraw = null of the trial image, despite not being programmed to overlap in the builder. I.e. in the builder there is 1 s of static component followed by 1.5 s of image presentation, such that there is no overlap. Whereas in the output, the static component seems to stop at the end of the image presentation. I wonder whether in the builder the static component and the image presentation should in fact overlap, rather than being set as one after the other?

Or if there was a way to have a fixation cross “in the background” that covers up the gap left between the fix_static_exptrials and the trial_image? I am using a jittered ITI that consists of a text fixation cross (which would be immediately followed by the ‘fix_static_exptrials’), although they are not being accurately reflected in the log output. But as long as they are presented within a certain range of values (1.5 to 2.5 s) I can live with the variability.

Cheers,