How to present 1000+ images in a run

Hello everyone,
I come from a vision lab and I'm trying to build a task to present 5000+ images while monkey passively viewing the fixation to get the reward (a passive viewing task).

The presenting speed of images is too fast (sometimes 100-ms onset and 100-ms offset), and I need to monitor the eye even when images offset, so I can't put one image into one trial(Due to a minium ITI longer than 100ms in my computer). Combining ImageChanger and RewardScheduler(with LooseHold modified by myself to update fixation judgement in real time) by a Concurrent adapter solves the problem, but it take too much time to creat scene and save trial record (if I put 500+ images to ImageChanger, it would take 20+ second to save 'one trial').

Is there any way to improve the current method? Can I provide ImageChanger with TaskObject (instead of providing image filename) so that I can pre-load images when loading condition file? Thanks for your help!

Thanks for the help!
Moon Li

Unless you turned on the [Save stimuli] option, saving a trial should not take that long. The delay must be the image loading time mostly.

The option of preloading images is already there. You just need to learn some MGL functions, as well as how to use the userloop function. See the attached example. Update your NIMH ML first before running the example, since I modified ImageChanger slightly to present multiple pre-loaded images.

By the way, this method circumvents NIMH ML's object tracking, so the resulting data file will not be replayed correctly with mlplayer.

Thank you for your generous help, Jaewon! I'm using userloop function now so that everything seems to running correctly