I am having a peculiar issue with preloading audio and images. For my artificial grammar learning experiment, I am hosting many, many resources in a locker (3 images per trial, 3-4 audio files per trial in Blocks 1/2) using the .zip method, as suggested by the documentation.
The load time seems extremely variable, with load time in between items which can last up to a minute, which I definitely want to avoid. I’m concerned about if/how I can support 3 more blocks of this.
I am currently unable to replicate the problem. Do you still experience varying preloading issues?
Actually I am able to identify preloading issues, having to do with the fact that you create elements with invalid filenames. For example, the last line of your CSV file for the training block 2 is:
ABA,2,B2FB2,comp,huspfb_1.mp3,huspfb_2.mp3,-,-,B2_husppos.png,-,-,"He said HUSP . Now, what did that mean again? I think that means big. These moons definitely look pretty HUSP .","Ok, help me out here. Which moon is TRAGOT than the first moon?",MIDDLE,RIGHT,MIDDLE
So Audio2, Audio3, Image2 and Image3 are all -, which are invalid filenames. PennController cannot find files for the elements you create in your second Template command that reference those columns, and so it fails to preload them. When the corresponding trials run, they try again to preload the resources for 1min, but obviously fail, and then crash when it comes to inserting the elements that aren’t associated with valid files