FOR BEST RESULTS, CLOSE UNNECESSARY TABS. Lag from other tabs, especially Scratch tabs, will seriously mess with the test. Adjust "Repetitions," which controls how many times the scene is rendered by each technique. Increasing repetitions increases lag but improves the precision of the timing measurements.
This is a speed test between: Method A: @ChromeCat_Test's "-Rex- tri fill improved" Method B: @KryptoScratcher's modified Azex 3D On my computer, Method B is consistently faster. Within the project are a bunch of other tri-fillers I tested. You can try swapping those into the "Method A" and "Method B" custom blocks. I've linked them all below. Pre-rendered geometry data from scene 2 of @TheGoodGuy2000's "good triangle speed test": https://scratch.mit.edu/projects/771305317/ which itself was captured from @ggenije's 3D game "Crystal Seeker": https://scratch.mit.edu/projects/463553665/ List of tri-fillers in the project: @KryptoScratcher's modified Azex 3D https://scratch.mit.edu/projects/885002848/ @tsf70's "Azex 3D Varient" https://scratch.mit.edu/projects/775301512/ 's modified Chrome_Cat tri-filler (unshared) 's "-Rex- tri fill improved" 's "Faster Tri Filler" 's "Faster triangle filler" 's "Azex 6.0 (more speed)" The speed tester was developed by me to compare the speed differences between scripts. Both visualizations use nonlinear coordinates to make the datapoints more visible. The scatterplot to the top right plots each trial as a point, where the two axes represent how long each method took. The line graph on the bottom graphs the time difference over the trial count.
