Update 10/10/23 - Latest submission in category had me retime the top 3 runs with VidTimer for better accuracy than my previous method. Used the consistent timing of 15 frames prior to the screen starting to blacken as a start time to avoid human error with starting the timer. Still used the frame of collecting the orb as an end time.

With this data, my run was shown in actuality to be a .083 rather than a .09.