As I was going back through some of the IL runs to figure out how to beat them, I noticed some problems with retiming IL runs. Not that they were retimed improperly, but that there are inherent problems with trying to retime IL runs that make it difficult or near impossible to do accurately. There are three main issues:

• Using the dust as the start frame
• YouTube dropping the frame rate
• The original recording’s frame rate The problem with using the appearance of the dust as the start time is that it appears when 0.033 is on the timer, giving an advantage to retimed runs. This problem is compounded by YouTube forcing shorter videos to be 30 fps. This means that the first frame the dust shows up in video could be when 0.050 is on the timer. The use of snapping into the door as the ending frame does correlate with the timer, but with the 30 fps from YouTube, a frame could be lost if you’re unlucky. These problems could probably be circumvented with some effort, but the real issue is the original recording’s frame rate. I don’t know if this applies to Android as well, but for iPhone the recordings have really strange frame rates. Just looking at four clips I made when looking into all this, without any editing, these were their frame rates: 58.68, 59.47, 59.67, and 59.75. I don’t think there’s any reasonable way to get a perfectly accurate retime of an individual level when you combine the two frame rate issues. This isn’t really a problem unless it’s a time in first place, especially in the variable category. For one, there could be a time that is impossible to tie or beat because of the issues I outlined. And retiming essentially puts those runs into fixed timestep. I don’t really know what a good solution is, but I do think something should happen. Despite the rule requiring IL runs to include the in-game timer, I think that removing the runs, or anything to that effect, isn't an great solution. Hopefully I’m not just rehashing an issue that’s already been brought up.