bismuth9
patreon
Production update 2020-07-30 and a release date!
🕑 Added 2020-07-30 15:00:21 +0000 UTCHello!
First of all, apologies for the late update. I kept putting it off for some reason, but at least I have some good news to compensate!
So, I've been doing some pretty good work on the video, but I've also been kept busy by other stuff (mostly the apartment, again). The next week's gonna be especially intense for me, as we're repainting the apartment, I'm finishing this video, and I'm editing a lot of content at once for Kosmic! My girlfriend took a week off work, too, so I might sprinkle in some vacation time with her as well.
So, I guess I should throw the release date at you already: Sunday, August 9th! I'm staying faithful to what I said about an early August release, but delaying it a few days until my girlfriend's back to work. Again, there will be a viewing party on my Twitch stream, and I will publicly announce the release a few days in advance. I'm thinking of trying something a bit different. I noticed that the peak time for YouTube wasn't the evenings, but actually the middle of the afternoon. Even though most of my audience is from North America, a good portion is European, so I think combining a decent time in NA and a great time in Europe is overall better than a great time in NA and an awful time in Europe.
For the video itself, I've been hard at work when I could. A lot of work didn't immediately translate into edited video, but since last update, I edited 10 more minutes, so I'm now at 26:30 done. As for the final length of the video, that's still up in the air, but my best guess is about 33 minutes.
What I mean by work not translating into edited video is that I ran into some issues where things weren't as I thought they'd be, and I had to understand why, then rework that part of the script to incorporate this new understanding. This happened twice this week, but the most notable example was in 8-Fort, on the conveyor at the end.
Mitch and Tompa (my two main helpers) disagreed on small details about Mitch's strategy approaching this part. Tompa said that because Mitch arrived a bit late to the final room, his left-right movement to gather speed to slide under the spikes ended up losing him time. Tompa also said that the jump Mitch did also cost him time, because he lost the additional distance from the conveyor belt pushing him forwards, and that there's virtually no scenario where jumping there is better than simply staying on the ground. Mitch, however, was on the fence about whether or not it would have been better to simply hold right in this case, but he was adamant that his jump did not lose additional time.
It turns out, they were both half-right about the situation. I did some meticulous testing and found out that yes, indeed, Mitch arrived a bit late to the final room and he would have been better off simply holding right (by about a handful of frames, assuming a crouch around the same place he did in his run). However, in his left-right movement, jumping actually SAVED time compared to not jumping.
This result baffled me, as I could understand the tradeoff between distance from the conveyor belt's push and speed from more time spent accelerating, but my calculations pointed towards the extra speed being worth only about 3/4 of the lost distance, in the end causing a very marginal loss. However, after painstakingly crunching the numbers, I realized that accelerating for longer made the extra speed count double, because by the time Mario starts losing speed, the other strategy already has been slowing down for a while, so the speed difference is both the extra acceleration and the deceleration between the two different points where Mario ducks. It's hard to grasp, but after I realized that this is related to how the braking distance is a factor of the square of the initial speed, it all clicked in my head.
Anyway, this is the sort of thing that slowed down production, but is going to be worth it in the end, because the final product will be more polished. What came out of this is a cool visualization of the 3 different approaches and how they would have fared in this specific situation with the P-switch timer (below is a still frame of the result).
Another major thing I figured out is how to properly use Dotsarecool's method of importing game data into videos using After Effects. It took me quite some time to figure out the whole process and add it to my toolkit, but I got past the first learning curve and I feel confident enough in it to integrate it into my workflow quite nicely. If you're wondering how that works, check this amazing video out: https://www.youtube.com/watch?v=yuVx4QI6fIM
Alright, I think this update's long enough. I hope you are having a good week, and that everyone's eager for a release next week!
