Hey guys. I created this little comparison video to analyze our runs. You can see it here.
Then I noticed, that I should probably add the Real Time to my run, and I am not sure how it's properly done, which are the correct starting and ending frames.
I discussed my steps in the video description, can anybody help me get the real time put on the leaderboards?
I started time when the menu disappeared and ended when the final time showed up on screen.
I'm noticing that your screen fades quicker than mine does, making your real time much faster. In level 2, we grab the heart on the same frame, but you exit the level 3 frames ahead.
Right! That is very interesting. I wonder what happens if we start a deeper investigation into this, including other runs. The real times might depend on some loading differences, but I'd bet that the game "removes" the loading times.
What could end up happening is that real time will be rendered useless?
Anyways: You use the same timing method as me, but I start one frame earlier, which makes my time a 2:34.000 flat. @koma how can I get that time edited in on the leaderboards?
Edit: Or we might have to remove the loading frames in future, as we do it in Turrican 2, which is a lot of work to determine the real time. I'm too lazy to do that right now but it would be damn interesting.
Okay, I manually retimed both my run and @xatrixx 's run.
I added up the final time of each of the four levels. I used the first frame the screen fully fades in as a starting frame, and the first frame of the screen fading out as the stopping point. I did not include the transitions between levels because that would be extra work to retime, and it's unlikely that I saved or lost time to xatrixx in those places. Here are the final results:
First, my run:
And xatrixx's run:
It turns out that my run was about 0.3 seconds faster than xatrixx when the fades are removed. My first two levels were a bit faster, my third level was slightly slower, and my final level was faster. I only saved 0.17 in level 4. I got a 5 hit demon kill, but lost a bit of time to a slow bridge skip. I don't know how I lost time in level 3, but it's probably from the bat boss.
Wow that is even more interesting! Thanks so much for the work. This probably means that we are at a point of having to gain < 0.7 seconds to get a 2:26. I still wonder why the fade times differ, but I don't think it will ultimately matter.
Any chance you have cut together a video? I'd be super interested in the x/y positions on a frame per frame basis, to do in-depth ananlyis on how to improve little by little.
Falling onto stairs seems to be slightly faster than walking onto them, which is why I saved time by jumping off the platform in level 1 later.
It's possible to quick kill the second bat by just hitting it whenever you can and reacting to it's movements. There is no easy setup like in level 1, butt hat would save 0.3. My third level and final levels have slow movement. Together, a 2:26 is possible.
My final boss loses at least .3 to my best I'm sure. So 2nd bat plus that and there's your 2:26
So since there is a loadless and a load time difference, should we remove the real time from the leaderboards anyways? It would be quite some work to maintain a loadless time, and I think for now an in game time is sufficient. However, we could add loadless time instead of real time for the top 3 runs at least.
That being said, real time as it is does not really reflect the performance of the run, so it has to go.
