This video shows a Prince of Persia (v.1.0, PC) level 8 speedrun I made today (2010-08-05). The time was 1:39.64, which is a new unofficial world record (as far as I know). At the same time, I improved my unofficial world record in individual natural segment total time (with minimal starting healths) to 13:42.73.
I used the level 8 start savegame from my "Prince of Persia v.1.0 in 13:45" natural multisegment speedrun. When I reached the level 9, I saved the game normally by pressing ctrl-g once, at the earliest frame possible. The time spent on level 8 was regarded as the difference between the times in the two savegame files.
The game was run in DOSBox 0.73 (core=normal, cycles=6000, sbtype=sb16). No cheating or tool-assistance was used. The Sound Blaster was configured to be Sound Blaster 16 to make it possible to disable the music without disabling the sounds. If I hadn't disabled the music, I would have had to pause the game for the duration of the music when leaving level 8 (not to waste time ofcourse). DOSBox was also used to record the run. I uploaded the original video clip on YouTube.
THE STORY BEHIND THIS RUN:
My old level 8 record run, used as one segment in the multisegment speedrun "Prince of Persia v.1.0 in 13:45", was 24 frames (2.00 s) slower. I didn't realize the slowness of the level 8 starting strategy (learned from SprintGod) until March 2009, when MoscowTracer suggested jumping by the first guard. I had already tried that when I was making the speedrun, but that strategy seemed slower to me as you have to spend more time pushing the guard right. Unfortunately, I didn't actually measure the time spent. When I measured the time spent in an unoptimised variation of that starting strategy, to my surprise, I found out that it took about the same time as my old strategy. I then optimised the jump by -strategy, and managed to save 13 frames of time.
In mid July 2010 I finally started to work on this run. I decided to optimise the whole old run, instead of just doing the beginning faster. Finally on July 27th I had found a startegy that saves 24 frames of time in total, and had made a run that saved 23 frames of time - I had wasted the 1 frame by saving the game 1 frame too late at the start of level 9 -.-. I then reached the same time at least 6 more times (messing up at various points or just having the orange guard (I named him Annoying Orange) take a hit instead of parrying) until I finally managed to make this run.
TIMES:
- time left at lvl 8 start: 53:42.98 (515)
- time left at lvl 9 start (new): 52:03.34 (40)
- time left at lvl 9 start (old): 52:01.34 (16)
- time spent between lvl 8 start and lvl 9 start (new): 1:39.64 (475)
- time spent between lvl 8 start and lvl 9 start (old): 1:41.64 (499)
- coherent natural multisegment speedrun to level 9 (new): 7:56.66 (679)
- the new individual segment total time (with minimal starting healths): 13:42.73 (512)
- the (old) coherent natural multisegment speedrun (whole game): 13:44.73 (536)
NATURAL SEGMENTS
By 'natural segment' I mean the smallest segments that the saving system of the game divides the whole game (from the beginning to the end) into. The 1st natural segment is from the game start to the saving moment on level 3 (can't save on level 2), 2nd natural segment is from the level 3 start to the saving moment on level 4 and so on. The last (12th) natural segment is from the level 12b start to the end of the game (time does not run while Jaffar is dead, but you have to survive to the level 13 and then save the princess).
By coherent natural multisegment speedrun I mean that you start the next natural segment using the same savefile (by contents) as you created when ending the earlier natural segment.
HOW THE TIMES WERE COUNTED:
The moment when time has expired was regarded 00:00.00 (not -1 frame). One minute was assumed to be 719 frames. The time left at the beginning of each session (except for the first one) was read from the first 4 bytes of a savegame file. The first byte was assumed to tell the minutes left rounded up, so the byte should be zero only when the time has expired. The third and fourth byte were assumed to represent a 16 bit integer in little endian format. The integer was assumed to tell how many frames of time was left in addition to (the value of the first byte minus one) full minutes. So if there would have been exactly n minutes left, the integer would have been 719 (not 0).
1/60 min (=719/60 frames) was regarded as one second. The time left before the first frame of the game was assumed to be exactly 60:00:00 (could be different in real?). The time left after Jaffar's death was read from the savegame file that was created just before entering level 13.
SEE ALSO:
Prince of Persia v.1.0 in 13:45 (1 / 2): http://www.youtube.com/watch?v=zYF3OMJeWms
Prince of Persia v.1.0 in 13:45 (2 / 2): http://www.youtube.com/watch?v=FJRmrFzwxPM
Loading...
1:01
Indy beats sober Biff in Last Crusadeby Ma5terViking6,597 views
cool!!!!!
kevin740329 4 months ago
wtf?!!!!!!!!!!!!!!!!!!!!!
MegaIdontfuckingcare 11 months ago
Cool!
gertetu 1 year ago