Experienced Forum User, Published Author, Active player
(420)
Joined: 9/21/2009
Posts: 1047
Location: California
I recall that NintendoG8mer "proved" that Corona Mountain isn't able to be entered early, but can someone double check now that Dolphin emulates SMS better? I doubt it will change anything, but it would still be a good idea to find out.
1. Why did you use the console-verified encode?
2. Both runs don't start at the same time.
Overall, this "comparison" video is pretty useless. =/
I encourage you to re-do it, though!
^ all this, plus the final .m64 hasn't been submitted yet...
Experienced Forum User, Published Author, Active player
(420)
Joined: 9/21/2009
Posts: 1047
Location: California
UMAD TURTLE?
Edit: So after playing around some more, Mapler and I both understand treasure hunting physics a lot better now. We were able to kill the 2 robots by 2:04 and still beat the turtle with this "average" position cycle. After optimizing the end, this gave us a time of 13:85. Pretty awesome for only a 3 frame cutscene skip delay!
Experienced Forum User, Published Author, Active player
(420)
Joined: 9/21/2009
Posts: 1047
Location: California
Update: Just got 14:10 (not optimized) on what I thought was a "meh" (i.e. ~15 seconds) emerald position cycle. I'm not 100% sure I'll be able to get 13:xx, but I'm sure 14:10 is improvable. And yes, this is the same cycle I thought would be a very high 14 bordering 15. I managed to use a weird "turtle jumping" strategy that saved some time. I'm glad we were able to find a pretty good cycle given that the soft resets changed everything.
Edit: 14:00 =)
Experienced Forum User, Published Author, Active player
(420)
Joined: 9/21/2009
Posts: 1047
Location: California
CoolKirby wrote:
sonicpacker wrote:
Hopefully the desync issues with SMS will be solved soon as well.
That would help too. Before the input backshifting fix and nitsuja's save fixes, I remember MUGG saying how difficult (an understatement, I know) TASing this game was. It's still hard to avoid desyncs in this game even with the TAS-input revision, so hopefully some game-specific fixes will merge with this Sirena Beach fix and the other important changes to make a revision perfect for running this game.
Yeah, MUGG, Rene, and myself (probably others too) have all had issues with getting SMS to playback correctly since Dolphin TASing improved a little over a year ago. I wonder why SMS has so many more problems.
Experienced Forum User, Published Author, Active player
(420)
Joined: 9/21/2009
Posts: 1047
Location: California
MUGG wrote:
This is getting a little confusing so...
There's the official branch, a TASinput branch and a more savestates branch (which supposedly has a worse tasinput window?). I'm not sure I follow on all the details..
There is official. There is more save fixes. There is TAS Input. The more save fixes alone has a much worse TAS Input window. The TAS Input revs includes the save fixes. What I was saying is that bzb *might* have forgotten to include the save fixes ONLY for the 32bit TAS Input rev fix (I know for a fact the save fixes are in the rev I use). I'm not sure if he did forget or not, but you shouldn't be getting desyncs.
Experienced Forum User, Published Author, Active player
(420)
Joined: 9/21/2009
Posts: 1047
Location: California
MUGG wrote:
Actually, forget the DTM... I can probably redo it someday, maybe in the fullgame any% TAS or somewhere. I TASed another level today, and it desynced halfway through. I redid the part and it desynced again near the end.
Given this rate of desyncs I'm getting, I want to stop TASing Mario sunshine for another few months. The person saying 3.0-381 (TASinput) was 100% desync-proof is wrong with his statement, it seems.
How long were the TASes? Did you actually use the TAS input or just the rev it's included in? Maybe bzb forgot to include the more save fixes branch in his 32bit fix. I know for sure I haven't had even one desync in any game since the save fixes (including 2 times I tried SMS).
Experienced Forum User, Published Author, Active player
(420)
Joined: 9/21/2009
Posts: 1047
Location: California
jlun2 wrote:
sonicpacker wrote:
By the way, there used to be a page on .dtm documentation that I found shortly after making this list, but it seems to have disappeared from the Dolphin google-code page.
All I found is this incomplete documentation on .dtm
Yeah, the page used to look much different. It had a lot more info on button presses and such.
Experienced Forum User, Published Author, Active player
(420)
Joined: 9/21/2009
Posts: 1047
Location: California
I would try but I only have the U iso. About a year ago when Mapler and I couldn't find anything about .dtm's, we made this notepad file showing what button presses are in hex form. Maybe it can help you:
0000000080808080 - Blank
0100000080808080 - Start
0200000080808080 - A
0400000080808080 - B
0800000080808080 - X
1000000080808080 - Y
2000000080808080 - Z
4000000080808080 - Up DPad
0002000080808080 - Right DPad
8000000080808080 - Down DPad
0001000080808080 - Left DPad
0004FF0080808080 - 100% L
000800FF80808080 - 100% R
0000000080FF8080 - Up Analog
00000000FF808080 - Right Analog
0000000080018080 - Down Analog
0000000001808080 - Left Analog
00000000808080FF - Up C
000000008080FF80 - Right C
0000000080808001 - Down C
0000000080800180 - Left C
By the way, there used to be a page on .dtm documentation that I found shortly after making this list, but it seems to have disappeared from the Dolphin google-code page.
