Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
I wasn't planning on submitting it, I was just wondering if the current publication file can be replaced with this one so it is more representative of reality. And if so what is the procedure for this?
I see the 'resyncing movies officially' thread, should I post it there? That looked like a discussion thread to me.
I'm not looking to get any credit or anything, I'm just wondering about the policy.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
http://tasvideos.org/userfiles/info/21288338289808557
Ok here is a faithful re-sync of flagitous' run. It plays each level exactly as he did. It is about 100 frames longer then his run. about 30 frames is lag, and about 70 is loading time.
So my question now is what should I do with this? Is there a procedure to replace older runs with ones on more accurate emulators? 1.5 seconds seems like a lot over a 1:46 run, it would be nice to have a more accurate representation.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
What is changing when those lag frames are being added in?
In my above example, if the machine state is identical at the start of the 12 frame sequence up to the jump, what is changing in instruction execution when the lag frame is added in the newer emulator? Can it change the final outcome?
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
Hi,
I have a question about working with older runs made on older emulators.
I am wondering if it is always possible to achieve the same state an older run achieves on a newer emulator if you can correctly account for added in (or I guess in rare cases subtracted) lag frames in the right locations in the movie.
For example, if an old movie has the player running right for 12 frames before pressing A to jump, and the new movie on a new emulator adds a lag frame within those 12 frames, making the movie file 13 frames before the jump, will the resulting state at the jump always be identical? By state I mean RAM or even instruction level execution.
I ask this as I am having trouble syncing some parts of flagitious' Umihara Kawase run made on snes 9x improvement 9 to bizhawk 1.9.1. I have tried repeatedly, so I am wondering if it should at least be possible in principle so I know how much effort to put into it.
But I guess it's also a question I am interested in in general, since I have only limited technical knowledge about emulation.
Thanks for any replies!
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
http://tasvideos.org/userfiles/info/21207921798009862
closing in on a final version. right around 13 minutes now. A couple route changes led to significant savings.
ars will be working on adding the entertainment value back into the boss fights.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
http://tasvideos.org/userfiles/info/21137506145337278
10 seconds faster then the published run. My guess is that ~15 seconds of savings is possible.
I'm not really familiar with, or much interested in, the specifics regarding what versions count as different runs or not. As of right now though I do believe the run can be ported to DS while avoiding the peculiar case in the any% run, so the option is there at least if nothing else.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
Oh I put the total time instead of the level time.
It's still possible to do tas time for DS since the physics are faithfully enough ported, one can just compare every level. Well let's see.
For now here is a file with ~ 5 seconds improvement over the published run, it's a start
http://tasvideos.org/userfiles/info/21071311858542402
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
http://tasvideos.org/userfiles/info/21056573281904883
Here is a rough draft complete re-sync. It was not simple to do. There seem to be some peculiarities in lag and loading times that make it not easy to predict.
I will be posting a comparison of level times to get an actual comparison and see just how much extra lag there is.
Level - orig / new / diff
00 - 671 / 553 / (-118)
01 - 784 / 690 / (-94)
02 - 633 / 555 / (-78)
03 - 928 / 923 / (-5)
04 - 559 / 580 / (21)
05 - 479 / 487 / (8)
06 - 449 / 450 / (1)
07 - 544 / 556 / (12)
08 - 7734 / 7882 / (148) (Boss)
14 - 801 / 807 / (6)
15 - 963 / 950 / (-13)
16 - 921 / 920 / (-1)
17 - 793 / 805 / (12)
30 - 649 / 599 / (-50)
33 - 1007 / 1008 / (1)
37 - 1262 / 1228 / (-34)
38 - 1076 / 1222 / (146)
41 - 1061 / 1038 / (-23)
42 - 649 / 657 / (8)
26 - 957 / 956 / (-1)
31 - 1167 / 1170 / (3)
46 - 997 / 984 / (-13)
47 - 1319 / 1248 / (-71)
48 - 935 / 934 / (-1)
56 - 13422 / 13488 / (66) (Boss)
57 - 47707 / 47705 [end]
All that adds up to about -70 frames. So considering both runs end at about the same frame, overall ~70 extra lag/loading frames is not so bad.
Based on the run so far, there is nothing preventing the run from being ported to the DS and saving a lot of loading time, so we might yet find ourselves back on the DS.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
AndiHoffi wrote:
I am wondering about a detail in the medusa-fight: AFAIK, you shoot as quickly as possible. Shouldn't it help the to shoot the first shot from a longer distance and the last one from the shortest distance possible?
Since the first shot can be released sooner and the last shot takes shorter to reach the target, this should gain some time.
I vaguely remember testing this and I'm pretty sure she takes damage on the first possible frame for every hit, so shooting earlier would have no impact.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
jlun2 wrote:
What's that passcode at the end?
That is most likely the prize code for a second competition that was planned but never happened.
An interesting side note. There are unused Pizza Hut sprites in the game data that show up in the pallets for the carnival world and the game intro screen, so possibly they were originally a contest sponsor.
https://code.google.com/p/treasure-master-hack/wiki/Bonus_world_2_code_analysis
Here is the machine code if anyone is interested. The difficulty as micro500 noted is that not all of it is necessarily exactly as originally programmed. I put the level end trigger in a seemingly likely but still aribtrary spot. Also the code that resets the spring flag can also be changed and still work.
In some of his replies on twitter, Steven Ruddy originally (a couple years ago) mentioned that he might still have the passcode that loads bonus world 2 in his archives, but never responded with a firm yes or no if he found it or not. So that is another loose end.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
I managed to figure out the working code for the second Bonus world.
Here is a savestate:
https://www.dropbox.com/s/7j3de93jsb8al1s/Treasure%20Master%20%28U%29%20HACKED_latest.zip?dl=0
The savestate works on a completely unmodified rom. Basically after I worked out what the code should be, micro500 was able to make it pass the internal checksum after the last step to make it run just as if it was input from the beginning.
However, finding out the pass code that let's you play the level through a normal play through of the game is quite another matter. Micro500 will be working on working backwards from the end code to what the original input was, but there is no garauntee it is possible, and a brute force search might still be needed.
In the mean time, enjoy playing this challenging level unseen for 20 years!
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
I have a minor bug to report.
I am using bizhawk with tas studio, and I notice that when I try to copy and paste from one .bk2 file to another I get 2 extra frames, and the first extra frame has an up in it. Actually it doesn't always have an up, it seems random.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
Well I ended up back here afterall.
here is the complete resynced any % run. The first level is still off but the rest is very close.
http://tasvideos.org/userfiles/info/20648132473254955
It's nearly 200 frames longer then the original, and ~150 of that is lag.
It really is too bad one small difference on the DS keeps it from going there. Otherwise it would be about 10 seconds faster!
It's still possible the sightseeing run can go to the DS, if the particular instance of that enemy capture is not used. Porting the run is not too hard, all you have to do is delete the lag frames to sync it up. Its a bit easier to tas on bizhawk anyway.
Oh and Samsara and ars4326 will be teaming up with me for these runs. Team Mission InTASible will have a submittable run yet!
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
Samsara wrote:
I'd like to keep looking into something brought up in the publication thread:
deg222 wrote:
2) Assembling the Iron Sword skip - there's a way to skip the message half way through the song... although I've never been able to do it consistently and could never do it on my real runs. Maybe a button needs to be pressed on a certain frame? Anyways, there's an alternate consistent way of doing it, but you need to have a spell with you, so that's not relevant here.
deg222 wrote:
But as far as the Iron Sword skip, I felt like holding a direction before hand made a difference? I'm not sure, it never felt consistent.
This would be a solid 3-4 seconds of improvement if it's possible to get without the use of a spell, but it might still be an improvement regardless. The fact that holding a direction apparently makes a difference makes me believe that it's RNG-based, though given my limited knowledge of the game I wouldn't know where to start with that. I tried brute-forcing it by just holding up for varying lengths of time, but wasn't able to get anywhere with it.
How exactly does the spell system work? If there's another easily obtainable key along the route, could you figuratively pick up Fleet Foot, use it somewhere to save time, then use it further to open the magic menu and allow for the song skip? Or does the spell actually disappear and leave you with no magic?
The fleetfoot spell is a one time use spell. Once you use it it disappears. It lasts as long as you remain in the same area (don't enter any doors)
About this song skip, I have no idea. This game as several frame rules in place in various areas. It might be associated with one of those, in which case it wouldn't be RNG related at all, but truly I have no idea.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
There seems to be a minor difference in the DS version that might make it fatal to the run.
IN 0.56 of the published run, the fish is critical to the route with the speed boost you get from it. But in the DS version this fish is captured 2 frames early, robbing you of the critical speed you need to make the route work. I'll do a bit more testing, but I did a frame by frame comparison with the snes, and it is exactly the same up until the fish gets captured.
Might be back to the snes afterall!
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
WIP of first 2 levels. It's going slower then I thought, but I'm skill learning the system so hopefully will speed up going forward.
Level 2 was tough, the published any % seemed to rely on a speed boost out of nowhere to line everything up, I couldn't recreate it.
http://tasvideos.org/userfiles/info/20607371199308230
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
Since the current any % run syncs almost to the frame (aside from lag) I will be developing a new run for the DS version.
My plan is still the same.
1. resync any%
2. splice in sightseeing
3. optimize level by level
anyone interested in teaming up or have any advice on how to convert bk2 file inputs to dsm so I don't have to do it all by hand feel free to pm me.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
KennyMan666 wrote:
Y'know, if someone is going to make a new Umihara Kawase TAS, it might be an idea to switch to the DS port. It has a slightly smaller viewing area, but much faster loading times and less lag (and a better timer!).
Hmmm, are the physics faithfully ported? The TAS relies on some tricks that I could easily imagine not making over to the DS.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
http://tasvideos.org/userfiles/info/20562643658346167
Ok, so I am slowly trying to re-sync the umihara runs in bizhawk 1.9.1.
It hasn't been easy, so far I only have level 1, and as you can see it's a bit slower then the published run, almost certainly due to lag differences.
My plan here is:
1. re-sync the any% run
2. branch off of it with the remaining sightseeing run levels.
3. re-optimize one level at a time.
It seems RNG is not at issue here, so picking and choosing levels seems like the way to go.
If anyone wants to help out, I'd gladly team up. Once a complete working file is done, progress should be much better.
Also if anyone who actually knows what they are doing in bizhawk can offer any tips, I'd appreciate that too.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
Nach wrote:
The YouTube embeds in this thread can now be played back in 60 FPS on supported browsers.
Regarding DTC6, how does starting in March sound?
How do you guys feel about using lsnes for DTC6?
I think I would wait a few more months before taking another beating : )
I don't know anything about lsnes but could learn it.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
I was completely unaware something like this existed, pretty cool!
2^64 is a daunting space to check. Even if only 0.00001% gave you something that looked like valid machine code, just checking those cases to see which actually work is still daunting.
What is the machine code the first prize code generates? Can it be used as a guide to reason out what the resulting other machine code would be, for example the last instruction might be returning to a specific place in the rom?
A very interesting challenge, I will definitely be keeping track of it, good luck!
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
Randil wrote:
Great job! You work very fast but manage to keep a good quality throughout the run. And thanks Samsara for the encode.
Some thoughts that popped up in my head while watching this (I will refer to the timings in the youtube video since I watched the encoded video). Some of this is nitpicking, but I thought I'd raise the points anyway:
*I think you can make a wall zip at the end of 1-3, zipping through the block on the left and wrapping to the right side of the screen and reaching the door a little bit faster.
*Maybe you can zip through the wall on the left on 2-1 at around 06:54 in the youtube video? I'm thinking that you turn around at this point, zip through the wall and pop out on the far right side of the screen. Might save a few frames maybe? I seem to remember testing this a while back, but I don't remember exactly...
*I think you can zip through the left wall at the store at 12:26, saving a few frames. At 18:13 you zip through the wall at the store (with the same layout), which is a little bit inconsistent.
*Same thought on the room around 12:31, I think you can zip through the left wall here too.
*13:47, aah, the platform pause trick :) It still makes me smile!
*This is a crazy idea, but at 17:52, would it be faster to drop down the pit in the middle of the room, zip through the wall here and pop out right on the door? Just a thought.
*Maybe a possible wall zip around 17:56? If you drop down to the lava and zip through the blocks to your left to reach the door.
Anyway, I will vote yes since this is a very nice improvement to the published run (and most of my points are nitpickings that don't account for much potential time lost). Awesome job, I think you are becoming a very competent TASer :) Maybe Little Nemo next? I think SDA has found some nice wall zipping glitches for that game that are absent in the published TAS.
I had tested some of these, as well as other little time savers. Unfortunately they seem to throw off 3-4 quite badly.
The 2-4 zips I thought would save time but things didn't work out. The swords in the later rooms end up in a different pattern and I throw away the time saved waiting for them since I can't take any additional hits.
Thanks for your comments, I think a V2 of this TAS might be possible, but first someone would need to figure out exactly how the level 3 boss works.
And next will be umihara kawase, expanding into snes should be fun.
Editor, Experienced Forum User, Published Author, Expert player
(3536)
Joined: 11/30/2014
Posts: 2733
Location: US
Hurray top 3!
Congrats to team 1, impressive.
I would like to thank Samsara and ars4326 for giving me the opportunity to team up with them in this contest, it was fun and I think our team was quite strong.
Thanks for hosting this Mothrayas. Game choice was good and I think the length of time was right for the game length.
