Posts for CasualPokePlayer


1 2
9 10 11
27 28
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Patashu wrote:
Dare I ask... Why *is* this a softlock?
Looking at tracelogger it seems more to be just a crash (also ACE to boot, lol)
E982:  10 00     stop $00            A:e8 F:10 B:21 C:d0 D:dc E:d1 H:e8 L:e8 LY:91 SP:dfee  Cy:33270232
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Already open issues on it, it's not really resolvable without UI to change sync settings without the core loaded (which is an open PR). You can manually change settings in the config file in the meantime, or just give it all the firmware it's requesting.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
RGL wrote:
There seems to be a menu bug. You get stuck in the Menu for challenges and the game wont react anymore. This thread on the MupenPlus Git exactly describes that problem: https://github.com/mupen64plus/mupen64plus-user-issues/issues/586 Something about a Counter Problem? Hope that helps. Here is a TASproj to recreate that problem. Just load the save Data and go into the Challange / Race menu. https://drive.google.com/file/d/1luGFeoLQhWt56EN7AEtHMDLxSf_N4NQq/view?usp=sharing
CountPerOp refers to a hacky number that is forced to be the amount of cpu cycles every instruction takes. This is not at all accurate to console, but mupen has bad timings everywhere so this can help some games. Our version of mupen (April 28, 2013) predates this hack option (added on December 13, 2013). Does the same issue occur with the Ares64 core?
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
RadRedRooster wrote:
CasualPokePlayer wrote:
Looks like the ROM is just bad. The headerless hash should be MD5:6288936a2c92b11096cb4f825f4d5070 / SHA1:a4346a6d83a148933879f7a6a2cf67c1979704ee according to no-intro: https://datomatic.no-intro.org/index.php?page=show_record&s=45&n=1737
Thank your very much! In the future, I'll compare any ROM not working with information found on DAT-o-MATIC first.
Checking with a good ROM, there does seem to be another issue which causes a green screen. The iNES header reports MAPPER010 (MMC4), but Punch Out uses MAPPER009 (MMC2). Simply changing byte 0x6 of the ROM header from 0xA0 to 0x90 will correct this issue.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Looks like the ROM is just bad. The headerless hash should be MD5:6288936a2c92b11096cb4f825f4d5070 / SHA1:a4346a6d83a148933879f7a6a2cf67c1979704ee according to no-intro: https://datomatic.no-intro.org/index.php?page=show_record&s=45&n=1737
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
What does ROM analysis say (click the checkmark/question mark icon in bottom left corner)
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
DeathKontrol wrote:
The final time is 150153 frames, which is 41:53 using Wikipedia's listed GBC framerate of 59.7275hz.
Note that 59.7275hz is not the framerate per se, but rather the nominal framerate. The GB/C runs at that rate when the LCD is enabled, however, the LCD can be disabled, causing it and the PPU to go to sleep until it is re-enabled. When disabled, all concept of a framerate is lost, so the GB/C has effectively a variable framerate. The true time would be derived from the cycle count tracked by the emulator, likely making the true final time a few seconds less. Also if the movie doesn't have a cycle count in its header, the judge will fix it, so don't worry if that occurs.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Bag of Magic Food wrote:
Will the password generators page make a comeback, or was it decided that it's outside of TASVideos's scope and needs to be hosted elsewhere? I did find a page of similar generators on cah4e3's site, and there is also now a GitHub page of source code for generators based on Bisqwit's YouTube series. It would be neat to learn how to put together more such generator forms as more games are cracked...
https://github.com/TASVideos/tasvideos/issues/878#issuecomment-1008175032
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Alyosha wrote:
CasualPokePlayer wrote:
Going through this movie, it seems to be very well optimized, and a solid improvement. There is one minor odd thing however. You can clear the first screen of the title sequence 20 frames sooner. Doing this in the movie through causes it to desync from general RNG difference. This seems like a rather long wait for some sort of RNG manipulation, so I'm wondering what was exactly aimed here, or if this was actually just a mistake.
It's to control the spawn point of the third pearl in level 2 while still getting good mushroom RNG. You can't effect the spawn location of the pearl after the level starts, so you have to set everything up ahead of time. I did this by varying the start presses (which controls the pearl spawn), and then seeing if I could get good mushroom RNG by varying movement right at the start of level 2. The big delay at the start is the best I could get so that everything works.
Thanks!
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Going through this movie, it seems to be very well optimized, and a solid improvement. There is one minor odd thing however. You can clear the first screen of the title sequence 20 frames sooner. Doing this in the movie through causes it to desync from general RNG difference. This seems like a rather long wait for some sort of RNG manipulation, so I'm wondering what was exactly aimed here, or if this was actually just a mistake.
Post subject: Mupen64 is accepted for submissions again!
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Years ago, Mupen64-rr movies were disallowed from submissions, due to unreliable video dumping. Years later, A fork has been made, which among other things, fixed various issues with video dumping, with another fork by us further fixing video dumping issues. With these fixes, Mupen64 movies are now re-accepted for submissions! More info can be found in this thread: https://tasvideos.org/Forum/Topics/23131
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
The ini file does not contain configuration data, rather it acts as a gamedb (containing rom hashes and sometimes save type overrides). By default it is compressed (although it can be uncompressed in settings). The .cfg file stores configuration.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Go to the forked release for the prepackaged plugins for encoding, not the upstream release.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
phoenix1291 wrote:
This fork?
feos wrote:
we decided to host a fork
The link for plugins on that page is "If you want everything to run the emulator, see this http://repack.skazzy3.com/" and don't work.
Yes that fork, and that repack page is not linked on that fork, it's only linked in upstream's releases. Our fork currently has a single release, and release contains all the needed plugins. Note too the angrylion's plugins were modified by us to implement some functions needed for dumping with mupen. Just use the package released by us for encoding by our standards and it will have all the plugins prepackaged and ready for use.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Mupen Plus plugins are not compatible with Mupen, as they use different plugin specs (Mupen uses the old zilmar spec that Project64 also uses, Mupen Plus has its own spec). The Mupen-rr release in our linked fork includes cxd4 rsp, and can also be found in our fork of cxd4 rsp https://github.com/TASEmulators/mupen64plus-rsp-cxd4/releases/tag/r682
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Just export it to bk2 and play that bk2 back. If it still does that, upload the bk2 and post it in this thread. If it doesn't occur, then savestates somehow got fucked, which doesn't seem likely anyways given waterboxing.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
https://tasvideos.org/Movies-Saturn It's not a particularly popular console, so there aren't a ton of movies for the system.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
https://github.com/CasualPokePlayer/mupen64-rr-lua-/releases/tag/1.0.9-speex-fixed_audio_sync I have fixed audio sync too, this should address the audio desyncing issues (tl;dr I ripped out the old audio sync logic and put in new logic which should make more sense and actually sync to audio). EDIT: These fixes have been merged into upstream now!
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
https://github.com/CasualPokePlayer/mupen64-rr-lua-/releases/tag/1.0.9-speexv2 I've solved the popping issue. It seems the resampler was at fault here, so I hacked in speex's resampler, and the pops are gone with that in.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
lsnes is indeed still the only SNES TAS emu capable of subframe input. Although it wouldn't be "hard" to add that BizHawk really (the baseline functionality is in, it would just be hooking everything up). SGB emulation in lsnes is really only "good" with the BSNES+Gambatte combo. Which is essentially the same SNES emulation quality to the older BSNES core in BizHawk and the same GB emulation quality as BizHawk 2.2.1 and below Gambatte (before PSR took over BizHawk-Gambatte development / lacking in sinamas' 2019 updates to Gambatte). It's not terrible, decent I'd say, although it's unlikely to spit out many console verifications, besides for games that don't care initial timing is just wrong.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
The purpose of a verification movie is to verify that the SRAM state is indeed valid and reproducible, and is not something hacked together externally. The verification movie does not need to be optimal at all, it could be some casual playthrough with max speedup and whatnot, it just needs to produce SRAM the final movie uses.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
Very unlikely to work, considering how old the latest stable is. Odd that game doesn't work for you on 5.0, seems to work just fine for me.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
I suggest trying the latest stable release instead (5.0) as the latest dev builds have more desync problems. Also make note of the other things in that post and the post I linked.
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
What Dolphin version are you using? (latest dev or latest stable?) Do you have any problematic settings enabled? (e.g. Dual Core, which is not at all deterministic.) Do you have a portable.txt file in your Dolphin folder? (it is highly recommended to do so as to not mismatch configs between different versions, which can cause "magic" desyncs). Any non-default settings? Also see https://tasvideos.org/Forum/Topics/18330
Emulator Coder, Experienced Forum User, Judge, Published Author, Experienced player (609)
Joined: 2/26/2020
Posts: 698
Location: California
neo_omegon wrote:
If you move not diagonally in the tutorial, you can skip one tutorial message like this. -> https://youtu.be/6HvpAZmUs5M?t=267
I've actually learned this not too long after this run was published (along with other things I've missed). Of course I wouldn't really want to redo this run without a very good understanding of RNG and ways to manipulate that in a TAS (which will likely require extensive botting due to how RNG values get generated).
1 2
9 10 11
27 28