Posts for Kirkq

1 2
19 20 21
31 32
Posted: 10/12/2009 6:11 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Bomberman Hero Movement optimization 1.0: - Dropping/Laying a bomb typically takes 0-2 frames. - If you change directions very sharply and it makes you slip, change direction partially for two frames and then change direction all the way. (The game only reads directional input every two frames.) - When jumping to grab a ledge, dropping a bomb at an appropriate time will typically let you grab the ledge about 5 frames faster. - When jumping, throwing a bomb will cause you to start falling faster. This is only rarely useful. (See end of Crysta Hole.) - Sometimes if Bomberman is pushing up against a wall/door waiting for it to open, he will start off walking slower than if you walk past the door on the first frame possible without running into it. (See Dark Prison if this is unclear.) - You can gain full momentum during a jump by dropping a bomb midair. There are two ways to do this that are only very slightly different. 1: "A" / / / / / Drop will turn a short jump into a full jump without holding A 2: "A" until you're doing a full jump, drop when it lets you. -Similarly to this note, make sure you are never in falling animation for the most part. -Coming out of the fan in Secret Room I dropped a bomb to regain momentum. Similarly at the beginning of Fog Route (I think?) I dropped a bomb to gain momentum because it starts out falling. -In Water Pool I jumped before falling down a hole to increase my falling speed. The end of Secret Room, dropping a bomb while walking uphill (only) is the fastest way to gain momentum. This is all I can think of at the moment without rewatching the whole thing. I hope it helps. If you have any questions about tricks I'm more than willing to help. I'd recommend you review my progress so far and see how I use various tricks to deal with certain situations.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 10/9/2009 9:12 PM
Post subject: Encode of Test Run
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Here's an encode. VBA records pretty fast so I don't mind providing updates. http://www.youtube.com/watch?v=ylkNdlGj7lA The enemy manipulation looks about as good as it gets. You can hardly notice any slowing down. In fact, disregarding any changes in routing, I'd say it looks very near flawless. (No visible improvements, though I'm sure a few frames could possibly be shaved through better overworld manipulation.) I also enjoyed the second mettool battle in the tutorials. Pure delicious optimized shooting. Good stuff Bag, you've definitely done a ton of research to get to this great execution.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 10/7/2009 3:25 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
I progressed a couple weeks ago. I'm at Nitros; I have to recreate a 3 pylon kill equal or better than my test run which is proving to be difficult. I may make it my goal to finish Nitros this weekend. I'll post a WIP when I've made some substantial progress probably. University leaves it hard for me to set aside a block of time to focus on TASing difficult tasks. This is a personal problem with my time management skills that I am looking to improve.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 10/5/2009 3:33 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Could you detail how much time was spent on tradeoffs? I may get around to watching this during the upcoming weekend. Congratulations on finishing a ~5 hour submission.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 10/4/2009 11:12 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
More than likely the categories for N64 Zelda runs will be as follows: OoT any% (RBA etc.) OoT "all dungeons, etc." (no RBA) OoT 100% (probably if someone decides to do it) MQ "all dungeons" MM any% MM 100% MAYBE a MQ 100%, but it's a stretch. MAYBE a MM minimal game time, but I doubt it. The "all dungeons" is a bit trivial at this point because MrGrunz has managed to clip most of the boss rooms, but that is for discussion elsewhere. From what I've heard MQ any% is essentially the same run as OoT any%, because Bottle Adventure works the same. You'll notice we only have one of Pokemon Red and Blue published at this site. This is because Red and Blue will have almost exactly the same run, just on a different version. We have a Pokemon Yellow run because it demonstrates a radically different glitch. If the OoT and MQ any% are essentially the same run on "different games", both will not be accepted. As proved by JNX's and PDot's runs, if you cannot match the current quality standard for a game. It simply isn't going to be accepted. It isn't about "Setting the record for Master Quest". It is about producing a run that meets the current quality standards which have been set by Swordless, AKA, Bloob, MrGrunz, and others. It is NOT about "having the best time." It is about demonstrating the best quality of a run. I have not viewed your runs, but I just want to add that Bloob's all dungeons run does not have the pause fix, so any time you "gained" on his run, should be translated accordingly. You are not beating him by 5 seconds every time he pauses. The run is simply 5 seconds faster due to a technicality. If you gain 5 seconds on a pause and lose 2 seconds on optimization, you aren't actually producing a better run than his. A publishable run should beat or match his quality everywhere applicable for it to be "faster" and publishable. While it is good that you are constantly trying to improve, it is not good that you are trying to find the fastest ways of getting your own personal published record. There are no shortcuts in TASing, you just have to be persistent until you can match the quality standards. Obviously Ocarina of Time is a very difficult game to TAS very well, as most members of this site can't even TAS to the quality standards set in place. So my suggestion is to keep practicing in OoT until you can match the standards, or find another game. None of this is meant to condemn your efforts, I'm just trying to give you the big picture of what I (and many others) tend to look for in a publishable run. If you want to make a run ignoring what I've listed, no one is going to stop you or condemn you. It just probably won't get published here. We don't publish records, we publish quality.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 10/4/2009 9:54 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Thoughts: 1: I highly doubt the BoB 100 coin star is anywhere near maxed right now. I would almost bet that you can start koopa the quick, grab 100 coins, and get to the top of the mountain before he does. 2: The star on the island in BoB could possibly be improved by using a method similar to the current Chain Chomp strategy and clipping the side of the island closest to the box. 3: Has anyone experimented much with BLJing with a metal cap through water? 4: Is it useful to use the BLJ momentum transfer into the lobby anywhere? Does this affect the route? You guys should be sure to keep accurate documentation of who came up with the strategy for each star.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 10/3/2009 9:42 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
andymac wrote:
You must be doing it wrong. I never got a single blister while attempting control stick spinning. The technique that I used was to put the control stick between the bases of my two middle fingers, and spun my hand wildly.
That still seems like you would develop a blister between your fingers. A kid I knew actually tore a half dollar sized piece of skin out of the middle of his hand attempting Mario Party spins. I guess that's why they started selling a glove.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 10/1/2009 7:58 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
LOL at boss fight. 14200, just so I'm not focusing on this all run. Does the y position you're at on the screen when you hit the wind ball barrier really influence how fast you can get by the camera freeze? The camera freezes as soon as it is hit right? So basically you either traverse y length further before hitting it and the camera scrolls y, or you're y less far and the camera has scrolled y less far. So nothing really makes it faster unless you can walk further in the nonlimiting direction? (In this case if you could gain some position in the x direction somehow.) 14800, care to explain why you stopped moving for a bit? I assume it was enemy manipulation, but why exactly was this the fastest method here? I applaud your Stom fight. The grinding at about 24000 looks very technically well done as does the run overall thus far. The improvements are very noticeable, and the grinding sections are much more watchable up to this point.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/27/2009 10:42 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
I managed to formulate a successful Serpent kill tonight resulting in max speed and max-1 power (by complete chance). Now I just need to get into battle on the correct RNG, and we will be on to the Dragon. EDIT: *Sigh* It's going to take a ton of effort to pump the RNG 600 forward, at least we have the whole walk from Pandam to Padal to start. Really late reply, I missed the last post I apologize. In simple terms, once in any battle in a dungeon, the RNG only increases due to the attacks. The monster attack types are set for the entire fight based on the previous battle you were in. The RNG is basically a table of values at this point. A defend moves the RNG by 1, an attack moves the RNG by 3, Healing moves it by 48 or so, the enemy attacks move it by some set amount. You can only choose amongst very limited options to move between values in the table. This makes it entirely impossible to manipulate multiple misses in a row at will.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/25/2009 6:38 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Number of rerecords does not matter. We want good, fast videos, it does not matter how many times you reloaded/rerecorded. The ".smv" file records controller presses, and plays back controller presses. If your buttons are done from power on with no cheats used, they will be viewable and "validated" by us. You should use the Secret of Mana (U) [!] rom. Once you make a short video, upload the ".smv" file here: http://dehacked.2y.net/microstorage.php, and we can make sure it plays correctly for us. We validate by making sure the controller input plays back in normal circumstances. You may want to use these buttons to help see things: , = Display Input . = Display Frame Number Like funnyhair said, Open Read Only = WATCH movie Not Read Only = RECORD movie Use Shift / F1-F10 to save and reload Use Shift 8 to switch Open Read Only <--> Not Read Only Rules: MUST "Record from Reset" and "Clear SRAM" Warning: A good TAS for Secret of Mana will need to watch memory to manipulate random actions.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/15/2009 10:01 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Get the emulator version with lua support. There should be an option titled "Run lua script" or something, but I'm not familiar with Gens systems. Copy the "lua file" into notepad and save it as "Something.lua" Run the lua script. What scripts actually do is modify RAM values. In this scenario, modifying the camera apparently does not affect the game in the slightest (lag, etc.), which is why it is possible to apply this to a video without it desynchronizing.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/13/2009 2:24 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
For compression, the last step (using virtual dub) will be useful. If you want high resolution, the only way I know how to do it is by the site guidelines. If you're getting a sound desync check out the emulator specific settings. (It could also be your movie player causing the problem. Try VLC and/or MPlayer, both have issues sometimes.)
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/13/2009 1:42 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Snippets from Encoding Procedures: http://tasvideos.org/EmulatorResources/MakingAVI/Common/Windows.html ------- # FFDshow: Download and install. During installation, install all components. For "select supported video codecs", make sure to select H.264. The rest of the options should be left to the default. # Configuring FFDshow to capture: After installing, go to the start menu > ffdshow. Select VFW codec configuration. Under the encoding tab, select H.264 lossless. Under FOURCC, select H264. Other settings can be left on default. Select the Decoder tab now, and click Codecs in the left hand list. A new menu should now appear. Find H.264 in the list and make sure it is set to libavcodec by clicking on it in the decoder column. (This only has to be done the first time.) ---------- Create initial h.264 lossless capture. In the case of emulator specific settings, check the Emulator Settings - Windows article for help. 1. Run the emulator and load the ROM. 2. Increase the emulation speed to about 400% (the command to do this varies by emulator but often it is the '=' key). This will speed up this initial encode by making the limiting factor your cpu. 3. Turn off frame display. Configure sound (This varies by platform, but typically it is the highest setting in the emulator, 48khz stereo, or 44.1khz stereo. In the case of NES it is 48khz mono). Pause the emulator and load the movie file. 4. Select the Record AVI option. Select the mencoder folder to save the AVI (or whatever workspace folder you like). Call it capture.avi and save. Select FFDshow in the dropdown box. Click ok. 5. Unpause emulation. Record until the ending and one loop of the ending song if there is one that loops. Don't worry if you record a little extra afterward as you can clean it up with mencoder in the next step. (This varies slightly for each emulator.) -------- After you get this largely oversized Raw encode, I suggest opening the video with the program Virtualdub, trimming off any extra content with the home, end, and delete buttons. Direct Stream copy implies no compression, Full Processing mode implies compression. Audio: Compression: MPEG Layer-3 Video: Compression: x264 This should be good enough for youtube.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/13/2009 12:37 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Please run a movie while you're doing testing, so if you get a successful result we can study the conditions that led up to it.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/10/2009 9:52 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
If you're going to wait over 6 months before doing another revision, you're welcome to submit your good runs in the meantime. Even listing that there are better tricks available is fine as long as you did your best with the knowledge you had at the time. We're always looking for the best runs that have been played as long as they are played with the best knowledge at the time. (Many people find possible improvements during a run, but don't redo it at the time because the time saved is small for the work required.) If you're redoing soon, you may as well wait until you finish at this point, but I would've liked to have seen your run published back in November. I do respect your dedication though. =D
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/3/2009 8:59 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
I'm just saying if you want to compare to the quality of TAS that is probably going to be published, you need to have much more patience and pay closer attention to detail. It would take me at least an hour or two to TAS just that stage well, even after having TASed over 20 minutes of gameplay so far. It takes a lot of practice and patience to produce a good product, and that is something you need to learn if you want to improve your skills. I'm not saying this just because I'm the one doing this run. It is true in general.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/3/2009 6:17 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
1: 1600 You never have to slip when turning except rarely after landing from a jump. Change direction halfway for 2 frames and then change direction all the way. 2: 1700 Your second bomb is thrown with a low momentum jump. Also, you should most almost bomb things by laying a bomb since it wastes at most 2-3 frames. (Obviously not the first box, but definitely the second.) 3: 1830 Should have laid a bomb to kill the robot. 4: 2050 I feel like you should have been able to hit both monsters with the same bomb. (I realize you have a limited number of bombs, but you didn't really utilize them as well as you could have up until here.) 5: 2200 Altering the timing of the bomb drop can help you grab the ledge earlier. 6: 2250 Poor ledge grab as mentioned. 7: 2400 Never fall off a ledge, either jump or drop a bomb ASAP to preserve momentum. 8: 2550 If you had jumped up to the higher platform earlier, you could have turned around a lot faster with a ground turn around instead of an air turnaround. 9: 3150 Drop a bomb instead of throwing it to kill the robot. 10: 3600 I think dropping a bomb would've helped you get your momentum back sooner after climbing the ledge. 11: 3700 The chicken was handled poorly. I think the best method would be to drop a bomb the first time and then come back and drop 1-2 bombs on the way back. 12: 3850 The reason you bombed the thing under the platform separately from bombing the platform later is beyond me. 13: 4100 You slipped/skidded again. 14: 4200 Lay a bomb instead of throwing. 15: 4700,4800 you have low momentum during two jumps. You can drop a bomb on the first frame possible in the air to regain momentum, or you can run just a few frames longer. 16: 5200 I think if the bomb impacted the wall blocking the chickens it would explode on impact. Not sure if that is true with this type of wall or not. Actually, I would think you want to hit that wall as soon as possible, because killing all the chickens is the limiting factor to getting out of the level. So I think you should have thrown a bomb at it at around frame 5000. 17: You really had no reason to go down into the pit to kill the chickens. The pathing was pretty decent, the directions chosen for movement weren't bad though I think you could do better. You missed a lot of small time saving techniques that I have found and utilized, and you made a couple poor strategic choices (11,12,16,17) 300 rerecords is pretty lax for a minute of gameplay. While it wasn't bad, I'm pretty confident I could beat you by at least 10-15 seconds.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/3/2009 4:51 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
How feasible is an idea of actively reading memory and playing the game with a script designed to do actions at certain points in the game? This may have better results than trying to sync the entire keypress file.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/1/2009 3:48 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Dropping the name of a text file 5 feet would be negative work if you define the system correctly.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 9/1/2009 3:34 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Not changing the name in a text file?
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 8/30/2009 3:57 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
zvsp wrote:
If I record from power-on state, I must use Direct3D8. But if I do so, the graphic is so far from perfect. In either, the movie may be reject. Do you think this is allowed as a special case?
I will restate once again, if the only issue is the reset, get rid of the reset. The site currently has no method to analyze a movie starting from a save state in Mupen to my knowledge. Maybe this is beside the point or wrong, emphasis on my first statement.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 8/29/2009 12:01 AM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Turn off sync sound. This is most likely the problem.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 8/26/2009 7:50 AM
(Edited: 8/26/2009 9:45 AM)
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Reset recording still has some glitches, so I personally would recommend doing it without resets if they are causing problems. You should most always use the graphics plugin that best displays the game's graphics accurately. There isn't much that can be done to fix this at the moment to my knowledge. If your choice is between accurate graphics emulation and ability to record resets, I think it is more important that the graphics are accurate. To me a run that doesn't use resets is equally impressive to a run that does. Skipping cutscenes does not add anything impressive; it is just a faster run by technicality. If resets worked fine I'd say definitely use them, but since resets don't work right it doesn't seem worth the hassle or worth delaying the run. This site prefers the (U) version of games if there is no gameplay difference, but if you insist on using the (J) version it will probably still be published. I'm looking forward to your next update.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 8/21/2009 11:02 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Yeah, one falling flag is pretty amazing. I'm also impressed that HappyLee's contributions match the previous improvements by a team of three. Can someone double check HappyLee's 1-1 vs. the 1-1 that was supposedly 1 frame short and see if an improvement was found hiding somewhere? Personally, the jumping in fortress levels is somewhat interesting to me as it distorts the sound channels. The jumping in this run wasn't distracting to me, possibly because there was a lot going on with fireballs. What exactly happened in 8-4 with the shaded/dark ground? Was that an encoding issue or a graphical glitch? Yes vote for significant technical improvements to an already tight run.
Twitch Stream, Twitter, Youtube, TASVideos Profile
Posted: 8/20/2009 9:42 PM
Quote
Kirkq
Experienced Forum User, Published Author, Player (208)
Joined: 7/7/2006
Posts: 798
Location: US
Well done. A full second is a pretty fair gain for a 4th(?) version. The entertainment was sufficient. Yes vote.
Twitch Stream, Twitter, Youtube, TASVideos Profile
1 2
19 20 21
31 32