Joined: 7/26/2006
Posts: 1215
http://tas.monotoli.org/mov/oot/wip/lozoot-wip5-ds.mkv (10MB) Well I'm wondering why you didn't roll in the store. And while I like the idea of not changing camera angles between the garden signpost and the store shore... I'm not sure that was the fastest way.
Active player (441)
Joined: 3/21/2006
Posts: 940
Location: Toronto, Canada
SpiDeY wrote:
quick question from a guy who doesn't enter this topic since last year: why are we back to the beginning?
Well, it's a very good place to start...
My current project: Something mysterious (oooooh!) My username is all lower-case letters. Please get it right :(
Mitjitsu
He/Him
Banned User, Experienced player (532)
Joined: 4/24/2006
Posts: 2997
I still think the WIP Comicalflop showed earlier was faster.
Joined: 11/11/2006
Posts: 1235
Location: United Kingdom
Well it would help if there was still a link of the run so we can compare.
<adelikat> I am annoyed at my irc statements ending up in forums & sigs
Editor, Experienced player (730)
Joined: 6/13/2006
Posts: 3300
Location: Massachussetts, USA
Here's there be scary WIPs I'm pretty sure it is still linked a few pages back. Microstorage has this habit of not disappearing, unlike other upload hosting sites.
Homepage ☣ Retired
Joined: 7/26/2006
Posts: 1215
Dashiznawz on the left, Comicalflop on the right. (13MB) no fancy left-right audio stuff, just two different selectable audio tracks. Anyway, this video speaks for itself, comicalflop's is consistently faster in every area. Oh, DS's capped with jabo which has the ugly pause screens. CF capped with direct64. I never noticed the black bars were different from plugin to plugin.
Joined: 5/24/2004
Posts: 262
Nice, bkDj. Do you do all this stuff with straight mencoder or do you use some sort of frontend so you don't have to memorize all the mencoder switches? More on topic: Everyone working on OoT should keep up the good work.
Joined: 7/26/2006
Posts: 1215
Andypro wrote:
Do you do all this stuff with straight mencoder or do you use some sort of frontend so you don't have to memorize all the mencoder switches?
nah never took the time to get mencoder working, I just rehash old scripts. I record what I want from mupen using x264 at qp 18. then I use an avisynth script. the side-by-side looks like this. vid1 = DirectShowSource("C:\emulation\mupen64\tas\lol.avi", fps = 60) vid2 = DirectShowSource("C:\emulation\mupen64\tas\lol2.avi", fps = 60) StackHorizontal(vid1, vid2) LanczosResize(640,240).Fadein(60).Fadeout2(60) Then I just call x264.exe on it, one-pass usually with a constant rate factor (crf) of 26 (I have the cmd line saved in a text file so I don't have to memorize anything). I pull the wav from the avi and have lame turn it into mp3. mkvmerge does the rest. mencoder would probably be less steps if done right, but I have shortcuts to everything and so really it's like a 10-click process. Hope that helps.
Mitjitsu
He/Him
Banned User, Experienced player (532)
Joined: 4/24/2006
Posts: 2997
Thank you for uploading bk ;)
Joined: 5/17/2007
Posts: 124
Location: Cincinnati, OH
For the record I was never able to watch comicalflop's WIP it desyched as soon as he got control. After seeing it put up against mine I am embarrassed I even tried. I really don't belong on this site. I guess wanting something and trying your hardest at it really doesn't help if you are talentless.
Joined: 7/26/2006
Posts: 1215
DaShiznawz wrote:
For the record I was never able to watch comicalflop's WIP it desyched as soon as he got control. After seeing it put up against mine I am embarrassed I even tried. I really don't belong on this site. I guess wanting something and trying your hardest at it really doesn't help if you are talentless.
stop with the self-deprecation. now that you have a good example, try meeting or beating it. Doing this 3-minuteish game segment like that will provide a good foundation for the effort required for the following hour-and-change of gameplay. (I'm guessing there's a good 30 minutes of cutscenes or more)
Joined: 5/9/2005
Posts: 748
You should have seen comicalflops first TAS. Sure it was great in the end, but he restarted the game from mid way through (ish) about 9 times. His TAS skills are due more to dedication, than pure talent. Thats the point, TAS take gaming skills out of the equation. (Not sure if it sounds like I am but just in case... This is not a Bash against comicalflop in anyway.)
Joined: 5/24/2004
Posts: 262
bkDJ wrote:
Andypro wrote:
Do you do all this stuff with straight mencoder or do you use some sort of frontend so you don't have to memorize all the mencoder switches?
nah never took the time to get mencoder working, I just rehash old scripts. I record what I want from mupen using x264 at qp 18. then I use an avisynth script. the side-by-side looks like this. vid1 = DirectShowSource("C:\emulation\mupen64\tas\lol.avi", fps = 60) vid2 = DirectShowSource("C:\emulation\mupen64\tas\lol2.avi", fps = 60) StackHorizontal(vid1, vid2) LanczosResize(640,240).Fadein(60).Fadeout2(60) Then I just call x264.exe on it, one-pass usually with a constant rate factor (crf) of 26 (I have the cmd line saved in a text file so I don't have to memorize anything). I pull the wav from the avi and have lame turn it into mp3. mkvmerge does the rest. mencoder would probably be less steps if done right, but I have shortcuts to everything and so really it's like a 10-click process. Hope that helps.
Cool. I haven't gotten into avisynth much, but it is clearly very powerful. A frontend I use takes advantage of it though. Remember (you probably already know this): two of the best things you can do when encoding to save on bitrate or help with psnr is making sure it's mod16 in both directions and doing 2 pass for a much, much better distribution of bits.
Joined: 7/26/2006
Posts: 1215
Andypro wrote:
Cool. I haven't gotten into avisynth much, but it is clearly very powerful. A frontend I use takes advantage of it though. Remember (you probably already know this): two of the best things you can do when encoding to save on bitrate or help with psnr is making sure it's mod16 in both directions and doing 2 pass for a much, much better distribution of bits.
Well I always strive for mod16, since x264.exe tends to complain if it's not (most of my 64 encodes are 464x352). And sure 2-pass distributes bits well, but different videos require different bitrates (bitrate of super mario bros to look good < sonic 1 < sonic 3 < f-zero 64) and it's just much easier (and half the waiting time) to give it a crf that gives consistent quality among all videos. crf is still variable bitrate. and with "--refs 12" I'm not worried about bad motion estimation. Besides, even with a high qpstep, fast/complicated motion still looks like shit in a 2-pass size-oriented encode (see: gunstar heroes spinning credits and spaceship level). At that point why waste time doing multiple ugly multi-pass encodes before you get it right, when crf (quality-oriented) does the job fine on the first try. :) (can you tell I'm not a fan of detailless blocky encodes?)
Joined: 8/27/2006
Posts: 883
It's pretty hard to compare your run with comicalflop, that's the TASer that takes the more time to improve things. You shouldn't care too much, sure comicalflop will be faster when it's finished, but I won't be this year. So you should go ahead, continue to TAS, even if you are a little slower, you'll surely do a couple of tried and by the end of the game, you'll find that you should redo your entire run because your TASing skill will be greater that at the beginning.
Editor, Experienced player (730)
Joined: 6/13/2006
Posts: 3300
Location: Massachussetts, USA
Paused wrote:
You should have seen comicalflops first TAS. Sure it was great in the end, but he restarted the game from mid way through (ish) about 9 times. His TAS skills are due more to dedication, than pure talent. Thats the point, TAS take gaming skills out of the equation. (Not sure if it sounds like I am but just in case... This is not a Bash against comicalflop in anyway.)
You've hit the nail on the head, couldn't have said it better. (you are not bashing me, it was very well put, accurate, and not demeaning in any way.) And DKC2 102% needs a redo anyways, because Arne's any% brings new standards that my first run didn't match. You improve with time; I wasn't fully confident in myself with optimization until Feb when I started MK64, 7 months after I joined.
ZeXr0 wrote:
It's pretty hard to compare your run with comicalflop, that's the TASer that takes the more time to improve things. You shouldn't care too much, sure comicalflop will be faster when it's finished, but I won't be this year.
Old news, I'm not working on OoT anymore. (I'll be helping out where I can though.) My work up to where I stopped is still going to be in Guano's version however.
Homepage ☣ Retired
Joined: 5/17/2007
Posts: 124
Location: Cincinnati, OH
I redid up to kokiri sword and got it 100 frames or so faster than my old one. So compared to my old one it is a good improvement. Compared to Comicalflop's? Not so sure.
JXQ
Experienced player (750)
Joined: 5/6/2005
Posts: 3132
DaShiznawz wrote:
After seeing it put up against mine I am embarrassed I even tried. I really don't belong on this site. I guess wanting something and trying your hardest at it really doesn't help if you are talentless.
Two pieces of advice: 1) Keep trying if you're having fun. You're getting better each time. 2) Man the fuck up and lose this pity-party bullshit attitude. It's only hampering the progress you're making.
<Swordless> Go hug a tree, you vegetarian (I bet you really are one)
Joined: 5/17/2007
Posts: 124
Location: Cincinnati, OH
I didn't say anything self pity like in that post. I honestly don't know how close it is to Comicalflop's (his run desychs for me) I guess I could check the frames. EDIT: oh I didn't see what you quoted.
Joined: 3/15/2007
Posts: 131
DaShiznawz wrote:
I didn't say anything self pity like in that post. I honestly don't know how close it is to Comicalflop's (his run desychs for me) I guess I could check the frames. EDIT: oh I didn't see what you quoted.
I would really like to see you continue with your run, Dashiznawz. I enjoyed your original TAS, and I think that with time and experience you could be very good at TASing.
Editor, Experienced player (730)
Joined: 6/13/2006
Posts: 3300
Location: Massachussetts, USA
Let's compare frame numbers then. The very first frame of the cutscene where Link is opening the chest; you see the chest on the side like this: For me, this is frame 15979.
Homepage ☣ Retired
Experienced player (612)
Joined: 4/24/2005
Posts: 612
This is DaShiznawz' run, the very first frame of the cutscene where Link is opening the chest. Interesting enough, it's from another angle than that of Comicalflop's. You're about 33 frames slower, DaShiznawz. Don't give up though, you're really close to either matching it or surpassing it. I have faith in you, I'm sure quite a few other people do to. Also remember, TASing takes time so don't feel down because you're not the 'best TASer' because there is no such thing as the 'best TASer' around here. All we really have is experience, with the emulator and testing crap out in the game so take your time and enjoy yourself, the best you can.
Player (34)
Joined: 12/18/2005
Posts: 250
I know this may seem like a dumb question, but does the duration of the cinematic change if the angle is different? (That is to say, it's faster or anything in a certain viewpoint?) Sorry if it's a dumb question :X
我々を待ち受けなさい。
Joined: 6/26/2007
Posts: 147
meepers wrote:
I know this may seem like a dumb question, but does the duration of the cinematic change if the angle is different? (That is to say, it's faster or anything in a certain viewpoint?) Sorry if it's a dumb question :X
You have too little confidence. That is exactly the type of question you have to ask while TASing. The only thing I can think of is maybe some camera angles create a few more frames of lag because of what's in the background. The Kokiri sword doesn't have much going on in the BG, but it still might be worth investigating the amount of frames that follow the two screens there.
Joined: 7/26/2006
Posts: 1215
same number of frames. I can't remember any lag from this or MM's engine except when there are multiple explosions going on. It kind of helps that it has two thirds of the framerate of the average n64 3D adventure game to begin with.