Posts for kooz


Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
I very much enjoyed your WIP and look forward to seeing more!
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Wow. Wipeout is a pretty fast-paced game when you're not bumping into a wall every 5 seconds. I had no idea, because I absolutely suck at it. =D Pretty cool! Though this just gave me the thought that it would be sweet to see some Rollcage TAS'd.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
I absolutely love the reasoning behind this run, and content made with this mindset is rare but my favorite thing here at TASVideos. Just about to watch. From the submission text and comments so far, I can tell it's going to be epic. =) Edit: oh my god, that sound at 50:19 scared the hell out of me, hahaha Edit #2: Just finished watching this morning. Really entertaining and well done! Huge yes vote from me.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
This is so freakin' SICK! I've been wanting to see an optimized 100% run since my very first submission nearly 17 years ago. Beautifully done. A very enthusiastic yes vote from me!
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
SMB's fluid character movement makes for incredibly watchable speedruns (by humans like myself, who typically aren't able to enjoy hyper-glitched-type runs). Even the wall-clipping was easy to follow. I could watch this for hours - which is why I really love the 106% run that MatteSMB did a couple years ago (https://www.youtube.com/watch?v=EF7ISgbS2-w).
scrimpeh wrote:
Absolutely phenomenal once again. I'd really love if the run were submitted to the site.
I agree wholeheartedly. I'm not sure why he hasn't submitted them. libTAS runs have been acceptable on the site for a couple of years now I believe.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Yes vote from me! I'd love to see a glitchless run at some point too. On a side note, anyone know what is going on with the weird antialiasing in the encode? Some temporal AA or deinterlacer being applied perhaps? I'm curious to know at what stage that's being introduced - I've never seen it before.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
I will agree that SpinRite is a pretty powerful and useful tool, but would caution that if the disk is marking new bad sectors at a rapid pace, some of its aggressive data recovery/repair settings could cause a failing drive to fail completely. If the drive is still somewhat readable, I would recommend attempting to recover only your most valuable data before scanning the entire drive.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
CoolKirby wrote:
This sounds pretty interesting. Awaiting Spikestuff's temp encode.
Sounds interesting indeed - the music is awesome, and is this game's most redeeming quality, IMO.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Ferret Warlord wrote:
Having watched some other videos of this game, I think that maybe it's for the best that the sound effects didn't survive.
No kidding. I remember always playing with them turned off. Hell, the game itself refers to the sound effects as "NOISE" most of the time.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
slamo wrote:
Jumping that distance allows me to chain another jump to get over the next ledge. Instead of landing and jumping at separate times, chaining it combines the animation into something shorter to save a few frames. Falling off a ledge also uses an animation so there's no advantage doing it.
Cool. I had a feeling there was a good reason for that. Definite yes vote from me either way! =)
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Awesome. This was one of my childhood favorites. Quick question: On Level 11 (at 4:02) you jump off a ledge to tag the EPIC bonus and avoid a ball rolling back and forth. Wouldn't dropping off the ledge and taking damage from the ball be faster?
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
The Learning Company presents: "Ancient Empires" aka "Super Solvers: Challenge of the Ancient Empires!" I loved this game when I was a little kid. I replayed it a couple years ago, and surprisingly, it held up as an enjoyable (and occasionally still challenging) edutainment title. The animation framerate is god awful, but I still think it would be really cool to see it optimized.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Man, I was pretty excited there for a sec.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Oh my god... I know this run can't be accepted here yet, but this is without a doubt my favorite TAS of all time! Such amazing work. Please tell me you're going to do a 106% run next! Also, any chance we can get a 60FPS encode? =)
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
BrotherMojo wrote:
I got the impression that the source engine uses discrete physics ticks at a set rate.
Same here, though I assumed the tick rate was tied to the display refresh rate for the sake of efficiently conveying smooth motion. I believe the tick rate can be set independently of both timescale and frame-rate, however, which leads me to believe physics engine motions are vectors rather than hard-set keypoints.
BrotherMojo wrote:
Though again, this is just the impression I got from this guy. You'd probably need to do some digging to verify all this.
Haha, so true. Gotta love how we're both sitting here with the internet (answer) at our fingertips and yet we choose to speculate =D
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
I imagine this would be quite a bit different from the TASes we are used to seeing here. Having played plenty of Source engine games myself, I doubt turning the timescale down will bring gameplay to the granular level of frame-advance TASing, where you are dealing with the absolute limits of a game's engine. I'm sure camera movement would not be sluggish at a reduced timescale, nor would the framerate. These kinds of advantages, while technically still a form of tool-assisted play, I think might break the barrier of what could be accomplished in real-time, and perhaps may go against some of the philosophies here, I'm afraid. If for instance you play at 60fps, and you reduced the timescale down to 0.1 while still technically getting 60fps (and I assume therefore still recording inputs at 60fps), then playback at normal speed would mean 600 inputs/sec. *shrug* Maybe the only reason we don't have TASes with that much temporal resolution yet is because we don't have any systems emulated with re-recording that would benefit from it yet. Just some thoughts.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Awesome! Glad to see this run finally completed. Also thanks for the credit, though I'm not sure my contributions warrant being a co-author. Nevertheless, very well done, and an obvious yes vote from me.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Wow, those shortcuts were really nice. Great work... Plok is looking far more agile than ever before. As far as vectorization... I personally think the result of hq4x filter looks better than the result you got with Inkscape (with the downside of not having infinite scalability... perhaps a pre-filter with hq4x prior to converting to vector would improve results?). Not to venture too much further off-topic, but take a look at this article if you get a sec: Depixelizing Pixel Art I'd love to see that vectorization algorithm implemented in emulators... Its gorgeous. More so, I'd like to see it applied on a per-layer basis, so as to reduce artifacting as much as possible (i.e. a score counter should not be interpreted as part of the background, nor should sprites, etc...) I don't think any emulator does that yet, with any filter type. Though it could be implemented in a final encode, I believe. You'd have to dump an AVI for every layer (so, 5, I think?) and recombine them using video compositing software.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Beautiful run so far, beautiful encode =) I'm sure not everyone likes hq4x, but it sure made for some damn fine TAS viewing at work.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
VirtualDub will work perfectly for this. All you need to do is go to Video > Frame Rate (Ctrl+R) and in the "Frame rate conversion" section, choose "Process every other frame (decimate by 2)" This will properly decimate the video and you should have true 30fps output. =)
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
My video should be 30fps. Dolphin records at 60fps for some reason, when RE4 is only 30fps. I use a video editor to slow down the video from 60 to 30, halving the speed, but keeping all visible frames intact.
Not that it matters too much as I'm sure the final encode will be correct if/when this is published, but I also noticed what AjAX was referring to... Your videos are 30fps (29.970 to be exact), but content-wise they are only 15fps (so, half of what the game actually displays). Stepping through frame-by-frame, you'll see each frame is duplicated. Again, not really that big of a deal for WIPs, but I thought I'd point it out. Unfortunately I do not know enough about Dolphin or how it exports video, but I'm sure someone here knows how you can fix this. My guess would be the program you're using to decimate from 60 to 30 is not working as intended. If I'm not mistaken though, H/X264 codecs will automatically take redundant frames into account and not superficially increase filesize if you were simply to encode the original 60fps output anyway... BTW the TAS is looking awesome so far.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Nice ASCII chart! Also, neat graphics glitch in Plok Town, I've never seen that. Is that a side-effect of the mid-air sawblade activation you discovered? I'm both surprised and pleased that the Penkinos battle was as optimized as possible! I was specifically thinking of that fight when I mentioned how difficult the bosses are. Haha... I was completely thrown for a loop when you started skipping fleas and limbs in Venge Thicket. Dreamy Cove race was AWESOME. It almost looks like you never had to slow down.
I'm sure that it wouldn't be acceptable to publish a warped run and a warpless run here if there is no additional restriction on the warpless run. The warped run plays 20 stages out of a total of 37 stages exactly as it would be done in a warpless run. There are only 11 levels completely missing in the warped run. (And the remaining 6 stages are those where I use a warp.) The "no damage" restriction would enforce different strategies on those two runs, which would especially make the bosses look very different.
I am phoning in my "YES" vote for this proposed run in advance. =) Warpless + no-damage is the ideal run for entertainment in my opinion. Those were the primary goals in my very first submission (deservedly rejected, but not specifically for those reasons). There are enough obstacles to present some interesting challenges to a no-damage goal, as well as the additional gameplay mechanics (and of course, levels) that aren't present when warps are used. As I mentioned before, I'm really interested in how the strategy will work out in the early Akrillic levels where the amulet is not available to insta-kill fleas. Specifically Garlen Beach, where a zig-zag sort of route is required - I believe this will be the most challenging level to optimize. Manipulating the egg bounce/flea jump direction, in order to efficiently manage several at once (while they are also each in different stages in their "life-cycle")... it will be very interesting to see.
I also could verify that you can skip parts of the last levels.
Wow, nice find. It looks similar to the glitch I found to pass through spikes in Silent Running, but I don't see you taking damage... Is that just a small opening in the barrier (level design oversight), or did you discover some sort of maneuver to glitch through there? Side note: those on-screen variables are cool. On WIP30 @ youtube:
2 frames gained before the warp. 1.8s ingame time gained in the race (no more forced waiting!) ~4s gained at the boss fight, although it's not fully optimized.
Wow... the more I watch your WIPs, the more I feel lucky to have even had my run published in the first place... haha... 2 frames at Dreamy Cove don't seem like much, but comparing to my run, I see you managed to avoid hitting the left edge of the platform on your way into the water. I imagine you did that by delaying pressing D... why the hell didn't I think of that... *facepalm* It's not only faster, but it looks cleaner too. Bah! I can't believe my Womack fight was so poorly optimized! I'm sure all that time didn't just come from sliding down that hill either. From what I can tell, the strategy is the same (pre-fire a limb and use high jump to elevate it). I'm curious to know how you improved it so much, with apparently even further improvement to be had.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
It's looking fantastic. I can't think of anything else to say about the run so far, except "great work!" As far as the possibility of a "no warps" run, obviously I have expressed my interest in such a run. Though I don't really see much of a point in the "no damage" or "get every present at least once" goals... Entertainment-wise, I can't see them being that much of a contribution, and I can see the resulting TAS being down-voted due to what would be perceived by the community as "pointless" goals... (though I can understand where you're coming from) I'd say the present acquiring idea would best be left for a longplay, where the idea is as close to 100% exploration of the game as possible. If you watch the aforementioned vid, you'll see the author actually purposefully fails bonus levels in an effort to show the viewers everything in the game. I would watch it, but I doubt such a run would be accepted here. I am quite excited that you seem to be taking the idea of a warpless run seriously. =)
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Shell warps... haha, I'm learning new things about this game every day.
partyboy1a wrote:
So what's better? Take ALL the warps, or take non-shellgame-warps, or don't use any warps at all?
Don't use any warps at all! Entertainment-wise this would be an awesome run in my opinion... Realistically though, I agree with Sonnikkustar - you should definitely use any/all warps that can get you to the end of the game as fast as possible.
Experienced Forum User, Published Author
Joined: 8/3/2004
Posts: 99
Location: Michigan, United States
Nice work! Especially with figuring out how to maximize the results when sliding... and really cool shortcut at Zig Zag! You are destroying my run and I love it. About the previous movie ending early... I don't know the rules on it anymore, but I remember a HUGE debate on the matter in regards to SMB1... If I recall, as long as the movie ends at a point where any additional input cannot affect the outcome of the game, its acceptable... Am I right? Bosses in this game are a bitch... I had a sneaking suspicion you'd get hung up on some of them. =/ I have no idea if my playing was ideal or if you're way ahead, or whatever... but feel free to hex my boss fights in, if it will help. (lol, probably not though, judging by how much you're improving on my run). Don't give up!