Comments on: Dead Space Dissection: The Trouble With Ragdolls http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/ Breaking Into the Industry Sun, 28 Feb 2010 07:37:08 +0000 http://wordpress.org/?v=2.8.2 hourly 1 By: flYBOi http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-18190 flYBOi Fri, 01 Jan 2010 11:07:05 +0000 http://www.mandible.net/?p=112#comment-18190 I'm pretty sure Unreal Tournament 3 (2007) had showcased technology where you actually could feign death. Control of your character model would be revoked from you and your 'avatar' would just fall to the ground as per every other death in Dead Space. 'Recover' from your 'death' and your avatar would get up from the ground, rendering you control again. just my 2 cents... anyway, games nowadays integrate ragdoll physics and pre-rendered motion. Just look at Killzone 2. I’m pretty sure Unreal Tournament 3 (2007) had showcased technology where you actually could feign death. Control of your character model would be revoked from you and your ‘avatar’ would just fall to the ground as per every other death in Dead Space. ‘Recover’ from your ‘death’ and your avatar would get up from the ground, rendering you control again.

just my 2 cents…

anyway, games nowadays integrate ragdoll physics and pre-rendered motion. Just look at Killzone 2.

]]> By: IA http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-14608 IA Sat, 17 Oct 2009 22:46:57 +0000 http://www.mandible.net/?p=112#comment-14608 ... roflcptr eviscerate :D that just made my day! …
roflcptr
eviscerate :D
that just made my day!

]]> By: Nestor http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-6049 Nestor Tue, 31 Mar 2009 23:21:53 +0000 http://www.mandible.net/?p=112#comment-6049 Interesting, can't they just zero out to a default pose to recover from the ragdoll and start replaying the animation from there? It's kind of like the problem you describe of the death animation snapping in a non sequitur way but you're probably less likely to notice. In DS it could be a kind of spasm where they go stiff as a board before getting up again Interesting, can’t they just zero out to a default pose to recover from the ragdoll and start replaying the animation from there? It’s kind of like the problem you describe of the death animation snapping in a non sequitur way but you’re probably less likely to notice. In DS it could be a kind of spasm where they go stiff as a board before getting up again

]]> By: SnapDragon http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-3515 SnapDragon Sat, 06 Dec 2008 17:55:11 +0000 http://www.mandible.net/?p=112#comment-3515 Trespasser - a 10-year-old game! - actually did animate dinosaurs using inverse kinematics. And it worked really, really well (ok, there WERE cases where it failed spectacularly, but the same can be said for ragdoll physics). It's stunning that nobody's been able to match it since. Pity the game was pretty bad. :) (Granted, I doubt you could make human motion look natural using IK. But for zombies, or dinos, or anything else... it'd be great.) Trespasser – a 10-year-old game! – actually did animate dinosaurs using inverse kinematics. And it worked really, really well (ok, there WERE cases where it failed spectacularly, but the same can be said for ragdoll physics). It’s stunning that nobody’s been able to match it since. Pity the game was pretty bad. :)

(Granted, I doubt you could make human motion look natural using IK. But for zombies, or dinos, or anything else… it’d be great.)

]]> By: Zorba http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-3507 Zorba Fri, 05 Dec 2008 13:29:45 +0000 http://www.mandible.net/?p=112#comment-3507 Basically, yep. Although most poker players aren't concerned with tells after they've been shot. :) Basically, yep. Although most poker players aren’t concerned with tells after they’ve been shot. :)

]]> By: Lar http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-3501 Lar Thu, 04 Dec 2008 18:12:20 +0000 http://www.mandible.net/?p=112#comment-3501 So basically the zombies are bad poker players and have tells? =) So basically the zombies are bad poker players and have tells? =)

]]> By: Requies http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-3493 Requies Wed, 03 Dec 2008 09:10:44 +0000 http://www.mandible.net/?p=112#comment-3493 That was me above. That was me above.

]]> By: Anonymous http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-3492 Anonymous Wed, 03 Dec 2008 09:09:40 +0000 http://www.mandible.net/?p=112#comment-3492 You got a lot more involved in your "is it dead" detection. I operated under a different model, namely: Is it moving? Yes: Shoot it No: Did it drop loot? Yes: It's dead, collect loot. No: It's playing dead, shoot it some more. This works 95% of the time in Dead Space. You got a lot more involved in your “is it dead” detection. I operated under a different model, namely:
Is it moving?
Yes: Shoot it
No: Did it drop loot?
Yes: It’s dead, collect loot.
No: It’s playing dead, shoot it some more.

This works 95% of the time in Dead Space.

]]> By: Zorba http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-3490 Zorba Wed, 03 Dec 2008 06:10:07 +0000 http://www.mandible.net/?p=112#comment-3490 Not until we solve a lot of very complicated problems. As long as we're doing pre-canned animations, we can just assume that the animation will flow reasonably smoothly, since it's designed frame-by-frame to do so. Once we get into physics issues, we start running into <i>huge</i> issues involving AI required just to walk around. We end up with the same issues that the people trying to design humanoid walking robots do, only our robots are supposed to be able to do karate and shoot people as well. The advantage is that it's all in a world that we control, so we don't have to worry about high-precision sensors, but the disadvantage is that we're supposed to be animating a <i>lot</i> of these things. I'm not sure anyone has really tried - even Spore only has a relatively small number of canned animations. Not until we solve a lot of very complicated problems. As long as we’re doing pre-canned animations, we can just assume that the animation will flow reasonably smoothly, since it’s designed frame-by-frame to do so. Once we get into physics issues, we start running into huge issues involving AI required just to walk around.

We end up with the same issues that the people trying to design humanoid walking robots do, only our robots are supposed to be able to do karate and shoot people as well.

The advantage is that it’s all in a world that we control, so we don’t have to worry about high-precision sensors, but the disadvantage is that we’re supposed to be animating a lot of these things. I’m not sure anyone has really tried – even Spore only has a relatively small number of canned animations.

]]> By: Yonatan Zunger http://www.mandible.net/2008/12/02/dead-space-dissection-the-trouble-with-ragdolls/comment-page-1/#comment-3489 Yonatan Zunger Wed, 03 Dec 2008 00:59:17 +0000 http://www.mandible.net/?p=112#comment-3489 I assume that it's not practical to use the physics model for creatures apart from when they're dead? I assume that it’s not practical to use the physics model for creatures apart from when they’re dead?

]]>