Dynamic Animation of Spiders and Insects
- sinbad
- OGRE Retired Team Member
- Posts: 19269
- Joined: Sun Oct 06, 2002 11:19 pm
- Location: Guernsey, Channel Islands
- x 66
- Contact:
- deficite
- Gremlin
- Posts: 160
- Joined: Sat Oct 15, 2005 4:31 pm
- Location: TN, USA
- Contact:
That video is so cool. I especially like at the beginning the way the spider feels around the side of the cube trying to get good footing. I also like at the end how the spider crawls around the walls of blocks and keeps balance until the wall tumbles, the way the spider falls and tries to get the blocks off of himself to start moving again is really cool.
Oh, and the giant spider at the end creeped me out. I saw some blocks getting pushed at the bottom and thought "What is doing that?" and then all of a sudden this huge spider came out. Better than any horror movies I've ever seen
Oh, and the giant spider at the end creeped me out. I saw some blocks getting pushed at the bottom and thought "What is doing that?" and then all of a sudden this huge spider came out. Better than any horror movies I've ever seen
- PolyVox
- OGRE Contributor
- Posts: 1316
- Joined: Tue Nov 21, 2006 11:28 am
- Location: Groningen, The Netherlands
- x 18
- Contact:
-
- Gnome
- Posts: 351
- Joined: Thu Oct 21, 2004 1:34 pm
- Location: Diepenbeek, Belgium
- x 3
- Contact:
- Chris Jones
- Lich
- Posts: 1742
- Joined: Tue Apr 05, 2005 1:11 pm
- Location: Gosport, South England
- x 1
- Devil N
- Greenskin
- Posts: 123
- Joined: Sun Jun 26, 2005 4:32 pm
- Location: Groningen, The Netherlands
- Contact:
-
- OGRE Expert User
- Posts: 1067
- Joined: Mon Mar 29, 2004 8:49 pm
- Location: the Netherlands
- x 43
- Contact:
Wow, this is some really impressive work! I did notice that when the camera was a bit closer, the feet seemed to float. Was that caused by rendering errors due to the particles in the walk-on-the-ground-animation, or were they really not on the ground?
By the way, is it a technical limitation that the spiders are so slow at doing 90 degree angles, or is that a graphical thing because it looks nice? I never saw a spider hesitating before rounding a corner. Then again, I never saw such huge spiders either...
I really like it that you actually took a realistic model and lighting setup. Most animation papers seem to want to prove that an ugly model never looks realistic, regardless of how well it is animated.
By the way, is it a technical limitation that the spiders are so slow at doing 90 degree angles, or is that a graphical thing because it looks nice? I never saw a spider hesitating before rounding a corner. Then again, I never saw such huge spiders either...
I really like it that you actually took a realistic model and lighting setup. Most animation papers seem to want to prove that an ugly model never looks realistic, regardless of how well it is animated.
Spore actually does not really use procedural animation. It used meta-animations that are translated to different models, while these meta-animations are hand-animated. Only the translations are procedural.bharling wrote:awesome
reminds me of some videos from Spore. Procedural animation is the future i believe.
Please say you are going to publish the paper
My dev blog
Awesomenauts: platforming MOBA (PC/Mac/Linux/XBox360/X1/PS3/PS4)
Blightbound: coop online dungeon crawler (PC)
Swords & Soldiers: side-scrolling RTS (Switch/PS3/Wii/PC/Mac/Linux/iPhone/iPad/Android)
Proun: abstract racing game (PC)
Cello Fortress: mixing game and live cello performance
The Ageless Gate: cello album
Awesomenauts: platforming MOBA (PC/Mac/Linux/XBox360/X1/PS3/PS4)
Blightbound: coop online dungeon crawler (PC)
Swords & Soldiers: side-scrolling RTS (Switch/PS3/Wii/PC/Mac/Linux/iPhone/iPad/Android)
Proun: abstract racing game (PC)
Cello Fortress: mixing game and live cello performance
The Ageless Gate: cello album
- Sundown
- Gnoblar
- Posts: 24
- Joined: Fri Jun 15, 2007 3:32 pm
Thanks for the feedback, much appreciated!
It depends on the type of angle really. With positive 90 degree angles (i.e creature inside a hollow cube), its basically just for show - any faster and it would be difficult to see what exactly the creature was doing. Hence why I've chosen a larger spider for most of these videos!
With negative 90 degrees (surface of a cube), its more to do with the creature trying to keep 'balanced', i.e none of its legs are too far away from the body / too stretched. The end result is that precipices like that require a more cautious approach. Any warning flags raised in the regulatory modules automatically make the creature stop and re-adjust.
As far as technical limitations, the bottleneck is pretty much the collision detection, which limits how fast the legs can move. At the moment the simulation is frame-dependent, locked at 60fps maximum. For example, recording a video halves the frame-rate on my computer, so the simulation basically runs in slow motion, the end video recorded at 60fps and so unaffected.
I've ran the simulation on an extremely powerful machine, and removed the frame-rate cap. What you get then is more analogous to smaller spiders - they can easily navigate 90 degree changes of surface angle extremely quickly.
I hope that answers your question. Short version - its a bit of both!
I made that video about 6 months ago, but I think its due to the underlying creature skeleton not being perfectly lined up with the model itself, so the (invisible) skeletal limb terminates in a different place to the overlying mesh. Just a 3DS Max tweak really.Oogst wrote:I did notice that when the camera was a bit closer, the feet seemed to float. Was that caused by rendering errors due to the particles in the walk-on-the-ground-animation, or were they really not on the ground?
By the way, is it a technical limitation that the spiders are so slow at doing 90 degree angles, or is that a graphical thing because it looks nice? I never saw a spider hesitating before rounding a corner. Then again, I never saw such huge spiders either...
It depends on the type of angle really. With positive 90 degree angles (i.e creature inside a hollow cube), its basically just for show - any faster and it would be difficult to see what exactly the creature was doing. Hence why I've chosen a larger spider for most of these videos!
With negative 90 degrees (surface of a cube), its more to do with the creature trying to keep 'balanced', i.e none of its legs are too far away from the body / too stretched. The end result is that precipices like that require a more cautious approach. Any warning flags raised in the regulatory modules automatically make the creature stop and re-adjust.
As far as technical limitations, the bottleneck is pretty much the collision detection, which limits how fast the legs can move. At the moment the simulation is frame-dependent, locked at 60fps maximum. For example, recording a video halves the frame-rate on my computer, so the simulation basically runs in slow motion, the end video recorded at 60fps and so unaffected.
I've ran the simulation on an extremely powerful machine, and removed the frame-rate cap. What you get then is more analogous to smaller spiders - they can easily navigate 90 degree changes of surface angle extremely quickly.
I hope that answers your question. Short version - its a bit of both!
- deficite
- Gremlin
- Posts: 160
- Joined: Sat Oct 15, 2005 4:31 pm
- Location: TN, USA
- Contact:
- Aladrin
- Orc
- Posts: 465
- Joined: Fri Mar 10, 2006 10:22 pm
I was just watching this video and something bothers me. At 5:30, the wall falls as the spider is climbing it... I think it falls the wrong way. if I walk up to a stack of boxes and start to climb it, they are going to fall -on- me, not away from me, unless I jump at the boxes. These boxes fall after the spider has climbed for a while... The second set of boxes makes it even more apparent as the boxes continue to fall as it crosses the width of the columns.Sundown wrote:YouTube link : http://www.youtube.com/watch?v=I1P_B65XW4I
High Quality link : http://www.informatics.bangor.ac.uk/~llyr/divx.html
Maybe there's some physics here I don't understand, but it seems backwards to me.
Edit: Awesome stuff, btw... It creeps me out watching them... They're too real.
- cybereality
- Hobgoblin
- Posts: 563
- Joined: Wed Jul 12, 2006 5:40 pm
- x 12
- Sundown
- Gnoblar
- Posts: 24
- Joined: Fri Jun 15, 2007 3:32 pm
I'm not sure what would happen if a spider climbed a stack of small boxes like in the video, but my guess is that you are indeed right!Aladrin wrote:Maybe there's some physics here I don't understand, but it seems backwards to me.
The reason why it happens is the boxes are only shifted due to the spider tapping on them with its leg - the actual weight of the spider due to gravity is not taken into account (i.e gravity is switched off), unless the spider has too little grip i.e at least 5 of its legs are 'anchored' to the surface. At that point, gravity is activated and the spider becomes a temporary ragdoll, until it finds itself with enough potential footing once more (or the fall kills it).
It is possible to have gravity switched on at all times, which would cause the boxes to shift towards the creature (due to its weight) and away (due to the leg impacts). It would have a large impact on simulation speed though, due to the constant gravitational force acting upon the 56 or so rigid body parts that make up the articulated skeleton! I'll try making a test video when I have the chance..
-
- Ogre Magi
- Posts: 1266
- Joined: Tue Aug 12, 2003 1:53 am
- Location: Melbourne, Australia
- x 1
- Contact:
Looks like the next version of newton has a massive speed up for many colliding objects so hopefully you'll be able to do stuff like this and remain realtime (even multiple spidersSundown wrote:It would have a large impact on simulation speed though, due to the constant gravitational force acting upon the 56 or so rigid body parts that make up the articulated skeleton! I'll try making a test video when I have the chance..
- Sundown
- Gnoblar
- Posts: 24
- Joined: Fri Jun 15, 2007 3:32 pm
Quoting myself...
http://www.youtube.com/watch?v=NMh5sRXqHq8
The spider has exactly the same 'brain' as the tarantulas seen in the older videos, only scaled down and as a byproduct of the framerate, sped up. The result is behavioural animation more similar to that of a house spider - where corner navigation happens very quickly.
I've finally recorded an example of what happens when I run the simulation at 120fps:Sundown wrote:I've ran the simulation on an extremely powerful machine, and removed the frame-rate cap. What you get then is more analogous to smaller spiders - they can easily navigate 90 degree changes of surface angle extremely quickly.
http://www.youtube.com/watch?v=NMh5sRXqHq8
The spider has exactly the same 'brain' as the tarantulas seen in the older videos, only scaled down and as a byproduct of the framerate, sped up. The result is behavioural animation more similar to that of a house spider - where corner navigation happens very quickly.
- pr3dali3n
- Gnoblar
- Posts: 3
- Joined: Sat Jul 26, 2008 9:52 pm
- Location: Vault 13
- Contact:
- sinbad
- OGRE Retired Team Member
- Posts: 19269
- Joined: Sun Oct 06, 2002 11:19 pm
- Location: Guernsey, Channel Islands
- x 66
- Contact:
-
- Gnome
- Posts: 334
- Joined: Wed Aug 02, 2006 9:27 am
- Location: Toronto, Canada
This is fantastic. Are you releasing source or going to make this project commercial?Sundown wrote::http://www.youtube.com/watch?v=NMh5sRXqHq8
- Sundown
- Gnoblar
- Posts: 24
- Joined: Fri Jun 15, 2007 3:32 pm
I'm not sure what my University's policy is on that actually (these videos come from my PhD research). I'll definetely link to my thesis once this is all over though, and a paper that condenses it all for the lighter read!voxel wrote:This is fantastic. Are you releasing source or going to make this project commercial?Sundown wrote::http://www.youtube.com/watch?v=NMh5sRXqHq8
I'm currently extending the simulation to animate lizards, should have a video by the end of the week if all goes well.
-
- Kobold
- Posts: 30
- Joined: Fri Jan 19, 2007 10:42 am
Excellent, i've been watching this and hopefully your thesis gets the attention it deserves, specially if you can do this quadpeds too. What would it require to transform this for creatures that have higher center point of mass?Sundown wrote: I'm not sure what my University's policy is on that actually (these videos come from my PhD research). I'll definetely link to my thesis once this is all over though, and a paper that condenses it all for the lighter read!
I'm currently extending the simulation to animate lizards, should have a video by the end of the week if all goes well.
- NickM
- Bronze Sponsor
- Posts: 65
- Joined: Sat Jul 01, 2006 10:41 pm
- Contact:
- Sundown
- Gnoblar
- Posts: 24
- Joined: Fri Jun 15, 2007 3:32 pm
A more complex inverse dynamics system would be the main thing. Spiders, insects and lizards have a low center of mass, so you can get away with building the articulated skeleton as a chain of hinge joints (horizontal for joint connecting first leg segment to body, vertical for the rest).Entropai wrote:Excellent, i've been watching this and hopefully your thesis gets the attention it deserves, specially if you can do this quadpeds too. What would it require to transform this for creatures that have higher center point of mass?
The main reason I didn't attempt the extension is that the physics engine I used (Newton) would struggle to simulate the creature's articulated structure if gravity was enabled at all times. However, now that I'm using the Newton 2.0 beta it seems to handle things a lot more efficiently.