View Full Version : Dinosaure Walk cycle WIP

11 November 2008, 12:07 PM
Hi all

I'm actually animating an albertosaurus for a crysis mod called Lost World returns

I would like to have your opnion about my rought walk cycle

For info this dino weight 800 kg and is 3 meters tall

Here are the anims :

and in slow mod :

Any crits , advice are really welcome :)

12 December 2008, 04:08 PM
Its quite good actually - maybe you could add more details on some parts of the body like -front arms and haed /

good work anyway ;)

12 December 2008, 07:49 PM
thanks , really appreciate :cool:

I will add more overlapping to the front part and to the tails too , and also correct the toe when the foot is leaving the ground , cause actually at one moment he is on all his toes which his normally impossible

new version comming soon :)

12 December 2008, 10:18 AM
Good start. Something to think about is your spacing.
Right now everything moves in a steady motion, there is no variety. Think about his legs for instance. When he pushes off you want to show the weight and force of the push, and the acceleration of his foot leaving the ground.
But you have him moving in a constant motion. Try and break that up! This is something you could apply to the rest of his body too of course.

12 December 2008, 09:38 PM
I like what you have so far. I agree with adding a little more movement in the arms and tail. Also maybe a bit more weight in his feet, looks like he is kind of creeping right now.

Looks good though

12 December 2008, 11:26 PM
New version of the walk cycle
i hope you like it

as always need critics for improovement

12 December 2008, 05:41 PM
This feels like a very well-polished animation overall-- convincing weight for the torso, and there are interesting details throughout the movement of his body (neck and tail movements are fun to watch).

The only thing that doesn't sell it for me are the toes. They drag a little bit, and when the foot first makes contact with the ground, the toes actually go below the plane the foot rests on. A simple lifting of the toes would fix this.

(Also, first post-- hello everyone!)

12 December 2008, 08:32 PM
thanks :bounce:

here are two other versions :

Which one should i choose in your opinion?

12 December 2008, 09:14 PM
Probably v3-- the differences between v3 and v3.2 are very subtle, but I think the toe lift in v3.2 looks slightly unnatural.

12 December 2008, 01:37 PM
overall looks good and solid for me _/
good work :)

12 December 2008, 02:23 PM
Looks really good! Only crit from me would be he looks a little "light footed" in his step.

May need a little more show of weight when the foot comes down.

I think I like v3 better than v3.2, also...

12 December 2008, 11:48 AM
cant really comment on the technical side of things, but from a purely Dinosaur "fan" point of view he seems a bit too relaxed, especially if its for a computer game. Perhaps his head could be ducked slightly, flattening the line from his snout to tail, then the hips raised a bit more on the steps and the steps lengthened a small amount to make him appear to be "stalking".

I guess you will have various stances for "relaxed", "aware" and "chasing" if the AI accomodates this, so if the player happens upon an unaware Albertosaurus, it could be just meandering around until it notices the player, so this walk will look fine. However, for efficiency, I would personally just have it always "stalking", so that it always appears threatening.

its a really smooth and realistic looking walk though, I think I would probably end up overdoing the weightiness of it and overselling the idea that it was a lumbering dinosaur, though that would just be my personal preference :D

12 December 2008, 03:14 PM
That looks very nice! I'm not sure whether youíre going for accuracy, but here are some comments on its posture, (ps Sorry if they come across as negative, there not meant to be)

The hands are shown Ďpronatedí, in the 'bunny' pose. This means that the radius is crossed over the ulna and causes the palm to face down. No theropod is known to have this ability (even birds) and itís not how the arm articulates. The palms should face each other, like when you hold a basket ball. Only the quadrapedal dinosaurs have arms that articulate crossed.

The upturn in the neck is too strong. Tyrannosaur necks arenít particularly fixable, and what you have now is probably dislocated.

I know when animating walk cycle youíre told that at some point the leg straightens, but in most dinosaurs they canít straighten. The shin bones canít straighten more than about 120 degrees in relation to the femur. The only dinosaurs that can straighten their knees are the sauropods and some stegosaurs.

Try not to curl the toes back too much once the foot leaves the ground. During the walk the toes are taking the weight and the tendons are taking the strain. When the foot lifts off the tendons loosen briefly, they donít actively curl up. This means that the toes would be straighter for a short time after. If that makes sence? Go to the BBC motion gallery and search for Ostrich.

This is might be a rigging comment, but it Ďlooksí as though the hip bone is moving up and down during the walk cycle. The hip is fused to the vertebra and canít move. Or is an this an illusion created by the twisting of the spine?

Again, if youíre not going for accuracy then ignore my comments.

It looks really promising and flows very nicely. :)

01 January 2009, 04:57 PM
hi all thanks for you replies

so i tried to use the ostrich as reference , but as it doesn't have the same weight , things are a bit different .
here is the reply from our paleotologist :

Quick comments:
- An Ostritch isn't advisable to use as a model. This is primarily due to the significantly different hip structure, wider hips and lack of a tail. The lower leg motion of the ostritch and other running birds can be a useful reference (I believe I posted a roadrunner video in the animations thread).
- I'm not entirely convinced about the neck being dislocated, but it is certainly worth looking into. Generally speaking, the neck should always maintain a slight curve, but it can also be relatively strait. I'll take a close look at the neck on the 22nd.
- The point about the leg straightening if correct. I find it a bit confusing though: As you and I have already created an animation in which the leg never fully straightens (and where the bones never reach vertical either)???
- As for the other comments they are accurate. I recognised the author of the post from some of our reference material. He has done some very modern and careful drawings of reconstructions. When I come back I'll see about contacting him for extra techical advice.
- The fact that there are so few corrections from someone with the background he has should be taken as a complement.

Regarding the arms etc: There are actually a couple of small changes I'd like to make to the model. The most significant is decreasing the width at the belly. This isn't a problem when the animal is standing, but it causes the animal's leg to collide with the belly during walking movements (which will look strange in the textured model and would cause considerable chaffing in real life).

The other noticable one would be to bring the shoulder blades a bit closer to the ribcage (I was initially uncertain about this, but examining skeletons at museums recently convinced me that the separated should blades were an artifact of how the skeletons were mounted in museums). I won't be able to get at the references until the 22nd though.

We also have to ballance accuracy with appearance and modelling effort. At some point it is important to say "enough is enough" and bask in the glory of the animal & art.

also if you are interrested by fallowing the mod devellopent or even help us to devellope the mod , please visit this link :

thanks again everyone

01 January 2009, 10:07 PM
Regarding the leg staightening : I brought it up because a lot of people donít know about that. I suck at putting things into words so it probably didn't make sence. Yours was reasonably good but what I meant was that during the walk cycle the leg straightened a little ítooí much.

Regarding Ostritch legs. I agree with your palaeontologist, modern day birds have evolved different hips which angle the femur more horizontal. I suggested the Ostritch footage to help show the feet during walking/ running.

As far as Iím aware the S curve in tyrannosaurs isnít particularly strong and from what I have read tyrannosaurs have stiff-ish necks. Precisely how stiff, I donít know.

If you can change the model, Iíd say make the tip of the tail, lower legs and feet thinner. The lower leg consists of mostly tendons hugging the bones and there isnít much fat so that area would be thinner. (Except on the bottom of the feet, which have some fat pads)

Here is a link of possible corrections.

Note the definition of the bones in the eye area wouldn't be as strong as I have illustrated it.

Update: I paid a little more attention to the lower jaw line.


01 January 2009, 04:25 PM
hi all

I updated the walk cycle following your instructions about legs , toes and hands :

new version can be see here :

01 January 2009, 03:39 PM

Thats looking better, especally the feet.

I'd still lower the neck, neck flexibility does vary between the diffferent groups but from what I have read tyrannosaurs don't have particulary flexible necks. I think the main problem is the sharp upturn at the base. You have to image that there are vertebra in there.

Take a look at this skeleton of Gorgosaurus (a very close relative of albertosaurus, it may just be a species of it)

At first glance its neck apears flexible but if you look at the vertebra they are actually disarticulated, due to rigamortis. Most skeletal reconstructions (I'm not talking about meuseum mounts, thay often have problems) of tyrannosaurs have the neck at about 30 to 40 degress from horizontal. How much higher they can go from than that I don't know.


CGTalk Moderation
01 January 2009, 03:39 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.