Hi everybody ...

finally I´m back to my home .... and I´m ready to start my task
Updating SVN.

some questions ...

Is the camera untouchable ?
Do I must animate directly in the choreography ?
Do I use "hold" interpolation in all keyframes except lipsync and facial animation ?

Most of the camera options are to be left. Things that aren't specific to the shot (eg render settings). But anything specific, like the focal length can be changed if needed. Though try to keep it to a minimum. Camera position can be changed once you set the keys to hold for cuts.

The chor for everything but dialog or cycles. It makes it easier to see the context. If you need to reduce the clutter in the chor, the H key can be used.

Yes. Though maybe also hold for facial animation too.

Hi SOers ...

I would like to start this message saying that is a pleasure work with you, guys and gals, in SO Project...
The Project TWO gave me confidence to put in practice an old personnal project, and the project SO is motivating me even more.

Take #1 - frames 0/290 - Trot

You are off to a great start Marcos.

I agree the camera angle will need to be refined but as far as blocking out and working the first pass (you didn't specify what your focus is on here) its looking fine.

The dialogue seems a bit fast (as in she should breath a bit more in there) but if that is what you have to work with. Fight on!

P.S. Not sure who is doing the voice of Trot there but... I like!

Hi Marcos,
If you are still working on this shot, please move the boat up a little so it is not filled with water. I think someone will eventually have give the boat a double hull so it can appear to sit in the water without the water plane appearing inside the boat. Also please make sure Cap'n Bill's hands stay on the oars and give him a good pose at the extremes of each rowing action.

If you are done with it, that's fine. The changes I suggested can be done in the animation or cleanup stage. If you are done, please upload it to SVN.

Thanks!

A double hull would still do the same! I think we're going to have to get creative with the camera angles. Unless we go for some sort of displacement rig.

I wonder if Steffen's "Resurface" plugin might work.

That would be pretty cool if it did, but I'm pretty sure someone would have to beef up the mesh resolution of the water plane around the boat quite a bit.

A double hull would be simpler. Then you could just raise the inner bottom of the boat a little above the water and the outer hull would make it look like the boat is still sitting in the water.

How about modeling the water as attached to the boat ? & moving the water line surrounding the boat as well as rocking the boat. The camera could move (as well as the boat/water model) - and any backgrounds?

In my simple "proof of concept" example the boat is modeled with the hull below sea level. There is one bone to rock the boat, & 1 bone to rock the spline ring surrounding the boat - could add sprites for a bow wake, make better wave motion, model more splines for waves, add animated water material

Good idea Mark. Experimentation is needed.

Raising the bottom might do it actually, but it could make their knees go up to their faces.

That's another good one Nancy. Though the whirlpool is set in an ocean mesh already so I'm not sure we can rebuild it.

No need to mess with the whirlpool model - the boat with it's attached water is separate from the whirpool model - when it comes time to go thru the whirlpool - can have a pose to hide boat's attached surrounding water (or switch to the boat model that doesn't have water), while simultaneously the separate already existing whirlpool model becomes visible. And of course this switching can be aided by judicious use of camera angles, and cuts, if necessary.

Nancy's idea is cool but why not just have a boolean cut out of the boat in the water model. Then "translate to" and "orient like" the boat control bone to the cutout? Then you move and animate the cutout bone in the water.


Here's a quick thing I threw together as an example.

Click to view attachment

Here's the prj file.

Click to view attachment


George

choices, choices

took svn rowboat boat small - added bones, surrounding water, placed ontop of a water ground plane (proxy whirlpool), increased transparency of boat's water to 100% when sinking

I don't think you are understanding. I don't mind if you disagree, but it is frustrating to be misunderstood. I am attaching some pictures of what I mean. I went ahead and added an external hull to the boat. It shouldn't affect any of the characters' existing poses because the external hull was added to the outside of the boat. If it causes problems you can just revert the boat to its previous state.

I really like Nancy's idea because making water ripples for the boat would be much easier if a small water plane were attached to the boat via an action. Then anywhere the boat goes, the ripples will go with it. The only problem is that whoever makes the storm waves will probably use an animated displacement map for that, and I'm not sure how easy it would be to make the small water plane attached to the boat conform to the larger displaced water plane. Maybe Stefan's plugin could be used??

George's Boolean solution would work well if the water plane and the boat were in the same model - I'm pretty sure boolean cutters only work within a single model and do not interact between different models in a chor. Since the boat and the water plane are different models, and we do not know yet how a diaplacement map will affect the water's contour, I don't think a boolean cutter would work well in this case. But George if you want to make some storm waves and give it a shot (using different models for the boat and the water plane), I would be interested in seeing the results.

That's why I said a double hull might actually do it. I thought you'd add it on the inside to preserve the curve shape on the outside.....hence raising the characters feet and knees (not a huge change). But that doesn't look bad at all now I see it. The simplest and best solution I would say.

Edit: I wonder if the waves on the ocean model would still be hidden though. (I just checked and most of the waves should be lower than the inner hull. They're so far apart that we can use move the boat up if any do intersect.)

Edit2: I do like the boat movement in Nancys idea too. Maybe we can add those bones to the boat which ever way we go.

The only thing that needs to be with the water model is the cutout boolean. The boat does not need to be part of the model. All you need to do is keep a cutout in the water, Translate and orient the boat to the cutout. When you don't need the cutout, you just move it out of camera.

The only down side is that the boat cutout doesn't effect the water. It essentially cuts a hole in the water. It may not look too good. Animating the water would look better. I just don't understand how Nancy's idea works completely.

George

I've incorporated a small surround of water plane on the boat model (can be any size, any number of rings) - starting with using an existing spline on the boat. The model has a "hole" in the water where the boat sits. The water and boat share a spline. The whole model then sits on top of a separate waterplane/whirlpool model (if needed) but is raised above so boat doesn't penetrate - it is an illusion, helped by camera angle, transparency, texturing, and perhaps modeling small water plane so that outside spline ring dips down to meet real water plane.

I used a bone to rock the boat geometry. That bone can roll, pitch yaw, as well as translate. The next spline ring of water (extruded from the starting spline on the boat) is animated as well (currently just has 1 bone assigned to that spline ring). It too has 6 degrees of freedom.

That water spline ring can be manipulated more elegantly and made more complex. ie bones radiating out from center of boat to each cp - to do isolated cp wave/ripple/wake action). Or could just fake with moving the cps.

The model also has a pose that sets the transparency value of the water group.

It doesn't have to be an all purpose, perfect, water/wave model simulation solution - it just has to work in whatever scenes it's required, and accomplished by whatever chicanery necessary.

Whatever works is ok by me.

Hi friends....

I had a busy week and only today I could sit down and work on the scene...
I lost the discussion about the boat and water ..sorry ... I just fix the Z axis in the boat and it seems to fix the water inside it..


Take #2

I'm on the road again .....
my notebook, the night, a dark room and the truck´s noise ... these are my companions to animate the SO´ scenes

I'm redoing the first Trot´s dialog ( lipsync/block )

take #3

Getting there. Their arms just seem a little busy to me. I think Bill is rowing through all this so he can't use his arms at all!
I love the water mark you put on it. How'd you do that?
Keep it up!

I think he can get away with taking his hands off the oars to make the shot interesting, (lots of talking). Putting them back on may be a little difficult to animate... Nice time to learn how to use the FK/IK switch.

I used Adobe Premiere to insert them




I deleted Row Boat's action, changed the the Row Boat Small's rig and rename it to Row_boat_small_xtaz... before, the hands of Bill followed the oars .. Now, like in the real life, are the oars that follow his hands

take#4 is just a test to experiment the rig.. I liked it .. Now I needn't worry about IK / FK, but the percentages under the oars' constraint.

Looking cleaner alright. The row strokes seem abit half hearted though and the oars are hardly entering the water. Also, when you rotate the head and chest bones, more times than not, you'll need to move them too to make it look natural.

I asked for Martin and he allowed me to work the in-between in this scene.

Take #5 ( emphasis on Bill.. Ill work in the Trot tonight )

Is the camera moving around the boat or is it stationary and the boat is move to and past it? If the later then there is too much movement in too short a time for the amount of rowing effort seen.

The rowing action is too uniform. It looks like he is rowing through air and not water. The action should have a 'catch' - the oar entering the water, the action will immediately slow then build to the end of the stroke when the oar exits the water, halts momentarily then return to the smooth action leading to the catch.

Next the hands should be closer to the end of the oar handles and during the crossover, one will be above the other. To me the oars look too short and he is having to dig the oars in at shoulder height making the whole thing look out of proportion.

Something common in rowing is the amount of time the rower actually spends looking at the oars. You might want the Capt looking at his oars more then every now and then - on the key words - looking back at Trot.

(Sorry if I appear a little pedantic - I used to row competitively, pairs, 4's, and 8's)

The animation in general is pretty floaty at the moment. I imagine that it will sharpen up in the next pass!

Cheers

David said it better than I could.

I made another water test for this shot. This one doesn't have any fancy stuff. It is a grid with a bunch of bones controlled by four master bones. I think with textures and camera cuts, it may look OK.

I can't figure out how to keep the boat on the surface though. I tried Newton, but I don't think Newton Fluidsurface was made to make things float on deforming meshes. I can't think of any other solution than to just hand animate it. Any other ideas?

I think the stormy waves look pretty good.

To keep it on the surface - perhaps a system of constraining the small water plane attached to the boat could work (add bones to small water plane). But to me, just manually animating would work - it just has to work for this specific situation and specific camera angles ? right? - animate the ocean first and camera to look good, then animate the boat to look good. Using path? if you have to - or not. No? Too simple ?

You can constrain the boat to a spline which is animating - the wave?

I think a surface constraint would work best.

Rig the boat for the surface constraint (geometry bones and targets). Setup the constraints in the chor and drag the model bone below the water plane. The model should stay afloat. Then all you would need to do is animate the model bone of the boat from point A to point B or use a path constraint (path below water plane).

If you remember my car suspension rig on the monster truck, that's how I set that up to follow the terrain.

Here's a quick test of the surface constraints. You can also animate the offsets of the constraint to make the boat bob up and down.

Impressive. Finally a use for the surface constraint!

Bravo!

That was my first idea, but I can't figure out how to use a Path Constraint to constrain the boat to a spline on the ocean mesh. You used to be able to choose the target by clicking directly on the desired spline in muscle mode, but now the only thing I can click on is a path that was created in the chor.

I remember something about holding down shift or control when selecting a spline in a model.

Thanks Ken, that did the trick. The Path Constraint seems to be the solution that works the best. I had to have Particles ON for it to work with a dynamically deforming mesh though. <Note to self: Remember That!>

I tried the surface constraint too, and it worked OK, but the boat would get flooded for a few frames every time it hit a swell in the ocean.

Now to incorporate a giant whirlpool into it. I'm afraid the camera work on this shot is going to have to be completely redone

Here's another attempt with the surface constraint. The water plane is rigged and animated with expressions.

Wow, I love seeing a plan come together.

That looks really good!

Nice one Mark!

Here's a test for the whirlpool in the storm waves.

wow... what can i say... i´m impressed...

very good

It's amazing Holmes ....
Was it ( choreography ) uploaded to SVN ???

I'm on the road since last monday then I had to copy ( last sunday ) the choreography from my PC to my eee ( no svn installed ) ..
Do I continue animating in "my" cho ???

excuse me for the lack of updates. I had very little free time in these 2 weeks...

Take #6

Thanks Marcos, your animation looks great so far.
I have created a different choreography, so you don't have to worry about svn conflicts.
When you have finished, I will try to import Bill and Trot from your chor into this chor. I will try to use your camera too, but it may not work.
I had to make many changes to the camera because of the way the boat moves on the waves.

This scene is truly going to be a co-operative project.

Here is a draft of the full length scene. I'll try to use Marcos' camera if I can.

http://project.hash.com/movie/so/active/mo...c01/1_03_01.mov
(5.7MB)

Here's a test where the Ork is gererating a large cloud behind it as it flies.

umm..er..tis a mighty big wind.

What seems strange to me is that the cloud (presumably a moisture cloud) is coming from the ork (starts out high) - rather than appear as if water/moisture were being stirred up from the water as he flies over it, also causing turbulence in the water. I wouldn't think that the cloud would get bigger in time. But, yes of course appear bigger as it gets closer to the camera. I would think the cloud would maintain a relatively constant volume, and perhaps would be more watery droppy, swirly looking.

I like very much how one can see the cloud/turbulence thru the water and how the waves look.

(Is the ork supposed to be ...how shall I say this...passing gas? Is this a matter of artistic license ?)


EDIT: videos of helicopter over water:

helicopter rescue of boat people - churning of water starts at about 2:47, good churning is at 3:45

http://www.youtube.com/watch?v=I08M6gfnwDk...feature=related

Also another video - shorter with more dramatic churning - unfortunately its a helicopter sinking into the water

http://www.youtube.com/watch?v=UjEpGchUXKQ

I see the video of the helicopter, but I'm afraid the helicopter-blades-whipping-up-the-water-below thing is just not going to happen if I want to get my animation assignment done by this Thu. All the files are up on SVN though, if someone else wants to give it a shot. The Chor in question is named "1_03_01_water.cho"

If you can draw a series of pictures to show what you want it to look like over time, or give me some more clues about how to accomplish what you have in mind, I'll give it a shot. Otherwise, we can just delete the particle emitter.


I guess you have never actually seen an Ork in flight. Everyone knows that Orks fart clouds when they fly.... You really need to get out more.

Unfortunately the ankle bracelet will only let me go so far.



If one were to simulate the rotor effect of a helicopter on water, one method to try might be:

1) Have your sprite emitter (using particles you have now) travel at water level, and rotate around the y axis as it moves in it's forward direction. The Ork (if it's to be revealed in this scene) would be traveling in the same horizontal direction above the emitter, but could vary in vertical height above the emitter. The rate of emission of particles would increase as orkboy gets closer to the surface of the water.

The particle emitter could be another water plane - much smaller, but similarly setup as the ocean to produce mini interferring choppy waves - at a higher frequency than the large swells of the surrounding ocean, and that they would have some transparency.

OR

2) If we are premiering the first Ork in flight over water, and we count on the fact that the audience don't know nada about no dumb Orks, & given you are the acknowledged world's foremost authority ORK researcher expert, then perhaps, the easiest is to just bring your current Ork fart cloud emitter down closer to the ocean level and have it intersect the ocean surface the whole way as it approaches the camera, and make sure the particle death amount maintains a volume that is the clue as to how close to the surface this gilded lizard buzzard is flying.

You still probably should rotate the emitter as it moves forward to get a swirling, rotating trail effect (ie, have the emitter model be on a path, than is a circle, or spiral - and rotate the path as the path moves forward)