itsjustme

Ideally, you'd like to have a front and side view...those are hard to get sometimes. You can do it with a 3/4 view like this, but it takes a whole lot more guessing. With this view you can get the height of things in relation to each other (eyes, ears, mouth, etc) by using it as a rotoscope and then putting markers in the modeling window at each significant feature. If you can't get a front and side view, I would get the front and side view of someone with features that are similar (if they resemble someone famous there'll be a lot more available pictures), make the head based on that and then tweak it to resemble the person that you're really trying to model...it would eliminate some of the guessing. That would be my solution. If you're asking about modeling using a rotoscope, I would recommend Colin's tutorial at http://www.colins-loft.net/new_coop_tutorial/tutorials.html

I'd go with David Higgins' assessment...I can't really add anything except to say that I only had four votes, it was very tough.

You can render Targa images. In the "Output" tab on the "Render to File" menu select "Targa sequence" then in the "Range" pull down select the frame range you want to render out. You can convert the Targas to whatever file type you need in a paint program or Irfanview or whatever.

Congrats to amarillospider and a thanks to Rodney and Hash for the drawing!

There's some cheek movement, but not enough. You're right, I need to have some more going on. I'm doing some experimenting with the face rig and increasing some of the poses. I'll try to have something new up in a few days. Thanks for the crit.

If you need to make some quick iris' for the eye, I have a Script-Fu for GIMP. I tried to get it posted on the GIMP Users Group site, but it never got listed for some reason. I'll post it here for anyone that might want it as well. To use the script (after unzipping the file), put it (irismap.scm) in the scripts folder in the GIMP directory...on Windows XP it would be something like "c:\Program Files\GIMP-2.0\share\gimp\2.0\scripts". Once the script is in the correct folder, you can start GIMP and use the script by going to "Xtns\Script-Fu\Misc\irismap". On the menu that comes up, all of the listed variables can be edited by clicking on them. The default settings make one brown iris that is 400 pixels wide and 400 pixels in height with a background color of a version of purple (200, 0, 200). If you uncheck the "flattenlayers" button, the image will have twelve separate layers (so you can tweak individual layers before "flattening" the image). If you want to save your original iris with its' layers separated, you'll want to make it an XCF file. To save the file as one layer for using in Animation Master, you'll have to "flatten" the image by using the "Flatten Image" tool in the image's "Layers" menu. GIMP will use the color you have selected as the background color in your color picker as the background for the flattened image. Using this script you can only get an oval iris and pupil, but they don't have to be perfect circles. The smallest overall iris size this script can make is 400x400 with a minimum pupil size of 125x125...the maximum overall size and pupil size is 9999x9999 at the moment but that was an arbitrary limit I put on it, you can change that size by changing the setting in the Script-Fu if you have a need for a larger iris. Once I get a website up I'll have a home for this script, for now, here it is. ------------------------------------------ Edit ------------------------------------------ This script works with GIMP 2.0.4, and probably any other GIMP 2.0 versions. For a 2.2 version, check later in this thread.

He's looking better, you could probably still get rid of the spline marked in red on the attached image. The high patch count on the eyes is because of the eye tutorial I made...in the tutorial I used a high patch count to easily determine the size of the iris and to minimize the distortion of the eye when dilating the pupil. It's a quick easy way to make an eye and the hit you would take on rendering time is small. A more elegant solution would be Colin's Uber Eye, mine is more of a brute force solution but it works for me. As for the flipping of normals, flip the sclera, cornea and iris normals to point outward and the retina normals to point inward.

Nice start Gerry! From what I can see, I would now go through and start reducing the splineage. I think you could cut the number of splines on the lower lip by around half...I would get rid of the splines immediately to the right and left of the center of the lower lip to start and end some of the splines on the nose in hooks or eliminate some completely so that you don't have quite as much on the upper lip below the nose.

I was looking at it a little wrong initially, but, I would've continued the splines the way this illustrates in red. The green circles indicate the hooks that are causing the flattening. Porcelain (or however it's spelled) would get rid of most of your problems...just set it at "0" with the material applied. If porcelain doesn't cure the flattening problem, you'll have to move a few splines around.

What might help with the torso is to have the hooks that are connected to the patches next to the five pointers replaced with a control point and continue those splines down to where you have the hooks at the hips...making it a continuous spline instead of the hooks at the waist as well (I hope I made sense). The hooks tend to cause some flattening. The lines on the crotch and legs look like a shaded mode render of a model with porcelin applied...or is this a final render? If it's not a final render, see if it shows up when you render in that mode. Hope this helps.

Thanks for the input, guys. If the tongue weren't visible because of Bertram's missing tooth I might've been able to cheat a little, but, it's good for me to practice that as well. I'm doing some tweaking to the rig, increasing some of the pose percentages and some general experimentation...I should have something new posted in a few days. I'm glad you found the eye tutorial useful, Gerry. I don't get much in the way of feedback on it, so it's cool when I find out it helped someone. When I get things finalized on Bertram I'll definitely show my work on that as well...I've got a pretty involved animation planned for him.

Thanks for the crit Matt, I'll have another pass at it to see if I can exaggerate a few things.

Okay, after a couple more passes here's what I've come up with...I think it's a little better. Crits are welcome...zipped up DivX encoded AVI again.

I think I'm #114....not sure if it matters, but, just in case.

If I'm eligible, I'd like to be included in the drawing.

The images I found of a replica of the Hunley show some big rivets, but the hatches are more flattened. Let's see if I can link the image. I found it in the gallery of the site located here: http://www.charlestonillustrated.com/hunley/# The rivets look accurate to me. Very nice model so far and very generous for donating when you're finished.

Thanks for the input robcat, I'll have another pass or two at it and see if I can make it more natural.

I like these! Very nice, lots of character.

I went back and changed a facial expression in the last three seconds...I think it's a little better now. DivX encoded AVI once again.

It looks like the new file downloads fine, I guess I came too close to the limit...the file was originally 930k, there may be some overhead that I'm not aware of. If anyone downloaded the bum file, my apologies. It should work now.

After rigging up one of my character's (his name is Bertram) face using a slightly modified version of Victor Navone's face rig (thanks for sharing Victor and thanks to Shaun Freeman for providing a downloadable example), I wanted to see if the changes I made would work before moving on to other things. At first I just did a bunch of facial expressions, but I found that my timing needs loads of work. I decided to try a lip sync test so that I'd have a guide to go by. I've used this as a way to try to get better at animating, feel free to C&C I'm sure I could use it. Anyway, here's a final render of the test that I thought I'd throw out since so many others have shared theirs recently. Only the face itself is animated, there is no body or head movement...it's not much. Initially it was a Quicktime zipped up, but, I was having trouble getting the file to download, so I decreased the size by making it a zipped up DivX file.

What I'm "working around" is my own ignorance...my sentence structure isn't always the best. I'll try to be more clear in the future.

My ignorance knows no bounds sometimes, I misunderstood. I'll try to work around it...and look a lot harder next time I hit a wall. Thanks for clearing it up, the problem was giving me fits...and apparently, blinding me. Sorry to be a bother.

Where the problem comes in is when you are trying to set a key in an action or choreography not when making the constraint. Everything appears to work fine until I get the model into an action or choreography, then I partially rotate the bone and set a key for the frame...that's the key button I'm talking about. If you set a key on the frame of an action or choreography after partially rotating the bone, the bone being controlled jumps. Then if you keep hitting the button to set the key, it jumps more each time for the next couple of times you hit it. The first action has a model in it that has a "rotate like" constraint set to 50%, the second action has a kinematic constraint set to 50%, both are driven by the rotation of a bone. If the constraints are set to 100% this problem doesn't occur, anything lower and it does. I'll have to double-check the kinematic constraint for what you mentioned, maybe I missed something on that. Edit---The model in the second action didn't have the euler limits on it correctly...it should be fixed now. Here it is again as well.

The first action's model has a "roll like" constraint that does this...I forgot to mention that it only does this if the constraint is set to anything less than 100%. The second action's model is set with a kinematic constraint on "bone1" of 50% to the rotator bone. It also only exhibits this behavior when the constraint is set to less than 100%.