http://www.lowrestv.com/misc_movies/woman-head-rig2.mov
Ressurected this old file and converted to my 2.5D head rig. There was a brief glimpse of this in my AS Demo video but I hadn't worked out the hair yet.
So now the hair is on 3 layers. One over the face. One under the face but over the "head back". And a final one BEHIND the head and body. It was that LAST one that caused no end of trouble. The hair has to be in the head rig which is over the neck and body. So LONG hair "poked" through her head over the neck.
I have to tweak my custom script so I can "duplicate" a bone movement from the head to the body bone layer to move the back hair layer like the head layer. It isn't working 100% in this sample. Need to "force" keys on it.
-vern
My old "Toon Girl" character has a 2.5D head
Moderators: Víctor Paredes, Belgarath, slowtiger
-
patricia3d
- Posts: 726
- Joined: Thu May 31, 2007 4:38 am
- Location: India
- Contact:
-
b15fliptop
- Posts: 167
- Joined: Thu Mar 10, 2005 6:11 pm
- Location: Parts Unknown
-
Barry Baker
- Posts: 342
- Joined: Fri Aug 27, 2004 6:58 am
- Location: UK
- Contact:
If you have AS Pro 6 there is a sample of this rig in the VernonZehr folder in the content library (yes they left the space out between my first and last name 
).
This is all done with bones. There is no point motion except for a "set up" frame on frame 1. I use frame 1 to adjust the points to match the character. This allows me to use the same rig without having to make a ton of changes to the 200+ bones.
There is also a custom script I created that converts a bunch of bone motion. For example translations are converted to rotations etc etc.
No switch layers, no point morphing... all bones.
The lip sync is "maintained" during the turns because of "compound" bones (there's a tutorial for this on www.lowrestv.com). Compound bones are what makes the whole thing work. They are used on nearly every aspect of the rig.
"Compound Bones" are simply identical bones (same length, position, strength) on top of each other over the same points. Each set of duplicate bones are constrained to DIFFERENT bones and each share control over those points in varying degrees.
So the mouth has 3 sets of identical bones, two for the lip sync and one for the head turn (I think... there may be more). Basically there is a set of bones for the whole mouth (upper/lower lips), and a set for each side of the mouth separately (upper left/right, lower left/right) and finally a third set that controls the turn. Since all of these bones are controlling the points at the same time creating mouth poses on top of head turns blends together with no effort. So the lip sync bones are constrained to the mouth control/s and the bones that control the turn are constrained to the turn bone/s.
With 3 sets of bones all fighting for control at the same time the bone constraint values have to be very high in order to have the desired effect. So many of the mouth bones shoot way off the screen in opposite directions. I usually just turn off construction curves entirely while animating or hide controlled bones.
--------
Layers:
The layers are also a key part of the rig. There is a the main "face" layers that include the leyes, lips, mouth, teeth, ears and the "face" which is just the face skin. Then there is a "back head" layer under everything. This back head layer is mostly bound to one bone that has a tiny movement as the head turns. It's very subtle but completes the "effect".
One new thing added to this rig is an addition to the script and extra hair layers. There are 3 hair layers, one in front of everything, one behind the face layers and a final one behind the BODY which of course is OUTSIDE of the head bone layer.
This required some additional code added to the custom script that transfers the "subtle" bone motion of the back head bone to a matching bone in the parent body bone layer.
In reality all the head layers could be in just one vector layer... but this would be CRAZY hard to edit. All those points on one layer are very hard to adjust. The only reason I have all the head parts on separate layers is for adjusting points when developing the characters.
-vern
).This is all done with bones. There is no point motion except for a "set up" frame on frame 1. I use frame 1 to adjust the points to match the character. This allows me to use the same rig without having to make a ton of changes to the 200+ bones.
There is also a custom script I created that converts a bunch of bone motion. For example translations are converted to rotations etc etc.
No switch layers, no point morphing... all bones.
The lip sync is "maintained" during the turns because of "compound" bones (there's a tutorial for this on www.lowrestv.com). Compound bones are what makes the whole thing work. They are used on nearly every aspect of the rig.
"Compound Bones" are simply identical bones (same length, position, strength) on top of each other over the same points. Each set of duplicate bones are constrained to DIFFERENT bones and each share control over those points in varying degrees.
So the mouth has 3 sets of identical bones, two for the lip sync and one for the head turn (I think... there may be more). Basically there is a set of bones for the whole mouth (upper/lower lips), and a set for each side of the mouth separately (upper left/right, lower left/right) and finally a third set that controls the turn. Since all of these bones are controlling the points at the same time creating mouth poses on top of head turns blends together with no effort. So the lip sync bones are constrained to the mouth control/s and the bones that control the turn are constrained to the turn bone/s.
With 3 sets of bones all fighting for control at the same time the bone constraint values have to be very high in order to have the desired effect. So many of the mouth bones shoot way off the screen in opposite directions.
I usually just turn off construction curves entirely while animating or hide controlled bones.--------
Layers:
The layers are also a key part of the rig. There is a the main "face" layers that include the leyes, lips, mouth, teeth, ears and the "face" which is just the face skin. Then there is a "back head" layer under everything. This back head layer is mostly bound to one bone that has a tiny movement as the head turns. It's very subtle but completes the "effect".
One new thing added to this rig is an addition to the script and extra hair layers. There are 3 hair layers, one in front of everything, one behind the face layers and a final one behind the BODY which of course is OUTSIDE of the head bone layer.
This required some additional code added to the custom script that transfers the "subtle" bone motion of the back head bone to a matching bone in the parent body bone layer.
In reality all the head layers could be in just one vector layer... but this would be CRAZY hard to edit. All those points on one layer are very hard to adjust. The only reason I have all the head parts on separate layers is for adjusting points when developing the characters.
-vern
-
Barry Baker
- Posts: 342
- Joined: Fri Aug 27, 2004 6:58 am
- Location: UK
- Contact:
Thanks for that detailed description. I am studying the "Technician" example and I also watched your video tutorial today.
I think your technique is absolutely brilliant - I'm not sure yet that it is not just too brilliant for its own good! The amount of time it must take to build a character head can surely only be justified in a long film, or perhaps a TV series. It still looks very daunting to me to plan it out, but I'll have to have a shot at it myself sometime.
I think your technique is absolutely brilliant - I'm not sure yet that it is not just too brilliant for its own good! The amount of time it must take to build a character head can surely only be justified in a long film, or perhaps a TV series. It still looks very daunting to me to plan it out, but I'll have to have a shot at it myself sometime.
Yes it is time consuming. The FIRST version took a very long time. However I have gotten much FASTER on each subsequent character.
The basic rig is in place. I tweak a few bones, move the points on frame 1 to match the character I am working on at the time so the process gets faster and faster. I was able to do a totally brand new unique character recently that was intended to look like a celebrity. I did this in a fraction of the time required for the second or even the third character I applied this technique to.
The toon girl character in this topic was the started some time ago and put aside because I could not figure out how to do the hair. What I discovered was that all I had to do was drop the hair in the bone layer and the bones already in place took care of everything with little effort. I was surprised! And a little shocked!
My current goals is finishing up my "snout" rig for talking animals. I have several characters I'm working on with that rig. This rig is very different and more difficult. The "mouth" must behave more like a "nose". It can't just "float" over the face and requires a much wider range of motion.
This rig... is just so incredibly versatile and continues to improve for each "version". The last one I did turned out better than all of them before it. Instead of a "head" with limited "switch layers" for mouth shapes and views you have a virtually unlimited source of poses and combination of poses.
I have "long term" goals for this technique and all of the characters created with this rig including sales of characters like the "Alice" character above, which is what I named her, for public purchase at a fairly reasonable price. So often you see characters for sale that have very limited views and facial expressions. They are fine and do the job. With a character like this though... I think it raises the bar of what to expect from Anime Studio.
I don't think there is anything like this for Flash or other 2D applications.
-vern
The basic rig is in place. I tweak a few bones, move the points on frame 1 to match the character I am working on at the time so the process gets faster and faster. I was able to do a totally brand new unique character recently that was intended to look like a celebrity. I did this in a fraction of the time required for the second or even the third character I applied this technique to.
The toon girl character in this topic was the started some time ago and put aside because I could not figure out how to do the hair. What I discovered was that all I had to do was drop the hair in the bone layer and the bones already in place took care of everything with little effort. I was surprised! And a little shocked!
My current goals is finishing up my "snout" rig for talking animals. I have several characters I'm working on with that rig. This rig is very different and more difficult. The "mouth" must behave more like a "nose". It can't just "float" over the face and requires a much wider range of motion.
This rig... is just so incredibly versatile and continues to improve for each "version". The last one I did turned out better than all of them before it. Instead of a "head" with limited "switch layers" for mouth shapes and views you have a virtually unlimited source of poses and combination of poses.
I have "long term" goals for this technique and all of the characters created with this rig including sales of characters like the "Alice" character above, which is what I named her, for public purchase at a fairly reasonable price. So often you see characters for sale that have very limited views and facial expressions. They are fine and do the job. With a character like this though... I think it raises the bar of what to expect from Anime Studio.
I don't think there is anything like this for Flash or other 2D applications.
-vern