To close out the end of the winter session, I'll summarize the end results of both the curpigeon and my golem's work status thus far.
As a whole, I've technically built two rigs; the first being the 4 in 1 rig that took the whole of the entire session. The body types involved are the very thin to the fattest of the characters on the curpigeon project. It's not the most perfect rig in the world, as it still needs a low res mesh to be added so that animators won't have to fight with the rig doing calculations, and, unfortunately, some of the controls are stuck being inside the geometry, as the blend shapes made it difficult to move the controls into a better position, but overall the rig is a good success.
The second rig is for the fifth character names long beak, who's overal facial structure couldn't be added onto the multi rig. However, I simply stripped off the geometry from the main rig and reused it, with a bit of re-positioning. Do a bit of weight painting, and that character was said and done.
As to my Golem, I had to set it as more of a secondary project, but some progress was made. The constraint system for the stones was tweaked a bit more, so that the major pieces hold place fairly well, while smaller pieces have more wiggle room. I've also gone with using an IK system on the left leg's roots, so that the whole can be moved, or the roots can stick a bit when the leg is lifted up. I'll need to play with some influence objects relating to the vine curves, as a simple skin bind doesn't work as well as I would like.
Friday, December 20, 2013
Wednesday, December 11, 2013
Progress for the week of November 17th - 24th, 2013
Golem
While using the nDynamics system, It can simularte a spring constraint to get that floating aspect I'm aiming for. However, the points on the geometry tend to expand dramatically if there is rapid movement. Upon further testing, I've found that it works easier to use the stock settings on nCloth mesh, rather than to use one of the presets. I had thought that using the concrete preset would be best, but something in that setting results in the geometry expanding. Currently, stock settings between 2 ncloth objects cause them to phase into one another, and get their vertices stuck. For the time being, I will use the soft leather preset, as it seems to give a fairly decent result. I'll adjust it more as things progress, since I'm still noting some give in the stone's forms.
Curpigeon
More refining and fine tuning on the rig, being sure that things bend reasonably when the bird is flitting about. I ultimately went with a simpler system for the neck setup, using a broken chain system to allow for stretching. Added some small tweaks to the eyebrow controls, having a larger catch all control to move them about.
While using the nDynamics system, It can simularte a spring constraint to get that floating aspect I'm aiming for. However, the points on the geometry tend to expand dramatically if there is rapid movement. Upon further testing, I've found that it works easier to use the stock settings on nCloth mesh, rather than to use one of the presets. I had thought that using the concrete preset would be best, but something in that setting results in the geometry expanding. Currently, stock settings between 2 ncloth objects cause them to phase into one another, and get their vertices stuck. For the time being, I will use the soft leather preset, as it seems to give a fairly decent result. I'll adjust it more as things progress, since I'm still noting some give in the stone's forms.
Curpigeon
More refining and fine tuning on the rig, being sure that things bend reasonably when the bird is flitting about. I ultimately went with a simpler system for the neck setup, using a broken chain system to allow for stretching. Added some small tweaks to the eyebrow controls, having a larger catch all control to move them about.
Sunday, November 17, 2013
Progress for the weeks of November 3rd and November 10th
Curpigeon related:
At the request of the animators, I've retooled the wings on the rig, allowing for a more staggered range of movement. Now, if one were to move the arm control, the elbow and wrist would follow only in translation, rather than rotation. It took several trial and error sessions, and more skin weight adjustments toe get the wings to look passable, compared to it's earlier iteration. I've also added a wing fold control to the shoulder of each wing, though if the controls along the wing are not zeroed, the wings will break.
It has also been agreed on to do the singular rig for multiple birds, though a recent test showed that the other models built so far are not compatible with the base model, preventing the blend shape aspect of the rig from working properly. I may need to find another solution to this quickly.
Golem related:
more modeling work performed, adding in the arm vines network into the body, as well as the support vines for the left shoulder. I've also connected some of the brush nodes onto a simple control system to test the rate that the vines could move. While a nice representation of what I'm aiming for, the vines themselves need to be broken apart within the time of the controller, so a more refined control will be necessary for the construction rig.
The animation aspect of the rig may now have some new foundations, after I had asked one of my professors who had worked on the rock monster in "Galaxy Quest." He commented that the monster was actually a set of props built onto a single rig. The first set was just the big shape of the monster so that the basic movements could be done, verifying that an earlier idea I had for the rig wasn't far off. The second portion of the rock monster rig, which is what I've been struggling to get a good sense of how it would be set up, functions based on a dynamic spring system. The setup goes something along the lines of as follows:
A dynamic spring constraint is set between the joint of the rig system, and that of a particle. This will allow for the particle to slide about and jostle, while still being connected to the rig. From there, I would connect the stones I wish to have in place to the spring bound particle. via a group. From there, I would need to add another control to the stone for manual placement, should the stone stray too far from the course. I will need to test this potential system as well once time allows.
At the request of the animators, I've retooled the wings on the rig, allowing for a more staggered range of movement. Now, if one were to move the arm control, the elbow and wrist would follow only in translation, rather than rotation. It took several trial and error sessions, and more skin weight adjustments toe get the wings to look passable, compared to it's earlier iteration. I've also added a wing fold control to the shoulder of each wing, though if the controls along the wing are not zeroed, the wings will break.
It has also been agreed on to do the singular rig for multiple birds, though a recent test showed that the other models built so far are not compatible with the base model, preventing the blend shape aspect of the rig from working properly. I may need to find another solution to this quickly.
Golem related:
more modeling work performed, adding in the arm vines network into the body, as well as the support vines for the left shoulder. I've also connected some of the brush nodes onto a simple control system to test the rate that the vines could move. While a nice representation of what I'm aiming for, the vines themselves need to be broken apart within the time of the controller, so a more refined control will be necessary for the construction rig.
The animation aspect of the rig may now have some new foundations, after I had asked one of my professors who had worked on the rock monster in "Galaxy Quest." He commented that the monster was actually a set of props built onto a single rig. The first set was just the big shape of the monster so that the basic movements could be done, verifying that an earlier idea I had for the rig wasn't far off. The second portion of the rock monster rig, which is what I've been struggling to get a good sense of how it would be set up, functions based on a dynamic spring system. The setup goes something along the lines of as follows:
A dynamic spring constraint is set between the joint of the rig system, and that of a particle. This will allow for the particle to slide about and jostle, while still being connected to the rig. From there, I would connect the stones I wish to have in place to the spring bound particle. via a group. From there, I would need to add another control to the stone for manual placement, should the stone stray too far from the course. I will need to test this potential system as well once time allows.
Sunday, November 3, 2013
Progress Log for the week of October 27th, 2013
Not much to report on this week. Due to a recent job opportunity, I haven't had as much time to work on the golem or the pigeon. What little I did is as follows:
For the golem, more body structural formation via new vines placed and some joint placement tests. However, due to the abnormal limb sizes, I can't just mirror over joints, so I need to be cautious on how I orient them. I asked Jason Patnode in regards to how I could get an interactive system similar to that of the rock monster from Galaxy Quest. Though he demonstrated using a fractal texture and plugging that into the translation nodes of the stones, it doesn't quite match what I was aiming for. I asked another student in class, and they recommended using a rigid dynamic system, as it would be the only way to attain that free floating, yet interactive look I was aiming for. I'll have to make time to retry this system again, though I know that it will be very heavy on the system. Maybe I can bake in the animations of the stones interacting to help ease the load on the rig.
In terms of the pigeons, it has mostly been cleanup on the wings and neck area. Around Tuesaday, I received notes from the animators, requesting some more controls and deformation fixes. Stepping in to the class on Friday, there was a brief recap on said fixes, and some suggestions, as a scaling control system on the mouth, and a bending control on the top of the beak. Another suggestion, which I'm considering more than the others, is an IK/FK hybrid system for the neck, as well as a more detailed control over the feathers on the wings, as currently, the wings rotate based solely on the wrist and elbow. There was a request for feather bending, but I informed the animators that something like that is low on the priority list at this time.
For the golem, more body structural formation via new vines placed and some joint placement tests. However, due to the abnormal limb sizes, I can't just mirror over joints, so I need to be cautious on how I orient them. I asked Jason Patnode in regards to how I could get an interactive system similar to that of the rock monster from Galaxy Quest. Though he demonstrated using a fractal texture and plugging that into the translation nodes of the stones, it doesn't quite match what I was aiming for. I asked another student in class, and they recommended using a rigid dynamic system, as it would be the only way to attain that free floating, yet interactive look I was aiming for. I'll have to make time to retry this system again, though I know that it will be very heavy on the system. Maybe I can bake in the animations of the stones interacting to help ease the load on the rig.
In terms of the pigeons, it has mostly been cleanup on the wings and neck area. Around Tuesaday, I received notes from the animators, requesting some more controls and deformation fixes. Stepping in to the class on Friday, there was a brief recap on said fixes, and some suggestions, as a scaling control system on the mouth, and a bending control on the top of the beak. Another suggestion, which I'm considering more than the others, is an IK/FK hybrid system for the neck, as well as a more detailed control over the feathers on the wings, as currently, the wings rotate based solely on the wrist and elbow. There was a request for feather bending, but I informed the animators that something like that is low on the priority list at this time.
Monday, October 21, 2013
Log for the weeks of October 7th through October 19th
7th - 12th
With the slow progression of the bird models I focused more efforts on building my golem.
The base structure itself now complete, I've begun to add the neck vines and create a small base plate inside the chest to have the back vines stem from. Most likely, I will modify it to look like the underside of a tree trunk, and place the energy stone at its center, then build outwards to help fill the rest of the chest.
I was introduced to a potential Ivy generator, in the hopes that it could serve to build the golem faster. However, upon obtaining the program, I came to learn that the geometry becomes too dense, and there is no direct control as to where the ivy flows to, rendering it a potential hazard. Were I to use it, it could potentially bog down the system and make the rig neigh unusable, as the poly count could potentially skyrocket.
As to curpigeon, things remained quiet for the bird models, so I continued to refine one of the human characters a little more before moving to other projects over the weekend.
20th - 27th
Curpigeon suddenly became the hot project for me, as I was asked to create a rig for a different character than I had been originally assigned. Monday turned into mostly joint relocation, as I duplicated the joint system from an earlier rig, and set them to the new character.
Tuesday brought about some more coverage on nCloth, which will become something I use in creating the stone collision system on my golem. However, I will need to create a new, lower resolution of the stones in order for it to work, as the current poly count would probably bee to high for it to function efficiently.
Wednesday became a class once again focused on the new pigeon character, placing controls and setting orientations. Nothing too extravagant, though there was talk between Jason, Hector(a fellow rigger on the project), and myself of potentially creating a universal rig. In order for this to work, though, we'd need to be sure that the geometry on the other five birds could work with the most average form of one of the overall 6 bird characters. Still, the main rig needed to be built, and that was my focus for the rest of the day.
Thursday became a mad dash of weight painting and some experiments regarding a broken chain hierarchy, or as a classmate had described it, a ribbon system. This proved quite useful when constructing the tendons on the wings, allowing for some better deformation than just letting the arm and elbow joints control the geometry. It is by no means a perfect solution, as there are still folds in the geometry, but with some controls added, it should work well. Adding a joint system for the feathers is proving more difficult though, but I may have a solution.
The lead rigger on the project sent me a tutorial on the rig setup for the human character faces, recommending that it could be used on the pigeon's face. However, I found it can function more effectively as a joint placement setup for the feather joints. The only real difference is that I would skin the controlling curves to the wing's joints to deform it. With locators acting as place holders, plus some orientation constraints coming from the outer most locators inwards, the joints effectively held place, and the feathers along with them. However, the setup is not ideal, as it causes a double transformation that I didn't have the time to work out. The lead rigger sent an example my way of something he whipped up, using a combination of parenting and orient constraints. With some mild modifications to the setup, I was able to get the feathers to track more accurately.
With the slow progression of the bird models I focused more efforts on building my golem.
The base structure itself now complete, I've begun to add the neck vines and create a small base plate inside the chest to have the back vines stem from. Most likely, I will modify it to look like the underside of a tree trunk, and place the energy stone at its center, then build outwards to help fill the rest of the chest.
I was introduced to a potential Ivy generator, in the hopes that it could serve to build the golem faster. However, upon obtaining the program, I came to learn that the geometry becomes too dense, and there is no direct control as to where the ivy flows to, rendering it a potential hazard. Were I to use it, it could potentially bog down the system and make the rig neigh unusable, as the poly count could potentially skyrocket.
As to curpigeon, things remained quiet for the bird models, so I continued to refine one of the human characters a little more before moving to other projects over the weekend.
20th - 27th
Curpigeon suddenly became the hot project for me, as I was asked to create a rig for a different character than I had been originally assigned. Monday turned into mostly joint relocation, as I duplicated the joint system from an earlier rig, and set them to the new character.
Tuesday brought about some more coverage on nCloth, which will become something I use in creating the stone collision system on my golem. However, I will need to create a new, lower resolution of the stones in order for it to work, as the current poly count would probably bee to high for it to function efficiently.
Wednesday became a class once again focused on the new pigeon character, placing controls and setting orientations. Nothing too extravagant, though there was talk between Jason, Hector(a fellow rigger on the project), and myself of potentially creating a universal rig. In order for this to work, though, we'd need to be sure that the geometry on the other five birds could work with the most average form of one of the overall 6 bird characters. Still, the main rig needed to be built, and that was my focus for the rest of the day.
Thursday became a mad dash of weight painting and some experiments regarding a broken chain hierarchy, or as a classmate had described it, a ribbon system. This proved quite useful when constructing the tendons on the wings, allowing for some better deformation than just letting the arm and elbow joints control the geometry. It is by no means a perfect solution, as there are still folds in the geometry, but with some controls added, it should work well. Adding a joint system for the feathers is proving more difficult though, but I may have a solution.
The lead rigger on the project sent me a tutorial on the rig setup for the human character faces, recommending that it could be used on the pigeon's face. However, I found it can function more effectively as a joint placement setup for the feather joints. The only real difference is that I would skin the controlling curves to the wing's joints to deform it. With locators acting as place holders, plus some orientation constraints coming from the outer most locators inwards, the joints effectively held place, and the feathers along with them. However, the setup is not ideal, as it causes a double transformation that I didn't have the time to work out. The lead rigger sent an example my way of something he whipped up, using a combination of parenting and orient constraints. With some mild modifications to the setup, I was able to get the feathers to track more accurately.
Sunday, October 6, 2013
Log for the week of September 29th, 2013
While waiting for the final version of the pigeon models to be done, I was asked by the producers to help clean up two of the human rigs for the project. With those being more complicated that what I normally build, it took a bit to find my way around the inner workings to fix some of the issues. Beyond that, it consisted a great deal of weight painting.
Next up on the list is a bit of experimenting with fur in maya. I do not know if it will be particularly useful on my golem, but it doesn't mean I'll rule it out for aesthetic purposes.
Regarding the golem itself, I've completed the skull model and have blocked in more of the securing vines on the right leg, and build the back vine cover that will act as an anchor for the left arm.
Next up on the list is a bit of experimenting with fur in maya. I do not know if it will be particularly useful on my golem, but it doesn't mean I'll rule it out for aesthetic purposes.
Regarding the golem itself, I've completed the skull model and have blocked in more of the securing vines on the right leg, and build the back vine cover that will act as an anchor for the left arm.
Sunday, September 29, 2013
Log of September 22nd - September 28th, 2013
While no particular advancements were made in terms of the dynamics systems on the golem, there has been more modeling progress. An initial attempt at reducing the poly count on the stones had some headway, it was thanks to the usage of a newer tool that Maya has incorporated recently into, called Modeling Toolkit. Specifically, a sub tool called the transform constraints, where in you can create a cube with a reasonable subdivision count, and snap the points to the surface of a selected geometry. This will definitely cut down the nasty high numbers of polys, as well as allow the rig to run more easily come time for setup.
In terms of the rest of the modeling, I've near finished the skull of the character, and have decided to restructure the vines of the body to better fit the design.creating the curves based on live geometry, then attaching the paint effects brush afterwards, allows for an easier placement, and can allow for a more interesting layout of the vines.
Meanwhile, on the curpigeon front, my initial experiment with the arm tendon has failed. I had thought of using a squash and stretch spline IK system, similar to another project I had worked on briefly, but the curve involved does not maintain its line of sight with the two controlling joints, bending and flexing about, rather than just compressing as I had hoped.
I'm going to try using an aim constraint system instead to help simplify things a bit more, though I have my doubts about it.
As an aside, I have volunteered to assist in working on some of the human characters for the project as well. I will try to detail what I can regarding the setup, though from what I gather, it'll be more of an insight for me on advanced rigs than me actually fixing things right out the gate.
In terms of the rest of the modeling, I've near finished the skull of the character, and have decided to restructure the vines of the body to better fit the design.creating the curves based on live geometry, then attaching the paint effects brush afterwards, allows for an easier placement, and can allow for a more interesting layout of the vines.
Meanwhile, on the curpigeon front, my initial experiment with the arm tendon has failed. I had thought of using a squash and stretch spline IK system, similar to another project I had worked on briefly, but the curve involved does not maintain its line of sight with the two controlling joints, bending and flexing about, rather than just compressing as I had hoped.
I'm going to try using an aim constraint system instead to help simplify things a bit more, though I have my doubts about it.
As an aside, I have volunteered to assist in working on some of the human characters for the project as well. I will try to detail what I can regarding the setup, though from what I gather, it'll be more of an insight for me on advanced rigs than me actually fixing things right out the gate.
Sunday, September 22, 2013
Log for the week of 15-9-2013 through 21-09-2013
15th of September:
Tested out the paint effects brush in maya 2014 on an older rig. The paint curve stays in contact with the mesh which is a plus. However, leaving it in paint effects mode slows the scene down, so I'll need to figure out a better method of attaching it to the rig. I've also noted that the paint curve upon a mesh is entirely dependent on the geometry's UVs. Can't just paint on top of freshly sculpted geometry without at least doing an automatic unwrap.
Meanwhile, finding good reference for a camel skull is proving difficult, as there isn't a clear view of the underside where the neck bones meet the base. It may never be seen anyway, considering the setup, but I'd still like to at least have a rudimentary form in there. At least for now I have the bony structure where the jaw bone slips into roughed out. Just need to make some basic teeth and the lower jaw and that will be done.
17th - 19th of September
Built a few stones in Zbrush to use as the stones on the Golem. A bit of stretching and scaling, and I was able to block out most of the major stone structures on the character, sans a few points that need a little more tweaking.
On the paint effects front, I managed to figure out the flower controls, so I made something fairly close to what I am aiming for, though I'll still need to make a few pistols for the pupils. I've also tampered with making the vines on the character's back, though its going to take a lot more effort, since the branches have a habit of creating unneeded angles and bends. I may need to use zbrush to at least make a base, then build separate vine structures to use in the anchoring system.
Side note so far: I'll need to contact one of my advisers on how the team at ILM set up the stone creature in Galaxy Quest. I have an idea on how to set up my character, but I'd still like to ask them on their method. I'll do a test on my theory later in the week.
20th & 21st of September, 2013
I decided to do some tests with nCloth in order to try and have the stones collide and shift about. However, my knowledge of ncloth is lacking, so I'm not certain how to attach the ncloth objects to the joint system, let alone keep them from deforming a great deal when they impact one another. One thought that had come to mind is to have a base geo act as markers for the stones, and attach the ncloth stones to them with a semi loose connection, which would allow them to roll over one another more believably. I may attempt this again using regular dynamics, as they do not deform on impact, though I'm not sure of the calculation weight it would require. I'll need to set up dummy geo regardless on a low res level so the rig can run smoothly.
21st of September, 2013.
It may be possible to use regular dynamics for the major stone pieces if I play with nail constrains and spring constraints a bit. Testing on a larger scale is needed, but that'll require a larger, more complete rig than the 3 piece scrap I built to check the theory.
Tested out the paint effects brush in maya 2014 on an older rig. The paint curve stays in contact with the mesh which is a plus. However, leaving it in paint effects mode slows the scene down, so I'll need to figure out a better method of attaching it to the rig. I've also noted that the paint curve upon a mesh is entirely dependent on the geometry's UVs. Can't just paint on top of freshly sculpted geometry without at least doing an automatic unwrap.
Meanwhile, finding good reference for a camel skull is proving difficult, as there isn't a clear view of the underside where the neck bones meet the base. It may never be seen anyway, considering the setup, but I'd still like to at least have a rudimentary form in there. At least for now I have the bony structure where the jaw bone slips into roughed out. Just need to make some basic teeth and the lower jaw and that will be done.
17th - 19th of September
Built a few stones in Zbrush to use as the stones on the Golem. A bit of stretching and scaling, and I was able to block out most of the major stone structures on the character, sans a few points that need a little more tweaking.
On the paint effects front, I managed to figure out the flower controls, so I made something fairly close to what I am aiming for, though I'll still need to make a few pistols for the pupils. I've also tampered with making the vines on the character's back, though its going to take a lot more effort, since the branches have a habit of creating unneeded angles and bends. I may need to use zbrush to at least make a base, then build separate vine structures to use in the anchoring system.
Side note so far: I'll need to contact one of my advisers on how the team at ILM set up the stone creature in Galaxy Quest. I have an idea on how to set up my character, but I'd still like to ask them on their method. I'll do a test on my theory later in the week.
I decided to do some tests with nCloth in order to try and have the stones collide and shift about. However, my knowledge of ncloth is lacking, so I'm not certain how to attach the ncloth objects to the joint system, let alone keep them from deforming a great deal when they impact one another. One thought that had come to mind is to have a base geo act as markers for the stones, and attach the ncloth stones to them with a semi loose connection, which would allow them to roll over one another more believably. I may attempt this again using regular dynamics, as they do not deform on impact, though I'm not sure of the calculation weight it would require. I'll need to set up dummy geo regardless on a low res level so the rig can run smoothly.
21st of September, 2013.
It may be possible to use regular dynamics for the major stone pieces if I play with nail constrains and spring constraints a bit. Testing on a larger scale is needed, but that'll require a larger, more complete rig than the 3 piece scrap I built to check the theory.
Saturday, September 14, 2013
Progress as of the 13th of September, 2013
So as of this post, there isn't too much to report. I've begun to model the skull piece of the Golem in order to have the vines grow up onto it.
I'll need to cut out the back region behind the eye sockets where the jaw bone slides into, as well as create some simple teeth and the lower jaw, though I may just create the sockets for the teeth as well to sell the age of the skull a bit more. In the mean time, I will be exploring the updated paint effects in maya 2014 in order to get a better grasp as to how to generate the growing vine system that will form the "muscles" and various digits of the Golem. I'll also need to locate some flower models to use for the eyes.
The Golem's Start
This is the concept work for the Golem I intend to build from scratch.
Early concept for the golem
Rough sketches depicting some forms and potential ideas on its anatomy
more early work regarding the golem's anatomy
Gem stone mounting concepts and marking design based on Arabic symbol for "God"
Fleshing out how particular limbs may look
A more in-depth look at how its posture may be, as well as its back structure
Idea
on how the golem's left shoulder and left leg would function, where the
vince would attache from below the boulder on the shoulder (top left
image), and how the leg would connect to the tree stump to control the
roots (bottom right)
Basic sketch ideas of the previous idea
How the right arm would be structured
Basic sketch ideas of the previous idea
How the right arm would be structured
concept on how the golem's "eyes" may work
A more definitive form to the golem, this time defining the body further
A more spindly variant of the golem, meant to emphasize a deteriorated body
A bulkier body form, as well as a color palette test
Closer to the final form. Limbs look disheveled, but the design is still lacking
Finalized concept for the golem. Colors work and the body is more balance while being asymmetrical.
Introduction
Hello and good day to you, stranger. I'm Dan Rico, curator of this digital journal for my thesis project, titled "Lingering Creations". The project itself consists of two sub projects, with the first being dubbed "The Golem", and the second is me working with another team on an animation project called "Curpigeon".
Bit about myself. I had gone to study animation at The Art Institute of California-Orange County, graduating with a Bachelors in Science. After that, I was a freelance 3D Generalist for a little over two years before returning to school for my masters. Now, I'm trying to focus on becoming a Character TD, with this blog acting as a recording of any progress made for the my Master's Thesis project.
Subscribe to:
Posts (Atom)