Saturday, 27 February 2010

Week 4 - Constructing/Detailing the Torso

I was now to move onto the creation of Manny's torso. With his torso, I was aiming to create a fairly accurate rendition with as smaller number of Polygons as I could, so as to appear close to a real-life example without being overly complex - which would only lead to problems when attempting to apply with UVW Maps and Skinning later on


Starting off with a Box 6x6x6 H/W/D segments, I proceeded to select the four corner Polygons that would part of the beginning of Manny's front shoulder/clavicle area. I knew I needed to make use of all four viewports when adjusting these, so I made sure the viewport layout was as shown here

  

These Polys were subsequently move downwards to imitate the shape of a male's lower shoulder, going towards his pectorial area, while leaving room to repeat this for the rearward shoulder area and nape of the neck

  

A similar method was applied to the four Polygons nearest his waist/hips - these were brought inwards with the use of the Scale tool

  

I then needed to choose an area of four Polys situated in the centre of the upper torso surface which could be extruded for Manny's neck. Notice how, in this screen, there is the gentle slope of the shoulders either side of the ridge that is going right between the selection I have made here - this should be more pronounced when the MeshSmooth is applied

  

I used the Bevel tool to allow the Polygons to open outwards slightly when the desired height had been reached

  

 The same neck area (facing forwards) with a simple MeshSmooth modifier applied. Notice how the neck looks too streamlined and narrow across the depth of the torso? This will be adjusted to compensate for Manny's Head (which, incidentally, will be bridged onto his torso next week)

 

  In order to obtain a similar level of detail to the rear of Manny's torso, I took it upon myself to use a technique whereby selecting alternate Polys and Edges can allow me to manipulate the muscle structure and defined appearance, but avoiding a complex, multi-polygonal mesh that would be excessively overdeveloped for the purposes of this module

  

Here, I am selecting opposing edges and pushing them along both the X and Z axis, inwards and upwards to acheive the common definition found on the majority of male mannequins front torso - pectorials and laterals (I frequently referred back to my reference images taken of real-life mannequins to maintain artisitc accuracy!)

  

This was the start of (attempting) to re-create a moderately-defined stomach area, modelling a series of small adjustments with the use of the Cut tool to map out alternate Polys if I wasn't having any success with the ones pre-determined by the Primative Box Object

  

The use of the Rotate tool was important; I could gather various edges and, by rotating them thorugh the X axis, the appearance of the pecs and other features become more pronounced, without me needing to revert into creating more and more vertices and, potentially, causing alot of hassle for future changes that could always be made unexpectedly

  

My first series of adjustments rendered: it is clear where further manipulation needs to be made (more detailing across the stomach, pecs and definition for the clavicle and lats)


I decided to try the chamfer tool; it creates two edges out of one and can be used to extensively deepen a models formal structure, and surely a vote for improvement of Manny's muscle definiition!


Here, I am incorporating the Chamfer tool to be used across the lateral definition, adjacent to the central area, which I have tried to improve by raising and lowering the protrusion of each of the edges

 

A render of the result of this idea: sadly, not as immeditately effective as I was hoping for, but its clear how effective this tool is for lateral edges, perhaps rather than longitudinal ones


Approaching this from a renewed perspective, I then commeced with the removal of the upper Poly's, so that I can attach Manny's head at a later date. Also, notice the inclusion of a Chamfered edge at the foreward nape of the neck

  


  

These two screens demonstrate the varying degrees of trail-and-error needed to manipulate and modify various Polygons, Vertices and Edges in order to achieve the definition I had staring back at me from my reference images - the Chamfer tool was used to enhance certain edges and Rotate tool was again used to add futher detail to the laterals

 

The finished product - yes, I may have to add more at a later date, but not only do I not have the time to keep returning back the the same items again and again, perhaps it's a godsend that the module doesn't specify that character models need to be extensively detailed and finely modelled in every minute area...all I can say is, I'm happy with this right now!


Week 3 - Creating the Hands and Arms (using techniques such as 'Bridging', 'Attaching' and 'Bevelling')

At this stage, the arms and hands had been created using primitive objects; an extruded cylinder for the arm (with three 'mini' extrusions where the upper-arm meets the fore-arm, and at the wrist), and a box object for the hand (selecting various vertices to pull it into a common shape, and then using the bevel tool on the four exterior-facing polygons at one end of the hand to generate the fingers, rotating each Poly slightly away from the next to ensure equal spacing)
The thumb was created in a similar manner, but this time selecting a Poly to the upper right of the hand closest to the wrist. Rather than using the Extrude tool, I wanted to create the ovalised-shape of the end of the thumb, and this was achieved by bevelling out the final section to get a wider, flatter shape





Before joining the hand to the arm, I noticed that I would need to adapt the basic cylindrical shape of the arm-structure. By using the bridge tool, I could delete the Polygons connecting the shoulder, elbow and the wrist and, after selecting the Borders, I proceeded to create another section of Polys, but this time I could modify them into realistic muscle and increased mass, a la a common shop manneqiun

 

Next, I deleted the rear-most Polygons of the hand, got hold of the vertices that made up the border of this now hollow object, and began to manipulate them into a rough oval - this was so that the bridging would be more accurate and (hopefully) have as minimal distortion as possible



Unfortunately, try as I might, there was a small hiccup when bridging, as the number of vertices in both sections were different, leading to the discrepancy shown above: two edges from the hand had to be lead into one vertex of the wrist, which would lead to mild pinching...



With a MeshSmooth applied, the appearance of said pinching is ever-so more apparant. However, it was remarked to me that I could perhaps be attributed to the irregular construction of a mannequin, as per in real-life, and may even provide a greater adaptive quality when I bone the wrist to rotate as part of the animation

 

In this rendered image here, the arm is facing downwards, protrusion of the knuckles signifying this clearly, while the slight fold where the upper-arm meets the fore-arm at the elbow is another example of a potential adaptive quality when boning this area  in a natural bend (again, found on shop mannequins)


Using ALT+X, I made the arm-hand object transparent while I created the bones to allow me to bend the thumb/fingers into a natural, relaxed semi-clenched position and, although Manny's hands wouldn't be animated (other than rotating at the wrist), I still wanted to keep a smooth and progressive motion to reach that positioning. Here, the construction of the bones is visible; a single one going from the wrist, with another connecting a set of three bones (for each finger) and two bones (for the thumb) back to that wrist-bone

  

Once I was happy with the bones postions, I used the Wire Parameters option (having right-clicked on the first thumb-bone) to attach a dummy helper to control rotational movment along the Y axis. The green helper box allows animation of a skinned material to be manipulated by a boned section of the mesh, as well as allowing the user to merely seek out which helper controls which bone, rather than clicking and un-clicking each bone around the skinned mesh in order to find the correct one for that point in the animation!

  

This screenshot demonstrates the window in which the user can specify the order and axial-direction along which the Parameter Wiring will take place. Here, I am assigning the dummy on a Y-axis bezier-float euler to constrain the float-wire of Bone 13, at the base of the thumb (I will rename each bone properly before I clone this arm, honest...)

  

I crossed my fingers as I assigned the exact same command to each individual finger bone, using the same method as before, even going so far as to assign the outer-most thumb bone along the X-axis, so that I would rotate it slightly, to follow a human triat we have to be able to clech our thumb atop our fingers (say, when we're making a fist). Much to say, I was extremely happy with the result - no obvious distortion visible and no problems with the MeshSmooth disrupting either

  

A rendered shot of the hand, demonstrating how, after much deliberation, the decision to make each finger just that little bit irregular from each of the others has meant that the clenched-form isn't uniform - exactly what I intended to happen, for once

  

The other item that was needed before I finished with the limb was to add the attachment for when the arm will be rotating through 90degrees. This was a simple case of selecting four adjacent Polygons near the shoulder, along the same line as the closed-hand, thus allowing the hand to face inwards to the torso. The Polys were then extruded in two layers, with the top layer having both downward facing Polygons bevelled, as shown above

  

It was at this stage I attempted to add in the bone to allow the elbow the bend, but I couldn't, for some reason, get the bone to react to the dummy I'd set up for it exclusively, and when I tried rotating it, the bone just lifted out of the skinned mesh of the arm! Odd, but I'll come back to it another time...

  

Here is a render of the right-arm: the fold at the elbow and slight discrenpancy is obvious, granted, but the MeshSmooth really has enhance the appearance of the mannequin's intentional gloss-texture

Thursday, 4 February 2010

Weeks 1 & 2 - Getting to know the Biped/Basic Rigging/Skinning Techniques/Field Research

  Description

This module wil focus on developing and characterising initial designs, coming up with a credible and forthcoming prototypes, specifying the character/s attributes and behaviours: stance, gait, motion and so forth

Techniques also featured may include; articulation, lip-syching, gesticulation and demenour. Much of the time will be spent on devising, designing, testing, critiquing and refining the character/s on-screen

The Brief


An analysis of human behaviour and body language; involuntary behaviours and learned all provide signals and signs to let other know how we feel, those actions coming across in all manner of ways

The creation of a character that mimics the body language and/or movement of a subject used as research


My Ideas

I took the idea of creating this animated character, with all the human qualities of emotion described above, but decided to not go down the typical route. I could have easily decided to create a realistic male/female character doing an everyday activity, using the typical selection of 'stock' emotions: anger, happiness, boredom and so on.

But, my interperatation of this module took an unexpected turn as I walked around the Town Center and spotted all the clothed (and some not-so!) mannequins in the department store window displays; with their empty eyes and static, inanimate nature, I thought "why not take the most a-typical inanimate object we see (almost) every day and provide it with that exact set of emotivity and movement that gives humans our uniqueness"

This is how my idea began...


Initial Images (to illustrate my point)


 


 


I want to keep a decent amount of detail in the face and overall skull form - at this stage, I knew that the only area I wanted to be animated and given a realistic form, however, was the eyes; the part of the face which can emphasise and display so much more emotion than even the mouth, in my opinion...


 


 


I have a bit of a dilemma; I want to keep the time spent of modelling the characer (who shall be named Marty, for the sake of humanity) to be kept within a small time-frame so, after taking these two images here, I knew that I should only focus on the upper half of Marty. Concentrating on the torso, arms and head would mean I could devote more 'animation' time


 


 


I sourced these from online dressmakers and mannequins markets; I can see now that the intentional strong form of Marty's torso and definition of the face also will more than make up for the lack of legs and, upon finding the image above, I could design a simple 'stick stand' or 'tripod' on which to place him (freeing up animation time once more)



Although perhap looking inconsequential,  spotting this joint/attachment when de-constructing the mannequin in one of the local department stores, a strong action sequence of the 30 second or so animation came to me. If I designed Marty's arms and head to be attached in the manner as this image here shows ( here, the connection between the torso and lowers), the prospect of them falling off and him trying to collect them again could supply a comedic factor, his misfortune inversely promoting a greater sense of humanity, in an otherwise inanimate object!






These last two images (gathered from an online American film-props store, no less) brought the prospect of supplying Marty with human triats and actions home to me; although I'm only intending animation of his eyes and movement in the arms, there may be the opportunity to explore the possibilties for introducing the stand itself within the movement through the scene, contributing to Marty's emotional demenour and judgement.

Now I have come across the field research and seen what 'is out there' on the market, including all the modelling aspects and realistic physical-form too, I need to develop some rapid prototypes and design possibilities to finalize Marty as soon as possible - the ideas are there, just a case of nailing a few down to paper...