Animation Project: Dialogue #01 - Part 3

 

 Animation Project - Dialogue #01

Part 3 - First Full Pass

April 2021

 This is the last progress report for this initial segment of animating a dialogue piece. After refining the mouth animations, I went ahead to animate the body movements and set up the camera. Since I purposefully split the dialogue into separate parts, I moved the camera as if they are connected scenes without being one single shot. Once synced up, I did my best to convey proper flow and rhythm to the overall animation. The rig I have used for this project is the Polly rig from NoisyChain on Turbosquid.

https://youtu.be/wThv1hLdfTk 

It was much easier to block out the body movements based on the reference video I shot, with multiple passes. I picked the most impactful poses and tried to replicate as close as possible. However, I did take liberties and also tried to accentuate the character 's movements in order to make the animation more dynamic. Sticking too close to the real-world proportions would dampen the readability of the scene. Once blocking was done, switching to spline and editing the in-betweens was not too tough. 

There were parts where it was difficult to pace the facial animation alongside the body motions so that the character was still relevant to the camera. I only wanted a stationary camera for this first dialogue piece, so no fancy movements from the viewpoint. This was meant to convey the audience as actual onlookers in the scene. 

Additionally, one of the hardest parts was not going too far with the body animation. Sometimes during the process, I found myself exaggerating the movements too much, and at the expense of the perspective I had already established. The character would turn too far or flow way too bouncy or stiff at times. Upon re-evaluation, I went back to fix as much as I could, but evidently some parts are still a bit much for the tone of the dialogue.

Overall, this has been a teaching moment for my journey with animating. I do not think this project turned out as I would have hoped, but at least I was able to reach an endpoint. I will leave this project alone for a little bit in order to focus on other things. I would like to return to this in time, but for now I think I need to put my mind elsewhere, so perhaps I can come back with new and fresh ideas. I feel that in order to improve, I would be better off restarting the whole process with a different strategy.

Rigging Project: Pigsy

 Rigging Project: Pigsy

April 2021

This is my latest project, the first of many hopefully leading to a fun final product. Pigsy is their name, and they are a sentient piggy-bank set to be the main star in my developing animated short. This is my first original character built from the ground up from conceptualization to fully rigged model.

This rig has full rotation capabilities in the body, head, and limbs. Movement in 3D space can be achieved with either the main body control, or the root control. The face has some more specific controls, such as a head tilt (in place of a neck), nose tilt, and one of my favorites: eyebrow movement and manipulation. Finally, the eyes are 2D grafted onto the model, made by an image sequence which can be toggled to display one of the many eye textures I have drawn up. There are also a few more specialized features intended to enhance the animated scenes to come.

As this is my first original character made from the ground up, I was hesitant to dive into the vast pipeline from modeling to rigging to animation, but by keeping my ambitions in check, I was fairly confident I could do it all on without having to compromise too much. I am happy to say I was able to accomplish what I set out to create almost exactly.

Making the armature for Pigsy didn't take too long, as I actually started building it as I modeled the character parts. Once I got to the actual rigging process, I ran through the process as I had envisioned and just observed what worked as intended and what didn't. The legs, body, and main head controls went together flawlessly. Meanwhile, with the help of a few tutorials, the eye texturing and eyebrow controls turned out alright as well. 

 

 

The only real point trouble was the eyebrow manipulators, as I wanted the easiest way to get clean deformation without too complex a rigging system. The solution was lattice deformation, which turned out to be exactly what I was looking for. With one control on each end of each eyebrow, I can squash and stretch to just about any extreme, which should prove very useful for expressions in animation. I am very proud and excited to see what I can do with such a feature.

This rig was a nice challenge to have as my first original character. On one hand, I had a clear vision of what I wanted the character to be capable of. On the other hand, there were obvious tasks I would have to face and problem solve for. The finished product has just about all I had hoped for. 

I began development of my first original character and soon-to-be short animation a little over a month ago, and this is one of the largest steps so far. This is the first result of my labors. I hope to push forward and begin animation and scene building next, but before that, I must begin production of additional assets and other necessary materials. Luckily, the main model here is out of the way, so the process will be a bit less stressful in the days to come.

Modeling Project: Pigsy

 Modeling Project: Pigsy

March 2021

This is my latest project, the first of many hopefully leading to a fun final product. Pigsy is their name, and they are a sentient piggy-bank set to be the main star in my developing animated short. This is my first original character built from the ground up from conceptualization to fully rigged model.

This model is a modular build, consisting of only a few simple shapes. My goal is to have it fully rigged for animation, with a list of capabilities all my own in mind. Of the many individual shapes, the main components are the body, head, legs, and head accessories. 

 

The body, probably the easiest to make, is simply a rounded balloon-looking piece that I only needed to add a coin slot on the top and a nub for a tail on the upper back end. The head was a fun task, as it really helped me understand what subdivision surface modeling could do. With only a few protrusions from a simple sphere, I was able to get the basic shape of the ears. With only a few rounds of trial and error, a simple subdivision and polishing of the polygons got me to right where I wanted to be.

The legs were a piece I wanted to be as simple as possible. I wanted the overall model to be as close to a real-life piggy bank with cartoon proportions as possible. Seeing as some piggy banks don't even have separated feet components, I opted for an easy and simple design of a barrel-like structure. I personally found the shape to be aesthetically pleasing, and it fit in the other parts of the model. After completion, I duplicated and spread out four of the limbs.

The rest of the pieces were easy to achieve with simple tweaks to primitives. The eyebrows are elongated spheres, the nose and nose barrel come from cylinders. With a few fixes to the geometry, all came together very nicely, with reasonable topology.

 

Overall , there weren't many hardships as far as modeling is concerned. I took a few liberties translating my concept art to the 3D silhouette, but overall the model is optimized as best it can be for what I am aiming for in rigging and animation down the line. Next up comes rigging, and beyond that more asset modeling and animation.

Animation Project: Dialogue #01 - Part 2

 

 Animation Project - Dialogue #01

Part 2 - Refining Mouth Animation

March 2021

 The turnover for this progress report was pretty fast, as I mainly finished up the second and third segments of syncing, then strung them all together. After that, I went back through to refine the timing and streamline the breaks and flow. Now the full mouth animation is complete, aside from a few more tweaks I may make once I begin animating the rest of the body. So far, the pacing is still a bit fast, but it is much more readable and relative to the audio. The rig I have used for this project is the Polly rig from NoisyChain on Turbosquid.

After seeing how the first segment turned out, I had better foresight going into the completion of the other segments. The first was particularly fast-paced and hard to read at certain frames. As a result, I tried my hardest to slow down the key frames, and maximized my time allotted for each spoken phrase. However, with so much dialogue going through so fast, it was hard to determine which movements were major enough to stay, and which were unnecessary in the long run. After having cut out what I could, the animation has cleared up substantially. For now, it is workable, and I will move on, but there is still some timing to take care of, as well as some further cuts potentially.

For now, I am leaving the mouth animations as they are. I feel that in order to refine it even more requires playing to the camera rather than nailing down the minute details that might not even show through in the end. So, I will be moving onto the blocking for the whole body animations next. Hopefully that will give me new perspective on making this animation even better.

 

 

 

 

 

 

Animation Project: Dialogue #01 - Part 1

 

 Animation Project - Dialogue #01

Part 1 - Blocking and Mouth Animation

March 2021

This is the first progress report for a larger animation project. I am creating an animated dialogue piece for my portfolio. This particular animation has been split into three segments, each being subject to their own multiple phases/passes of animation. This will likely take a while to complete as intended, but I plan on reaching certain goals periodically, maybe not all in one go. The rig I have used for this project is the Polly rig from NoisyChain on Turbosquid.

The project as a whole, once finished, will have a cohesive flow of facial and body movements synced with the chosen audio clip. I plan to move from the mouth animations to the rest of the face, and then the full body. I also plan to move through each of the three segments equally in terms of refinement. As of now, the plan is to move from blocking - refining - timing - finalization for each of the aforementioned phases. So far, blocking has taken place for the entirety of the mouth animations, and refinement and timing are underway.

For this first part of the dialogue, I have synced the mouth shapes and movements fairly close to the audio clip. For reference, I did some blocking in real-time video, and also made sure to save specific mouth expressions to call back for an easier process. The in-between movements were tailored as needed depending on how fast or slow the audio pacing was.

 (Segment 1/3, refined and synced with audio)

 

The hardships of the initial stages for this project were mainly getting familiar with the base mouth shapes. I used the standard charts for facial animations and tried to get as close as possible using the provided manipulators on the rig. Furthermore, the rig is relatively new to me, and it took some time to get familiar with the controls. In fact, there is much more to explore, once I move on past the mouth to the rest of the face, and eventually the entire body controls.

For the refining stage of the first and currently second and third segments, it has been particularly tough keeping in sync with the audio while also keeping the expression readable. The timing is important because some parts of the speech are faster, and others are slower, not to mention the various pauses. As a compromise, some key shapes must be cut out or expressed in transition. From what I learned from lectures, sometimes less is more when animating dialogue. It is more important to accentuate the extreme poses and the natural flow rather than fit in every single movement.

The remaining two parts of the dialogue are blocked out with the basic shapes, however the syncing is not complete, so they will be included with the next post. In the meantime, I will be going through and streamlining each individual clip, then eventually blocking out shapes that will merge them together once rendered out for compilation in editing. As mentioned before, once the mouth animation is finished, or up to an acceptable level for me, I will move on to the next phase and repeat the process.

Modeling Project: Hourglass + Skulls February 2021

 

 Modeling Project - Part 2: Hourglass + Skulls

February 2021

As an addition to the hourglass project, I modeled a skull (with jaw and teeth) that would fit inside. This project was concurrent and was intended to compliment the aesthetic of the overall model project.

This project overall was inspired and modeled in reference to this image, depicting an ominous hourglass with skulls inside. I wanted to replicate it in a 3D space, and put my own spin on it with the realistic textures and particle simulation.

 The skull in particular was another means to practice and brush up on my modeling skills. This was no uniform modeling project like a box or machinery. I needed to visualize and understand the organic shaping and context for the curves and real-life measurements of skull and bone. I started out with orthographic references, plotting vertexes and extruding where necessary. To complete the final product, beveling and solidifying edges gave the smooth look. I took some liberties with the modeling in order to make the model fit into a 3D space. Some angles are sharper, and others are wider than their projected measurements according to the reference pictures. Overall, it was a successful attempt to replicate and also make the project my own.

 

Some noticeable hardships with this project were figuring out which edges needed more thickness, and which needed to be smoothed out more. One challenge was the gap on the side of the skull, where, from the front, its hard to notice, but on the side there is a noticeable dip where the dome of the skull expands into the front-facing features. Additionally, the jawbone has much more curvature and volume on the inside, which I did not realize until I later down the line. As a result, I had to do some extruding and sewing back of vertices to try to recreate the gaps which were lost. I am not sure what an easier alternative solution would have been, but the compromise seemed to pay off. After all, not all skeletons are the same. In fact, this model may well be too uniform for a realistic standard.

This project was a big step in understanding my current limits with modeling. This model in particular ended up very nice and about what I anticipated when I started. The topology was about as tight and well-formed as I could get. With more time and effort (and polygons to manipulate) I could definitely improve things like spacing, smoothness, and much more. I could have modeled the interior of both jaw and skull, not to mention focusing on the individual teeth. However, for what I was going for, this fits the bill pretty spot on.

 pictured: model without smooth shading

 Once I exported the skull model into the hourglass, it was smooth sailing. I created two skull duplicates to place in the upper and lower sections of the hourglass, clipping them into the "sand" geometry. Then, I gave them all a particle effect, paired with weight painting, to give them a bit of sand buildup along the sides. This certainly could have been refined more, and the sand could be made to look closer to the textured material, but in the interest of time and computing power, I settled for what I could. I replicated the collision effects from the previous plain hourglass render for the bottom-half components, so the sand particles would interact and fall off in a realistic fashion. In this version, the sand interacts with the skull, sand, and glass if necessary. Some tweaks to the dampening and friction were needed, and in the end it all came together practically perfect. This was a lovely multi-part experiment, and I am very thrilled with the results.

 

 

 

here is the final render, animated

 

 

 

 

 

 

Modeling Project: Hourglass - February 2021

 Modeling Project - Part 1: Hourglass

February 2021

For this project, I aimed to brush up on my modeling skills, starting out with a simple hourglass. In addition, I tried out and added a particle simulation to the finished product.

This model took only a week to complete, as far as this part of the project is concerned. The first step was fairly easy, starting with primitive shapes and slowly modeling them into the necessary forms. Only three unique shapes needed to be modeled individually, as the base and top share the same base form, as do the two poles. Meanwhile, the glass itself is one whole shape, with two sand shapes being booleans which were scaled down to give the illusion of mass within the glass.

After the initial modeling, I marked and unfolded the UVs, and proceeded to organize them for texture mapping. For this project, I used a single sheet for the hourglass components, and another for the sand. If I were to optimize it for more professional use, I would condense the UVs all into a single map. But for experimental purposes, I elected to keep the sand texture separate. The sand texture in particular was interesting to work on, because I used displacement and randomization nodes to help add more depth and detail to the render, despite the object having flat geometry, with a flat texture. It was especially fun to see that in renders in conjunction with the glass which gives its own distortion of the inner objects.

 

Once finished with the modeling portion, I wanted to take it a step further. Since this was meant to be a semi-realistic hourglass, I wanted to add sand actually falling from the top of the glass to the bottom. Initially, it would have been ideal to forego the faux sand shapes and instead use a particle simulation to then generate a mass amount of "sand" to fill and fall through the hourglass. However, seeing as particle simulations are very taxing on renders and computer processing and I am new to the subject of particle simulations, I opted for something much easier. Implementing a simple particle spawnpoint in the neck of the glass, I only needed to set up the amount, velocity, and constraints on the sand particles. This particle simulation only needs to flow in one direction, with the particles only requiring collision feedback from 1-2 surfaces: the lower sand, and the glass if necessary. In addition, I learned how to randomize factors for the simulation such as color, size, dampening, and friction upon collision.

 

Overall, this project was a very fun way to brush up on modeling again, as well as learn some new skills that may come in handy for future endeavors. The toughest part was definitely getting re-acquainted with UV organization. While not the perfect or most realistic model in the end, I think this hourglass is a good starting point for even better final products.