Hitting the Ground Running

Hello, and welcome to fall of my senior thesis! Our senior short is officially full speed in production, and we have a lot of work ahead of us. But at least our story is solid and our characters and style are pretty well developed; I'm proud of what we've achieved.

This is largely going to be an info. dump, as I have been keeping relatively busy over the summer and the last couple of weeks. As the lead of FX for our short, I'm in charge of... well, all of the FX. Lately I've been working hardest on Character FX for the sake of proof of concept and setting up those pipelines to plug and play when we need/want.

First and foremost, please make sure and check out our latest animatic as it's been through several iterations since I last posted on here. Eric, Sabrina, Christine, and our story team as a whole have made huge strides, and we're at a point where we're using it to create a shot list, layout, 3D Animatic work, and 3D camera work (which Hogan has been doing amazing at).

Let's take a look at some hair/pipeline work that I've gotten through since you last saw.

Starting with the worst, as I was trying to get alembics to play nicely with hair gen and simulation in Houdini from Maya.

One of my more sophisticated grooms for Daisy concept, which quickly got critiques - people like the clumping, but imagined her with longer, curly hair.

You may recognize this Roberto from an earlier blog post; this is the same simulation, but rendered with Arnold in Maya. It acted as my first major proof of concept of pushing hair gen through the pipeline from Houdini to Maya. Renders were quick and the hair interacted with Arnold lighting and textures.

And this was a rough proof of concept for Daisy hair, with a cleaner, longer, curlier groom based on critiques and a rough cloth sim to boot. The alembics finally play nicely and the hair network is set up for easy plug and play for any animation from Maya. This plug and play system works for both grooming and simulation, all in one network (but wait, there's more!). Sabrina is one of our main animators and busted out the animation using the Dana rig for me to get a slightly prettier render than Roberto for proof of concept. She also is in charge of doing research for, designing, and assembling the outfit in Marvelous, which I then was able to simulate.

I worked on some Ivy over the summer, and with the help of an amazing LinkedIn connection and these two tutorials, have made significant progress in making art directable and relatively realistic ivy growth. I wish I could shake the hands of those who helped me/made these tutorials, as they've been priceless and while this is only from a couple days' work, I feel confident moving forward and adding my own twists for what our short needs.

This is a rougher proof of concept - people wanted goop in the apartment destruction scene, so I busted this out to show that it was easy enough to get started in. The falling trails are not my favorite, but it was a solid start.

And now, for the 2D side of things.

I've surprised myself with how involved I've been in character and environment concept art. I'm no storyboarder, but it sounds like what I've put out has been very useful to our sculptors, as well as additional shape critiques using SyncSketch.

These were iterations of hair styles and facial structure/style for Syd, our main human character. The short is taking place in the early 80's so I tried to find something gender-neutral that would fit the time.

This is the most recent iteration of some orthographics/full body art for Syd. Mason, who's in charge of sculpting Syd, seemed grateful so I'm happy with that alone.

Svetlana has been kicking ass with doing countless iteration of the character sculpt for Daisy; this is just some of the rough drawovers/critiques I did for one version. But truly, Svetlana's amazing. She takes the critiques well and has gotten us to the point where we are nitpicking the sculpt.

While working on mood boards and style sheets, I drew some rough concept for layout and a couple of props. It helped us get a better sense of how we would populate the space, and added assets like the rubik's cube and lava lamp to our list.

And finally, the technical:

Everything above has been integrated into the Prism pipeline - something Hogan, Ryan, Mitchell, and Jasmine have been hard at work organizing, communicating with the creator on, and creating documentation for. It's going to be priceless in the long run, and I've already pushed these models through it on top of my Houdini pipeline work. I'm really grateful to the group members who made it so accessible, as well as the Prism creator who has pushed several updates through Prism per our request as we troubleshoot it.

On top of this, our production managers Christine and Ryan have been helping hugely in making us do weekly check-ins, a detailed calendar, and maintaining open communication. We actively use SyncSketch, Slack, the remote server to our university, and Discord and wouldn't be nearly as efficient without the work they've put into organizing it all.

We've got our fridge magnets...

...and our 80's telephone. As the lead of FX for our short I'm unable to do much of anything except FX, so sadly this is all I have to show for modeling. But it felt nice to stretch those muscles again.

And with that, you're all caught up! Thanks for coming to my TED Talk, and as the semester goes full force we'll certainly be seeing much more of each other. Stay safe, and see you in the next one.

"O" for Onward! (Please don't copyright strike me)

Hello, and welcome to another episode of productive quarantine! Allegedly, the stay-home order in Denver lifts in just over a week, but that does not change the fact that campus is closed, summer internships are cancelled, and all our lives are significantly impacted, whether emotionally, fiscally, other, or all of the above. That said, technology has really come in handy more than ever, and for the most part we have managed to stay productive, all things considered. And arguably even more importantly, I have managed to remain social, along with my friends and fellow group members, which has made a world of difference in realizing we are not alone in how we're struggling through this and experiencing the challenges of isolation and fear. There's solidarity, and being able to take a break from production and still enjoy each other's company through games or movie nights has sincerely boosted morale I believe. It certainly has for me, and gives me the motivation to keep pushing forward. I know this blog isn't necessarily about that aspect of production, but I think it's really important to acknowledge how myself and my group as a whole are handling this, and still pushing through.

So as my ability to remote access the university computers is still quite spotty, I have shifted my focus to concept art in order to keep producing something, even if it is just in a different production vein. So I drew some concept for our human characters; this model is our main reference for the non-binary main human.

With that in mind, this is what I put out for their design:

Obviously the left is essentially a carbon copy of the reference image, but it was a good starting point.

This was my first attempt at some dynamic poses, which are certainly a bit lackluster.

I got a tad more dynamic on the right, and emphasized the outfit, as well as bringing in Palmer, the zombie hand character.

And now, the S/O:

This was my main photo reference, namely for hair and the glasses. We are trying to normalize diversity in our short (the amount of reference for 3D characters that are black was severely lacking, which is upsetting), and personally I'm really excited for the hair simulation for the S/O! Here's what I drew out:

Once again with the carbon copy, but then from there I expanded my horizons a bit and was able to add variety to expression as well as explore a couple of outfit ideas.

Overall, thankfully my drawings landed well! There was some critique on avoiding gender binary colors on the non-binary character, and the overalls were immediately shut down as they were used in last year's short. However, everyone loved the yellow dress for the S/O, and just the S/O in general. The most intense overall critique was to soften the edges and shapes, and basically make our character look a bit more cartoonish. I did my best, and while it's not much I did produce these:

People were very happy, and I feel a bit better about things like eye and nose shape. I have more critiques moving forward, so this obviously is not the end. But my concept art has allowed our sculptors to have some reference to begin getting rough sculpts out for our humans.

To wrap up the drawings, I drew a character lineup to show our professors, as well as to get a sense of how the characters meshed together.

There's a reason I'm doing VFX, but this drawing did the job. I made the non-binary character's jacket jean rather than a bright blue, and our group has been going through several iterations of the Bat Vampire character; this is my depiction of our latest version. Discussions are currently being had as to whether or not we will be keeping "Batman", and whether Aldo will be made with fur or leaves. I have my own bias, but have no doubt that our group will succeed either way.

I know this post is already quite long, but allow me to show you really quickly the progress made on hair. I had a dumb scaling issue, but spent an entire day reworking the node network, fixing some VEX, and adjusting the grooming setup and plugin options. The node network is now plug and play; there are clear spots to put in static and animated meshes, and I have set up a version for people who want to groom that points out and explains which nodes they need, and how they work. I can then take their grooms and adjust as needed for the simulation. It's working much more efficiently now and I feel significantly more confident about how it will mesh into the pipeline.

Hopefully this makes it clear as to why I need others to be in charge of groom, but despite that the hair is looking significantly more realistic. The above is wispy, and I added grooms underneath; the below is lower res. and shows some blockyness on the separate grooms, as I foolishly rendered out the first frame, so the hair hasn't simulated and relaxed yet.

This is a rough render of a lower-res simulation - it shows a bit of bounce, but is still quite an improvement.

Okay, I've already written a novel of a post, so if you've made it this far congratulations! Moving forward I'll be shifting to doing some R&D for Vellum cloth simulations in Houdini and seeing if that is the best option for our short. I need a break from hair.

Pushing Through

Well, welcome to the apocalypse. We're witnessing history, and it's quite a time to have a blog dedicated to something completely unrelated. But like everyone else in the world I do want to address the whole COVID-19 topic. Denver is currently in lockdown until at least 4/11/20, campus closed just over two weeks ago now, and the historic Larimer Square block on which I worked part time has completely shut down, some restaurants temporarily, other shops and restaurants permanently. I no longer have a paid job, and remote accessing CU Denver's VPN has proven relatively inconsistent, partially because my home setup of PC and WiFi alike are lackluster - the PC's, VR equipment, licenses and even monitor setups that I pay so much tuition for are now gone or incredibly difficult to access. Like everyone around me, I'm in some economically and educationally challenging times. I understand that it will also show as people decide to keep pushing or give up as this constantly shifting situations presents massive challenges to motivation and productivity. Some will shine, some will fall, and I am anxious to see which I will turn out to be. I have confidence in my group members and am grateful that we've figured out methods of continuing to be social and productive, as both friends and team members. We got this. Things are rough right now but I am remaining as optimistic as possible.

So, now to the short. Because of all of the above, my productivity did drop and therefore I don't have as much to show but I do have a couple of things. I've been pulled away from particle and Pyro FX in Houdini as our group has asked for more focus on hair and leaf (yes, leaves - I'll explain why) simulations. So first off, I solved the issue we saw in my previous post - turns out my hair was being put on a static mesh, and so the hair was not following the animation. It clips out of existence when the animated mess passes through the points of the static mesh. So, after some node network changes, I was able to push out a full render rather than a Flipbook of our dear Roberto with electric blue hair.

Doesn't give us the best look, but it shows that the hair texture works, the hair and clumps are simulating without any strange collisions, and the physics follow the animation quite well. The group was happy, and asked me to continue R&D for making it dirtier, having clumps and debris, greasier spots in the hair, etc. Essentially, making it look less smooth and shiny. I feel like part of that is texture work, but already have ideas and R&D going for making those things happen.

Another request from the group was a collisions test. They wanted to make sure characters' hair could interact with other objects in the scene without too many issues. What you're about to see is not pretty, but it was the best I could do at the time as this was the midst of COVID-19 chaos.

So the pom-pom and the grid ball... I made sure to generate clumps and frizz and other noise to the hair to make sure interactions worked with hair that was more than the basic shiny hair gen. As you can see in the first video is that even seeing through the ball you don't see any hairs coming through it. The collision was relatively simple to set up and even with pretty low settings (due to my PC), they worked! So it's at least good proof of concept, and enough to start with and prove capacity to my group.

Finally, I have a human hairstyle. I hesitate to post this video, but I suppose we're here to see the good bad And ugly so here goes...

This is my bro dude hair style. We wanted our main human character to have rather curly crazy hair that goes up on top and maintains shape, and so while I had to absolutely crank the simulation passes down to less than 1/5 their usual capacity to get 24 frames on my PC, even with such low settings I got this bounce to prove that the hair would maintain shape. I do not know if the overall Jello feel is because of my low settings, though I assume it is. I also quickly learned that while I'm not too shabby at setting up the node networks and troubleshooting, I'm no hair stylist. So I figured out a setup where I could only make a couple of nodes editable and have someone else do the groom that I could then simulate and network out. Basically I make the tool and let someone else make the hair pretty.

So the groom has begun. The hair maintains its shape and the groom is relatively easy to change and brush and shape for someone that I were to just hand the file to. The simulation is going to be a challenge to set up, and collisions with this new network (shown in the bottom right to demonstrate its current level of complexity) will be potentially challenging, but it certainly isn't anything I feel I can't handle.

In other news, our group has decided to have our yeti character have leaves instead of fur - he is now a bush man. I am still trying to fight for hair, but in the meantime I'm going to try and help R&D making a good looking leaf simulation as proof of concept. I'm currently thinking I can use Houdini's vellum solver and perhaps instance leaf geometry rather than the hair splines that hair gen. automatically makes. TBD.

Meanwhile, the rest of the group has been making huge strides on script development, character design (we're finally reaching an end point for our 'smart' bat monster, which is a relief as it's proven to be the most challenging), and color palettes. I've added all of their names and websites, as well as a link to our most recent animatic on the right of these blog posts - please go check them out, as they have made so much more!

Here's hoping my slope of motivation stays on the rise despite the practically hourly news updates, and that we all get through this crisis as ideally as possible.

Until next time!