4 WAYS TO DESTROY SOUTH ASPEN HIGH
Quick Links to Posts:
November 9th, 2025 - Speed Trail Iteration Test Comp
November 2nd, 2025 - Speed Trail Variation
October 24th, 2025 - Set Extension Test Breakdown and Further Improvement
October 19th, 2025 - Set Extension Test
October 13th - Shadow Pre-Vis and Test Shooting
October 3rd, 2025 - Speed Test Renders and New Animatic
September 28th, 2025 - Speed Trail Render (Version 01) and Test Shooting
September 23rd, 2025 - Fresh Animatic and Speed Trail Test Render and Comp
September 16th, 2025 - Preliminary Shot List and 3D Plan
September 13th, 2025 - Animatic Pre-vis and AI Match Move
September 9th, 2025 - Speed Trail Test and Quick Render
September 8th, 2025 - Rough Animatic and Storyboards
August 19th, 2025 - Shadow Monster and Speed Trail Previs/Proof of Concept
July 23, 2025 - Project Layout and Ideation
November 9th, 2025 - Speed Trail Iteration Test Comps
For this weekend, I did some new speed iterations down the hallway shot, as well as new compositing for the lockers. Here is the original design from two weeks ago as reference. The lockers aren’t fully there yet, so more texturing and lighting tests will be done before the real renders.
These two iterations are based off the Sonic running trail and the Mekkari running trail. One involving lightning and the other a simplistic trail. Neither of these will be the chose option as they are too basic and don’t give the feeling for the type of speed we want.
This trail is loosely based off the quicksilver trial and involved more distortion form the trail compared to other iterations. Not the final one chosen, but a good design nonetheless.
These two are further iterations of the original design, which was custom from the start. The main change was the strides taken by our speedster. The left has a quick running motion while the right as a more galloping approach.
Taking the longer strides, I did different rates of decay and noise to have the trail last longer but keeping the same speed rate consistent. Here is the light noise version that has a slower gradual decay.
This next version has the same decay rate, but more noise as the age of the trail increases. This version was the unanimous favorite between me, the director, and the producer of Four Ways. This will likely be close to the version we choose simulation wise.
This weekend I also took some time to do a test simulation and render of our laser vision we are adding to Dr. Fenster. Here is a render I did, involving fire spread from the beam onto a collision object.
November 2nd, 2025 - Speed Trail Variation
For this week, I wanted to do some speed variation tests, just to get some more options and gears turning for what we might like. These have some post comp work done, with glow and lens dirt, but nothing major. Below is the original design that I made.
These next three are different alterations of the trail. The first is one based on the sonic running trail from the movies. The second is more towards quick silver, though the sim force is quite high, and his is more distortion than anything. Then the last is more towards Mekkari style with single lines following the path, nothing super detailed.
Utilizing some cryptomatte post comp workflows, I was able to take the render and combine certain elements from the other renders without doing a whole new render. Here are the two other variations combine two of the tails together. Original with Lightning and Original with Mekkari.
October 24th, 2025 - Set Extension Test Breakdown and Further Improvement
For this week, despite being sick, I was able to get a little bit for further development on this set extension speed shot. The main thing I added was a dust kick up trail behind the speed trial to add some more depth and realistic behavior on the environment. Firstly, here is a detailed breakdown and node tree of my current comp.
For the banner in the back, adding onto my clean plate, I took a online banner asset, combined and resized with our Rat logo to get the right placement. I then did a soft light of the background plate to get the right color and cast shadow from the lights onto the banner. I then added a drop shadow, using a gizmo, to add some depth. Below is the node graph for this setup.
For our lockers, Tripp is responsible for creating the model and textures we will be using in our extensions. The left is straight out of Karma, with a simple HDRI as the lighting. Since the texture isn’t progressed that far, I decided to add some post process scratches to it, as well as a drawn image of our logo. For getting our logo to look drawn, I leveraged Gemini for a quick recreation of our logo.
Here is the node setup for how I achieved this result in Nuke.
To give our lockers better depth, I rendered an AO pass outside of Karma, allowing me to control first hand, the amount of depth I want to give my lockers and ground. Here is the result of grading my object and ground based on the AO pass.
I then was able to add my simulation light, trail, and other post-process dirt that I have been using to achieve my speed effect.
Here is that node setup in Nuke.
My biggest change I added was a kick up dust trail following my speedster running down the hallway. Below is the process for how I achieved it. First I took my running character (same one used for the trail), then I used an attribute transfer to transfer the footstep to the ground plane. The result is shown below.
I then took this transfer, and using a SOP Solver was able to write a simple script to keep the transfer for the full running sequence. Then using another Solver, I was able to create a life attribute once it was created, allowing me to delete after any determined amount of time.
I then was able to create a POP simulation, which I could feed into my Pyro solver to ultimately get my kick up effect. Here are those results.
I sent this to the farm and got a good pass that I could then composite only my already made plate. Here is the final result with the dust trail added to my plate. I plan on refining this and using subtle bits of kick up for our speed trail to add some extra depth and realism to the scene.
I also completed another test extensions, utilizing Syntheyes and Photoshop to clean up the scene and seamlessly add the lockers in. Here is the final export and breakdown. This is not a final pixel product, but gets the idea across nicely.
October 19th, 2025 - Set Extension Test
For this week, I focused on some tracking and set extensions, as well as some realistic footage of the speed trail in our comps. From the footage that Tripp and Zach gathered, I attempted to track this shot below. This experience and practice with SynthEyes taught me a lot about 3D tracking. I feel that I know my way around SynthEyes now and can confidently say I am ready to tackle any shots we take for our film.
For this shot, I wanted to try some easy set extensions and alterations to our plates of Arnold. For our purpose, I decided to also do a clip with Aaron running down the hallway. Here is the original plate.
To start this comp, I took a still image, exported to photoshop, and using some cleanup tools, took out the paintings, signs and Tripp in the background. Using SynthEye’s auto/ML roto tool, i was able to get a decent alpha of Aaron, that I could cut out and put over the lockers. He is blurry with a lot of motion blur so the auto tool made quick work. Here is that cleanplate.
I then took Tripp’s Env that he made, including his lockers, and was able to render out the lockers to comp them in, as well as a light pass for the hallway interaction with the speed trail. Here is the final comp result.
Here is the full before and after for the whole comp. For this coming week, we are planning to meet with Marina and Chelsea and fully map out the layout of Arnold based on the story for our script, from this, Tripp will be able to fully mock up the environment. By that time, Felix should be casted, hopefully. I plan to further develop my kick-up pyro for the speed trail design, adding some subtle detail here and there for it.
October 13th - Shadow Pre-Vis and Test Shooting
For this week I focused on furthering some testing from previous weeks. This is the added trail and inkwash effect, fulling done in comp - nothing rendered.
I also added a small dust FX so my speed comp to add some depth.
We also went to Arnold and filmed some test footage, some of this will be reshot as there is a lot of motion blur currently.
Here is newer footage filmed with less motion blur.
I also began tracking the first shot, I am doing this manually, so I have complete control of the tracking markers. Here is my track currently set. For the res of this week, I plan on doing some set extensions on this clip and altering it completely. We want to be able to shot Marina and Chelsea the extent of how much we can change.
October 3rd, 2025 - Speed Test Renders and New Animatic
For this week, we did some test filming and test composites for some FX we are doing. My two consisted of the speed trail and glowing skin. Here is the two speed test shots I composited.
Here are them seperated into single clips
For each shot, I had two render passes, including the speed trail and emission casting. For the casting, I set up geometry similar to that of the plate.
Here is the compositing node tree for the speed trail.
Here is the glowing skin test that we did. I plan on doing one with Tripp’s eyes glowing, but that will be tackled next week. Here is the skin test.
To accomplish this, I took a few different types of cracked surfaces and images, utilizing some comp tricks to have the “subsurface” effect on the skin. Here is the node tree.
This week, we also compiled a new animatic, only a few changes and more FX shown in this version.
September 28th, 2025 - Speed Trail Render (Version 01) and Test Shooting
Since last class, I have dome some Speed trail rendering, testing the trail in motion, as well as emission/glow.
Here is the 1x, normal speed
2x Speed
4x - Favorite Result
Glowing Skin and Eyes Test - Testing rendered eye inserts as well as glow amount.
We also did some test recording for various shots that will require VFX. Here are some of the main shots I will be testing some VFX for. Including the glowing skin and speed trial.
Here is the Speed Trail shots, this is after Eve uses her shadow powers to same both herself and Taylor.
I set up my camera, matching the perspective, using fSpy to help.
Here is my sped up plate, with the motion blur as well as my newly made simulation, now with slowly fading pieces, this will drive my opacity in my material to add to the fade.
We also recently launched out gofundme for our film. Here is the link for it - https://gofund.me/609c574b3
September 23rd, 2025 - Fresh Animatic and Speed Trail Test Render and Comp
Since my last post I have bee working on refining the speed trail as well as diving deeper into Karma rendering - utilizing the best practices for the best result and quickest renders.
Some changes to the Speed trail simulation were slight, but I ended up doing two simulations, a refined point specific one and a overall wave one. Here is a look in the viewport as well as my current Solaris tree.
I I then rendered some layers and attempted a test comp, utilizing AOVs and other various comp practices to mud the image up. Here is that test comp render.
My node tree wasn’t complex, though it might look it. I took my three layers and did some heatwave warping, color grading, and other muddy gizmos.
Here is a simple breakdown showing the build from base render to full comp.
As well as this, I took our most recent animatic, re-pacing it, getting it from 15 minutes down to 11.5. The flow is much better and has better attention to dialogue and timing with that in mind. Here is that animatic. It is going to be refined and adjusted further, as our shots will change in a few parts, but as far as story, this gets the point across rather well. Special thanks to our Director, Marina, and our Assistant Producer, Sky, for these wonderful boards.
September 16th, 2025 - Preliminary Shot List and 3D Plan
September 13th, 2025 - Animatic Pre-vis and AI Match Move
For this week, we focused on developing some pre-vis for our animatic, mainly shots that in cooperated a lot of VFX or heavy on VFX. Here is the new animatic with inset pre-vis that we made. This is not yet paced, that will be done later this week.
For my tripping sequence, I did some test renders in Karma to show the movement of the particles in light and space. Here are those test renders.
Below is the all the pre-vis I was responsible for this week as well as a description.
This pre-vis to towards the beginning of the film, where we are first getting introduced to the kids. Taylor, the cocky athlete, is bored and to pass the time decides to use their powers and run across the gym, but trips over Felix’s backpack. She reveals that she is very powerful, yet rather clumsy.
This shot directly follows the kids hearing Principal Clarkson dying and Dr. Fenster is coming after them. Taylor leave in a huff, seemingly out of fear.
This shot follows right after Taylor and Eve escape from Lamda, instead of sticking around and talking with Eve, she takes off mid sentence.
For this scene, Remy is backed into Lamda, but Taylor comes speeding and talked him out, trucking him to the ground. This scene is subject to change, as the and story board isn’t fully fleshed out just yet.
On top of the pre-vis, I decided to test out some AI Match Move for some early R&D. We will likely need match move in many parts of this film, so having an early test of it is crucial. Here is the results and the node tree of how mapped the rig in Houdini.
I used this same method on this stock video a man running and used my pop trail method as well. I tested a render and composited it on the clip. Here is the result. This will be the similar method we use for adding the trail.
September 9th, 2025 - Speed Trail Test and Quick Render
Over these days, I have worked a bit more on the simulation, testing different forces and controls to add a new look to the POPs. Here is a quick render test for the original set up of the POPs I had in the prior test.
I built a wedge cache chain in order to speed up the look process, allowing me to see multiple different forces and only having to cache once.
Currently, in these simulations, the particle distance is far and creates gaps in the smokey wisps I want for the render. To combat this, I am setting up my points and samples differently, using less simmed points, and some trail and jitter tricks. This is quicker as far as caching and simming goes. Here is the node set up.
I also decided this was a good time to test some renders in Karma. Coming into this year, I wanted to work within Redshift, however we didn’t get license for this year, so I am having to make Karma work. I wanted to use Redshift as I have grown quite fond of it over the years, and really like the interface, look dev/material builder capabilities, as well as the viewport speed. Here is my simple set up in Solaris to make it simple for a quick render.
Here is a quick render of the current sim test. Some slight post process motion blur added.
September 8th, 2025 - Rough Animatic and Storyboards
August 19th, 2025 - Shadow Monster and Speed Trail Previs/Proof of Concept
During my limited free time this summer I have been working on a few approaches and pre-visualization of both the speed trail effect and shadow monster comp.
Here are some references for our speed trail effect we have been gathering.
We especially liked this effect for quicksilver, where the trail is attached to where he has been with some subtle movement after its “emitted.” The trail isn’t heavy in geometry or overly complex, but the wave of heat and bending of light is something we really like as far as look.
Presented below is the tests I have done with pops to build a speed trail based on scattered points, attached to a mesh. This would be a similar approach we could use, not building a CG double. But if we have a animation (which we can take off mixamo), we can “mimic” the overall movement of running, and work some compositing tricks to add it to our shots.
Here is the first test using pops. Mimicking the trail described above, using points attached to a moving geometry.
I took these points attached to the geometry, blasting away the others and using these are my emitted for the pops.
To do this, I painted where I wanted my emitter to be, and scattered points based on the group. Deleting the non used points.
Here was the super simple POP set up, adding drag and a small amount of noise to force the particles sporadically.
I also tested the look of adding trails to the points and carving various distances. This gives a lot of control for life span of the trail and what gets carved. So if we want to combine a Flash like lightening and a quicksilver heatwave, we can combine the two.
I basically take a POP sim and add a trail, carving based on the connectivity of the trails. For the future, should we want lightning/noisy trails, we can use this setup and add and per trail adjustments within the for each.
Here is a visualized combination of the two, with a stylized trail that breaks at various points.
Here is that trail stand alone.
Furthermore, I did some work and proof of concept when it comes to our shadow monsters. Here is the references I have compiled for the concept we are wanting to achieve.
This is a super rough version where I casted a shadow from a GS card in Maya, rendered a shadow pass, and comped it in Nuke. For the real deal, there would be more care into the actual casting of the shadow and the light within it, but as for a visualization, I think it gets the idea across.
Here were the two plates I took, one on a blue screen, and another against a blank wall, with a harsh shadow.
I first took a key off the bluescreen, uploaded this to maya, and casted a shadow from a card geo.
Since the wall is flat, and my shadow isn’t moving, it was very similar to the regular plate, but if I were to add geo, like we will in reality, the result will be different. Here was the cast shadow layer and the way I composited within Nuke.
One challenge I know we will face is cinematography of these shots, and setting up the scene the way that would be best when it comes to adding this effect. We have talked with our DP and are getting an idea of justifying the light for our scene and the possibility of doing more GS work with CG environments, which Tripp would take control over.
Overall, we have a good start to the project, with both some previs and initial ideas/approaches for the super powers.
July 23, 2025 - Project Layout and Ideation
Welcome to my senior thesis build log. For my thesis, I am teaming up with a few other VFX Students and Film students to create the film - 4 Ways to Destroy South Aspen High. A story about 4 teenage high school students with super powers.
The current crew/department heads are as follows for this film:
Director - Marina Frontani - https://www.linkedin.com/in/marina-frontani-807586252/
Producer - Chelsea Nott - https://www.linkedin.com/in/chelsea-nott/
Writer - Nevin Allen - https://www.linkedin.com/in/nevin-allen-aa4340234/
VFX Supervisor - Caleb Stovall - https://www.linkedin.com/in/caleb-t-stovall/
Associate VFX Supervisor - Barry (Tripp) White - https://www.linkedin.com/in/barry-white-0298b01b5/
Associate VFX Supervisor - Zachary Malich - https://www.linkedin.com/in/zmalichvfx/
Associate VFX Supervisor - Aaron Linde - https://www.linkedin.com/in/aaron-linde-b2911b253/
Furthermore, For our project, we, as a team, have set up a Collaborative Space as well as a naming convention for our project. On top of this, I have started an organization project on “Monday,” a collaborative space designed to keep our project running smoothly but assigning shots, and keeping the current WIP for each one organized and in one place. This way there is no issues in miscommunication on who is work on what shot and what has been done/not done yet. Here is an example of a sepereate project, called Celestial Court, in which I set up the same approach for my team