Continuing from the last entry I removed the slow motion. As it became clear later down the line, the lack of sound and any real build up to the surprise appearance of the larger creature were the major contributors to how jarring its arrival in to the scene felt.
In Unreal I added ExpotentialHeight Fog on top of the already in place Atmospheric Fog to give the effect more of a prominent presence as well as to attempt volumetric lighting. This failed however. The angle of the light and the consistency of the trees meant that whatever volumetric lighting beamed in to the scene it was too choppy and barely drew any notice. This I feel was for the best. Adding unnecesary effects would have only drawn fous away from the important aspects of the short, the animation.
A few small tweaks were added in 3ds Max to the animations. The humans hands were animated more. The mantis’ movments mid run were made more fluid as oppose to his stiff upper body rocking back and forth. The lizard was given more weight as it adjusted it’s position post attack on the mantis and finally, the human characters head was moved into a better position in certain shots to keep focus on the insect. On to the final stage.
I had a choice to make given the time. Particle effects or sound. Sound for me felt like the natural option, it would play a much bigger role in setting mood and tension throughout the scene that some dust from the felt and spit from the lizards mouth would. To begin with I made a list of sounds I would need then proceeded to record foley in the sound lab. I used brushes and the cushion of a stool for the grass, a crisps packet for the mantis crunch as it was killed, the leather cushion again for the steps of the creatures with varying levels of impact depending on if it was for the mantis or the lizard. A stick swung in the air to create the effects of the mantis’ swipe for the human character and the lizards swinging around of said mantis. Finally the mantis’ sounds and the lizards sounds. Hissing noises performed by myself. I also attempted a number of growls and roars for the lizard but ultimately those did not work out as I had hoped.
Wind, birds ambience, the lizards roar and a background track were sourced from royalty free websites. The foley sounds were edited in Audacity and the final scene was compiled and edited in premiere.
And that is all she wrote… Done and done~!
Since the last entry the fog has been dabbled with more to give it a more realistic looking aesthetic. The field of view through the murk has also been changed to allow the camera to see further in to the fog to catch the silouettes of the trees further away. The biggest change to the scene however is now there is grass. Moving, dense grass. Such a simple edition and yet it does so much to keep the floor from being so distracting as it once was. The perfectly smooth but textured landscape caught the eye far too easily and tore attention away from what was happening.
Lastly with the shots and the sequence practically complete I wanted to take the time to experiment with the addition of slow motion. Originally my intention was to take the animation in 3ds Max and stretch out the keys of the shot I wanted to slow down but that then meant I would have characters moving slower with background plantlife blowing through the breeze at a normal speed. I held to a hope that Unreal would have a sollution and sure enough, it does.
In the Master shot of the sequence I added a Play Track, a track that is set to a default 1.0 and can be keyed. When a key is added ot marks the play value so with a key at frames 100 and 200 set to 1.0 and a key at frame 150 set to 0.1, the play speed will slow down and pick back up again between 100 and 200. Using this I have introduced sudden slow motion at the moment of the larger beasts arrival in to the scene before biting the mantis. All in all it looks good! It gives the viewer more time to process the beasts arrival. Before it all happened simply too fast and as a result was often jarring to those I presented the scene to.
Using a collection of asset packs from the Epic Games market I have built up a scene for the short to take place in. As intended, a forest. The process was not exactly a smooth one as there were some issues with the lighting and shadows blackening the textures of the trees and the new skin created for the human courtesy of a fellow student.
This was resolved none the less by removing AO maps from said figures being affected. An unintended but welcomed addition for the environment was fog. I never planned to use it from the start but after experimenting, I believe the scene is much more contained and focused than it would have been without it.
We finally made it to the Unreal Engine! The animations carry over. The human characters textures are broken alas. Seems it was created without Unreal in mind. Whilst this is not an issue and the animation is the focus, a kind soul has offered to re-create his textures in a way that Unreal will accept as a material.
With my characters in place the main focus has been the camera as the environment will be built around the movement space of the actors in the scene. I created a level sequence and a single shot for the scene with cuts playing out between 16 cameras in total using the pre-vis in 3ds Max as a reference guide. This has also given me room to play around some more with certain shots and try slightly altered angles. The cameras though timely did not cause me issue setting them up, what did however was the render.
Image sequences would skip a great number of frames and the cause of this was the motion blur that Unreal likes to have on by default. Disabling this feature fixed my issue and from there the sequence rendered out nicely. Now the next big step, building and lighting an environment for my animation to play out in!
More of the animation in Max has been cleaned and polished. It says in my last entry it was all clean but there are always new flaws to be seen with each little bit of time spent away from the project. The roll is a prime example. I also took a moment to expand a little more on my report, laying out the foundations and giving the use of diagrams a try as a means to guide the format of my writing.
UNREAL! Yes, I have put the animations in to Unreal and they ‘work’. Some issues appeared in the human character. textures not carrying over and the skinning on the gloves leaves much to be desired. I have yet to try and fix that but if it turns out to be too much of a hassle I will just delete the gloves from the model as a whole. My next big focus is to create the scene. Also finish of refining the max animations for the creatures.
In the last entry I mentioned how I had cleaned the mocap, put it in to 3ds Max, cleaning it again in there and discovered my motion capture data was running at 120 fps as oppose to the intended 24. This is a mistake that I can say with a breath of relief has been corrected!
The animations have been cleaned, they run at the correct frame and are now in the overall pre-vis scene, tweaked and positioned in their appropriate places through the power of motion mixer.
The report has also been started! The contents are laid out, the introduction has been done. Things feel like they are coming together at last. My next major step now is to finally implement the animation in the Unreal just to test it works like I think it should. Then I have my hand-key animation to finish and on and on we go~
Following up from the motion capture shoot last week, this week has been focusing on preparing that data for use in my overall animation.
From Shogun, where the footage was recorded and the mocap data was captured, I then put these files in to Shogun Post where I was able to clean the data. It was a nice learning curve and it definitely have me a huge appreciation for how impressive this software really is! What a step up from Cortex in the previous year.
After the data was cleaned I moved on to Motion Builder so i could fit the motion on to the human characters biped. From there I could send the motion straight over to the human characters skeleton in 3ds Max and it was here I found an issue. For reasons unknown, the motion capture data applied itself to the biped completely wrong. the character was raised in to the air, limbs were hyper extended. While I admit I felt a moment of pure dread, I tested a theory and it paid off. Removing all the characters assets and leaving only the skeleton behind lead to the animation importing successfully. I am not sure what it was but something in the characters model was interfering with the import.
So now the human character has motion capture animations applied. What then? Well, then I set to cleaning the animation in 3ds Max. A task that has now been restarted as I failed to take notice of the fact my animations carried over at 120 frames per second as oppose to the intended 24. What this meant was now my animations, whilst looking clean, where in slow motion. I learned of my error and have since corrected my mistake. In Motion Builder I plotted the character at 24 fps but I did not change the fps of the overall file. This neatly brings me to what I am on with next. Cleaning the animation in 3ds Max with the power of Layers.
My project is behind on what is to be expected. It’s been a very reflective week to say the least… also a very busy one!
The hand-key animation for the mantis is almost done as a base to add finer detail in to. The addition of having the larger monster physically ick up the mantis is also resolved. Using a dummy as the mantis bipeds parent I have tethered said dummy to the larger monsters head, keeping it keyed in to world space until the moment those jaws clamp around the mantis’ body, from there its connection to world space is replaced with treating the lizards head as the core of the dummies movements, driving the mantis as a result.
The motion capture data has been shot and is ready to be cleaned up, the shoot was without any issues, each movement being shot between two and three times to give me a wider range of choice, the option to mix and blend various motion capture sequences in to an ideal final animation is not without consideration but we will see when the data is cleaned and actually applied to the human character.
Lastly there was the progress review. I have much more confidence in my ability to speak over my efficiency at making a sparkling, pretty slideshow for certain. Time will tell if that’s enough.
The biggest take from this week has been giving some hard thought to where I am at and where I feel I ‘could’ be.
Things I were informed about in my feedback were corrected. Certain shots lasted too long, the camera moved too much. traps I’ve fallen in to in the past but that’s the great thing about a second opinion!
Certain angles were changed completely. The human character running on to the screen, the larger monster mauling the mantis, a cut added in the opening sequence to build more suspense for the reveal on what is chasing the human. The scene certainly flows better and now that I am content with the camera work for the time being, I can work on the hand-key part of my project.
The mantis is the start. As its momentum is already in the sequence from the pre-vis, I’ve started with animating its legs as it pursues the human. Reference footage of a real praying mantis running helped a lot in how I’ve made its legs take strides, always keeping to a front leg and opposing back leg moving in close time with one another.
I’ve also put in motion my session for shooting the humans motion capture data. An actor is in mind, now I need only wait for my slot to get the animation shot.
This entry should be short and sweet. The human character I purchased came rigged and skinned but the two creatures did not. This week I did just that. The only hitch came from my silly habit of going straight to directly manipulating the weights of the vertices on the model to the bones they are to move with. This caused hitches, drawn out sessions of starting over. I learned a lesson from this, don’t underestimate the value of envelopes. When I finally did the skinning using the envelopes to take much of the mess around out from the process the models were skinned to an OK enough standard in much less time. Envelopes, they are there for a reason!
With that done, I moved on to the pre-vis. The only animation factored on to the models at this point being movement in the roots alone. I main focus was the camera. Trying different angles, settling on the ones I liked. The sequence runs at 24 frames per second and I was introduced to my first technical roadblock.
The giant monster will pick up and maul the mantis, I need to figure out how. Other than that, things look alright!