PROJECT CONTRIBUTIONS

IN DEPTH: PROJECT CONTRIBUTIONS

I was excited to arrive at Paradox and take over animation tasks on Crusader Kings 3.However, Crusader Kings 3 (CK3) was limited by an old 2d engine that was modified to work with 3d.

My first order of business was to streamline the export process. I believe in fast iteration, and getting the assets into the game to test quickly was critical. When I arrived, the process to see an animation in game could take up to 40 minutes. When I left the process was under 5 minutes.

        The former process consisted of these many steps.

        Adding 1 animation to the game involved opening 4 Maya rig files and creating animation names on the root joints of 4 separate rigs. We also had to make sure we were syncing to the latest rigs, and notify the character art team that we would be changing the rig files. Source control didn’t function so even if you had a rigged checked out, it could be overwritten by another dev.

        Once that was done, we had to open our Maya animation files and select the root joint for the head and body,  then keyframe the animation range on a specific animation channel that was added to the root joints.

        The export process also required the head and body to be exported separately. So after exporting the body, we needed to run a script to place the head on the ground plane. After baking the head animation and deleting any orient constraints we were able to run the script and set the head to the proper height. If the export succeeded we would now have 4 animation files. But we aren’t done yet, time to script.

To Script an animation we had to open 7 text files and modify them accordingly.  This doesn’t include opening several more text files if there happened to be a prop in the character's hand. Once all that was done, we could launch the game and see if our animations were in game.

        Keep in mind the animations weren’t fully scripted in game yet. This was just for testing purposes. The final scripting needed several additional states required to keep different body types/heights from clipping.  Something like if weight > x play fat animation if < x play normal animation etc.

        As you can see.  This process was very time consuming.  Additionally, creating up to 3 or more variations for each animation, male and female, was extremely labor intensive.  And oh yes the body types were scaled in the engine so I could not see them in Maya when animating. I had to guess how much to adjust each animation before exporting it to the game to actually check it.

As the single animator on the project in charge of all of this and more, it didn’t leave much time for acting animation..

        What to do?

        While I was onboarding another animator from a sister studio he mentioned he was actually a technical animator and not an animator. So, we discussed a way to transfer animations from male to female in the engine. This would cut the workload and animation data in half. So with the help of our AD, we were able to convince our sister studio to allow us to work on this small project together. Many games at Paradox were still using this engine so it would be beneficial to all.

        We quickly realized that the male and female rigs were in fact, not the same.  So we laboriously tested offsetting rotations and swapping names in script to make sure the animations could be transferred. We could not alter the rigs, since there were hundreds of outfits already skinned to them and the team didn’t want to have to re-skin them all. After some time we succeeded.  Which was a great help.  We were able to make one male animation, and in script transfer it to the female. We could of course still make separate female animations as needed.

        Step 2. Storing the animation name data in the rigs on an animation channel was not an ideal solution when dealing with hundreds of animations, it needed revamping.

I pitched using the built in Maya exporter to export game animation files. After putting in a request detailing the changes we wanted with the engine team, they decided to give us some time on this task. This now meant that I would just open the Maya exporter. Type in an animation name, a frame range, and hit export.  No longer needing to open 4 rig files separately. No more needing to hand key the range on the root joint of the head/body. No longer needing to contact the character artists that I was changing the rig file and waiting for a response.


        The second part of my plan was to automate the scripting part. Unfortunately the engine team was unwilling to spend time on this. However, I did not give up on this.  While onboarding the new shader programmer, I discovered that he had extensive experience setting up animation systems for a tv show.

He was also fairly surprised at the state of our animation systems.  He had a couple of days of on-boarding where he wasn’t really tasked with doing much. I was able to work with him during this time to get a maya tool that opened the text files and slapped in a test script with the press of a button.  So I no longer had to open 7 text files and copy and paste code in several places just to test an animation.

And so the animation testing process was now reduced from 40 minutes to about 5 minutes.


        While all of the above was going on, I pitched the entire concept, created and scripted expressive animations. What are expressive animations?
The game had static animations. Ones that would just play an idle animation in a pose.. For instance a guy looking angry would just be in an angry pose with an angry face, and just kinda stand there looking at you while you read the text.

Old static loops before/not my animation

New expressive anger animation that I created.. 

While looking through the animation script, I realized that there was a function for an intro animation.  This was just doubling the main animation at the time. I realized I could replace this double animation with an acting intro animation.  So now a character would emote before settling into the current idle animation.

        This was done without design or code needing to change anything in the game. I also figured out how to add 3 randomized animation intros to those events, I noticed the happy event animation played often. So I made 3 happy event animation intros and figured out a way to have the game pick 1 randomly to play. I also talked to design and got a list of the most frequently called animations so that I could make variants for them as well.

        Major Contributions to Wildstar 2007-2014. Many were done in my off hours and well outside my responsibilities as an animator.

This directly led to upper management focusing on enemy telegraphs. Leading Wildstar combat to eventually be completely telegraph based.

But let us take a step back. Once we were asked for more telegraphs I created a visual design sheet that allowed us to quickly talk about telegraphs and player movement.  All 3 examples I listed on the sheet made their way into the game. A Pacman shaped telegraph, a 3 bar telegraph, and a full screen telegraph with a “safe moving circular area”.

Pacman telegraph on my design document and in game, player is meant to dash into white area, the only place they don’t take damage.

Manager Comments from performance review:

        “The template sheet got people thinking in different ways when approaching creature telegraphs. Many people were focused on what the creature was doing rather than how the player was reacting.This added element of things to consider when designing telegraphs led to the creation of some of our most fun to experience telegraphs like the "Pac Man".You have been among the more passionate combat team members when it comes to submitting gameplay ideas. Some of your ideas have made it to the prototype state and beyond..”

        Additionally, I emailed the combat team 57 telegraph ideas that manipulated the player/UI in various ways. Many of my suggestions were implemented in some fashion by the end of development.

Hoverboards:

I lead the animation/testing of hoverboards in my free time, taking an abandoned prototype and tweaking and transferring 180 animations.  At least 200 hours of animation work was involved, working with the Lead Scripter, initially creating it as a costume for the race track instance so that no additional work was needed for it to appear in game.

When the hoverboard was turned into a mount, I took responsibility for all the changes that needed to be made to accommodate its new function. Adjusting nameplate, fx nulls, adjusting ride heights, modifying animations to work over water, creating new mount animations, working with props and concept art to create templates for board mods.

Hoverboards were then used in our vial marketing puzzle game. They were also featured in Wildstar’s most important beta. In “Winter Beta” they were used as a key incentive to help hit retention numbers. They were advertised in the patcher as an item you would acquire when you hit level 10. A Hoverboard video is on the main page of my website with the art director crediting me for its implementation.

FX:

Initially the game was going to use a procedural ribbon for weapon trails. It would require us to animate attacks in an awkward way to accommodate weapon trails and since the game was subject to lag spikes (due to it being an MMO) the trail would break during a spike. I proposed that we use geometry to create weapon trails, and this is exactly what the team would end up using.

Tools:

Created run speed tool that allows animators to match run and walk speed footfalls with travel speeds set in game.

Was the first person to use video recording on the animation team to document bugs for programmers. Enabling us to go frame by frame.

Contacted an outside technical artist through email to create a Softimage version of a secondary animation tool, Also later recruiting said Technical artist when he became available.

Worked with tech to create a ghosting tool, and orient constraint tool.

Projects|prototypes

Posed the Wildstar Box cover in 3d, during off hours under direction by the marketing team, the file was then sent to Petrol for paintover. Created many screenshots used for marketing materials, helped synergize scripting and machinima teams.

Created critical death giblet explosion prototype with the Rowsdower. This went on to be implemented on all enemies.

I animated giblets only in this clip

Prototyped dashes, combos,emotes, knockback, knockdown animations for player characters. Created Warrior 4 hit air combo attack after forum feedback indicated the main combo wasn’t exciting enough. Added several Medic class attack animations to the game to fill out the attack kit 

Created two storyboard/animatics as proposals for a launch trailer in 2009 during my off hours. 

Created a warrior sizzle reel during my free time that helped inspire and "set a precedent for in house video creation."- Cinematics Lead

Created initial kit for Draken male (was told my animation convinced upper management to keep the race in the game.)

Used rigid body simulation for Gronyx explosion death. This technique was also used on the critical hit giblet death for all characters in game.

        Research: Contacted former colleagues, sister studios, scoured gdc papers, and captured video footage from all genres of games. Reporting all my pipeline/gameplay findings to the animation team.

        Vehicle prototype: I was approached to help create a " Star fox type mini game inside of Wildstar. I created 11 animations with fx for the prototype. And transferred those to an additional vehicle model during off hours 

Animations: These are some contributions as an animator on the game.

        Created over 2800 hand keyed animations animating the base kit for 40 characters as well as adding animations to many others. 

Manager Comments from final performance review:

        “You have consistently demonstrated an interest in bettering the game. Often this is accomplished by your contributions as an animator but it also often goes far beyond that. You do research, participate in prototypes, pitch ideas and propose solutions to problems outside the scope of your job duties.

        (…)you have been good at hitting and often exceeding your deadlines and helping pick up the slack when others have fallen behind. You have also cared enough to point out and fix problems caused by others as you came across them in your tasks. You have been good about taking feedback and responding to it promptly without raising a fuss.

You[r] (...) knowledge of our combat system and what animation solutions we have at [our] disposal help make the art and design more harmonious. I believe you care about Wildstar and want to make it the best game possible.”