Post date: Mar 16, 2015 9:35:55 AM
Steven Russell
GDC 2015 Research Gathering + Things Learned
This year, I was attending GDC as a Conference Associate (CA), and was able to network with other CAs to make connections and learn more about what each of their branches of the companies they represent are specialized in, and how they contribute to the pipeline, along with other really interesting information regarding how they started up, developed updates, or generated content over time. Here’s some of my findings:
(Being a CA was friggin' awesome. Seriously. Would happily do it every year if I could.)
Making Champions From Concepts: Updating and From Scratch
Starting off with Riot Games, a studio that of which developed League of Legends, I was able to attend a few talks by them, as well as speak to several of their team members in person who were CAs as well.
Interestingly enough, the illustrative side of their departments has been incredibly busy lately pumping out splash arts of characters from within their game to showcase the characters, their story, and their abilities before you even look at the statistics. The reasoning for all the splash arts they’ve been creating lately is mainly to update incredibly old, previous ones. One of their talks was solely devoted to talking about the differences from then and now, and talking about what directions they went in to improve their work of “this single picture tells a thousand words.” Here’s an example of a before and after of one of the champions you can play as, Sion:
A sour-looking piece of art, huh? This was his old art, in which back when League of Legends was still a growing game from a no-name studio at the time, they could only provide paint-overs of their low-poly models. It’s a grizzly spectacle what with today’s tech and especially in comparison to the changes they made for him:
When the speaker at the presentation displayed this image after the previous one, people began to understand what he meant by major changes to the art. The Riot staff I spoke to further emphasized on what all they aimed for in the new art when I spoke to them after the talk, in which they said the major hits they want with their splash art is to create a personality from the get-go, displaying the character’s general emotional state, a less static pose, and displaying some idea of what their abilities could potentially be through painted effects or visual dynamic poses. In this case, Sion is shown standing over the viewer, stepping on the viewer’s chest, followed by the brutish hulking silhouette and mind-numbing gaze of an undead soldier from the pale skin, piece-together body parts with scars, and the soul furnace in his abdomen, an ability he has ingame, where he gets stronger from killing practically anything.
I asked how their development stages went, they said when they make new champions or environments, it jumps around to extremes from time to time depending on the parameters. Sometimes the abilities come first and the champion comes later, sometimes the theme of the champion comes first and the abilities that support it follow in suit. Normally when they try to imagine the size of these characters and view of them ingame from the top-down view, they’ll take a screenshot and paint the character over it. Here’s some early concept work for a void creature, Rek’Sai:
Notice the dynamic pose from the get-go? Though it lacks an environment, Rek’Sai was originally developed to look like this on paper, however as time went by and choices had to be made, the writers I spoke to said they felt the creature needed a gender, and the game designers saw flaws in the design with how some of the abilities would play out. Here’s the final rendition of how she (Rek’Sai) turned out:
Notice the similarities in the pose to keep the dynamic action? Now however, you can see her burrowing a tunnel, which is her prime playstyle characteristic - to burrow tunnels underground to move around. They added carapace coating to get the more alien-insect feel the void champions Riot has made seem to come with, and it seems a lot less human, more beast despite a similar body structure from the original.
Give or take, this was all gathered from several talks Riot Game staff members gave at presentations, along with information I gathered speaking to CAs who also worked at Riot Games. Overall, it’s interesting seeing their method of developing their artwork for their characters, even if it’s just the surface, and I’d love to know more in the future.
The Making of Unreal Engine 4
Next up - developing Unreal Engine 4. While I was working I was able to come in contact with a few CAs who guided me towards meeting some people who worked at Epic Games, makers of the Unreal Engine. I was able to ask questions on how they made the latest Unreal Engine 4 (UE4), and they were happy to provide info on how all they came into making it.
Believe it or not, UE4 wasn’t intended to be a separate installment of the Unreal Engine family. Rather, it was built in their office when they wanted to toy with the current functions of Unreal Development Kit (UDK) and see what all they could accomplish with making new addons or tools in a future update. That’s where things started taking off.
First off, the developers wanted to provide different material options for the upcoming changes companies and studios have been making to support them in the material editor - roughness, metallic, and refraction were some of the functions for example they wanted to iterate on, and make easier for the users to understand and work with, without breaking the system. They started approaching other tools the same way with this thought process, and rather than come up with office -made mods for UDK, it was suddenly a whole new system.
Granted before the realization that they made a new system occurred, one single man looked at Kismet, shook his head, and developed a homemade programming system utilizing the same C++ functions Kismet did, but in a cleaner, optimized way that didn’t require the user to bend over backwards - Blueprint. A single man who worked in the office programmed this entire system, towards which many of the coworkers found tons of use in. Around this time, the inclusion of Blueprint really pushed the old system to its limits, and the developers realized that this couldn’t be an update for UDK as slight updated modifications.
Enter UE4. The devs decided rather than stressing the older engine parts and rewiring the whole thing again to update the older versions, they issued out Unreal Engine 4 as a separate engine. It was updated in visual sophistication, addons, modified capabilities to manipulate environments, animations, and recordings in a phenomenal fashion, as well as sported a C++ editable system, Blueprint. Now Unreal became up to snuff with the industry as an updated system, than an outdated user-unfriendly interface.
It was cool hearing how Unreal Engine 4 was originally meant to be a mod built in-office for UDK that the devs went from “for-fun,” to “brand-new engine completely.” I’d like to have asked more questions, however I sadly wasn’t given the time to, but that’s what next time’s for.
Personal Take-Aways
Having been amidst all of those people, making connections and seeing the inner-workings of developers’ minds both indie and AAA, I’ve left realizing where I truly want to aim for in my post-college career as a Game Artist. I’ve had some phenomenal portfolio reviews from both people in the industry (actual VFX Leads!), as well as mentors through the CA Mentorship Program. Here’s my final take-aways a closing read:
Portfolio
Consider looking into Macro UVs for simpler-to-set-up-but-more-complex emitters
Look into throwing some animation in there for some of them, which can lead to the beginning and stronger build-up of the initial effect
Test yourself by seeing how far you can get away with an emitter in just the material alone, without the special nodes in Cascade - test, of course
Look into other VFX Artists’ portfolios, and don’t be shy to ask how they went about their process - normally they’ll be happy to find time to message you back
Look at some timing variation - don’t have a bunch of different VFX in the demoreel with “starting to be predictable” timing, in which all of your VFX have an impact on relatively the same size
Focus on adding STRESS to the last milliseconds before the final impact
Personal Realizations
Top company choices for future:
Mortiga Games (Gigantic)
ArenaNet Games (Guild Wars 2)
Riot Games (League of Legends)
Stylize much, much more with the effects - doesn’t hurt to have realistic weight and impact, but stretch the style further
Start building portfolio to be more magic-spell-y
Contact the Lead VFX Artists I came in contact with, and see if I can keep a flow of communication going, perhaps at least a mentorship, if not an internship