Conference BackgroundTriangle Game Conference held its second annual conference in Raleigh, North Carolina on April 7th and 8th. The conference was held in both the Marriott City Center hotel and the Raleigh Convention Center across the street (and also connected to the Marriott via underground thoroughfare). The name of the conference stems from "the triangle" formed by the three cities clustered in the area - Raleigh, Durham and Chapel Hill. The relationship of which is expressed by this map:
Within this area you will find dozens of game development companies, with a heavy focus on middleware development from companies like Epic Games and Emergent. The area also hosts many universities that have game development programs like North Carolina State and University of Chapel Hill. What you have here is pretty much a self-contained and sustaining game development ecosystem as the schools and companies work closely together to foster development of the next generation of game developers.
Helping in this endeavor is the Triangle Game Initiative, which is also the driving force behind the TGC. The Initiative is a group of local developers not only looking to foster development at the education level, but the government level as well with tax breaks for local companies, for example. TGI is showing the rest of the industry what is possible if local developers band together for a common goal.
When TGI sought out sponsorships from local colleges/universities for the conference, they decided there was no reason to go with just one, but form a group program involving all that included some form of digital media education, which is pretty much all of them - NC State and Chapel Hill in particular have been investing in computer graphics since the late 1960s. NC State also has a Digital Games Research Center which is the primary focus for students attending for the Game Development and other related Bachelor degrees from the college. Wake Tech Community College, the largest community college in the region, offers an AAS degree in Simulation and Game Development. Chapel Hill, as mentioned, runs a computer graphics program and Duke University boasts strength in computational geometry and AI, with a fully-enclosed, six-sided virtual reality environment.
All these schools are participating in support of the Game Development University, which were targeted sessions that can be attended by those who have purchased a discounted Student Pass ($25/day) and provide students knowledge they can apply to their students and continued growth in game development.
Conference OverviewSince this was my first TGC, I can't compare it to last year's. I can certainly compare it to others, but that's not always very fair given that every city has its own personality. What I can say is that I had a really great time, met a lot of awesome people, and learned some new things.
The venue for the event was primarily located in the Marriott hotel's ballrooms and conference rooms clustered together on the main floor. A few panels and the keynotes were held across the street at the convention center in one of two rooms. Despite the rather small size of some of the Marriott rooms, I didn't have much trouble getting in to any of the sessions I was looking to attend when arriving reasonably early (~15 minutes). I do wish they had been able to set up registration in the convention center lobby rather than the hall outside the Marriott ballrooms, as that made things a bit crowded and confusing ("are you standing in line?" "no, just chatting with some friends, go ahead"). Also the signs for the conference were not readily apparent, at least to me. I must be too used to big banners with huge conference logos ? While I didn't mind traveling between the convention center and hotel for sessions (it's not a long walk or anything) it would be nice to have it all in one place.
TGC had a pretty sizeable expo thanks to all the companies in the area. I didn't spend much time in there but the stalls were all set up well and there was still plenty of room for people to mill about. Exhibitors ranged from schools to indie developers to robotics clubs to large companies. There was a networking lounge towards the center that had beanbags and chairs (and lollipops!) that hosted gatherings for students and pros alike to partake in various activities like resume reviews. Companies on the floor were also taking resumes from students as well.
Obviously it's not a game development conference without some parties, and TGC did not disappoint. The local IGDA chapter hosted an opening party at a private bar for anyone to attend, which was a wonderful networking event for any students that happened to stop by (I met a few). The parties the two following nights were reserved only for VIPs and premium pass holders, but that served the purpose of allowing the industry pros a nice tight gathering in which to power network over Mediterranean food one night and yummy appetizers the next.
Select Session CoverageIn addition to the short write ups you'll find here, there are videos for several of the key sessions on the TGC page. You'll also be able to view some more conference images as well. Also, Lewis Pulspiher has the information-packed slides to his presentation "What video game designers can learn from 50 years of tabletop games" available on his site. This is also the part where you all get to thank dgreen02 for his work in writing up all these sessions!
Breaking into the Games Industry (Gaiiden)
Go Procedural: A Better way to make games (dgreen02)
Practical Direct 3D 11 Tessellation (dgreen02)
How to Qualify your Game Publisher (dgreen02)
Raising Capital to Build a Game (dgreen02)
Breaking into the Games Industry
I've listened to a countless number of panels like these, and they all generally spout the same stuff - generally. Yes, this panel was no different, but there are always some new things that crop up here and there that make me sit up and go "hey, haven't heard that one before!". Sometimes it's even as simple as being worded differently but in a way that makes way more sense. The panel for this talk was huge - with seven people on it. Given that panels can quickly digress into la-la land, I wasn't holding out much hope for an hour long session. But I was pleasantly surprised at the ability of each panel member to stay focused, on topic and succinct. So that's mainly where a lot of the greatness comes from in this session.
The panelists were John Austin (VP Tech, Emergent); Keith Friedly (Designer, Insomniac); Sandy Dockter (HR, Atomic Games); Paul McLaurin (Red Storm); Tim Johnson (Recruiting Manager, Epic Games); Ryan Stradling (Sr Dev Dir, EA); Suzanne Meiler (Dir Design & Game, Vaco). Each started by describing how they got into the industry:
Keith graduated from the second class of DigiPen in 1997 and actually took a job over in Belgium as a programmer. Yes, he was so dedicated at wanting to get into games he actually left the country. He returned to the US as a designer to work for Sony around 2000 before ending up at Insomniac.
Sandy did an unpaid internship to get the experience she needed. She recommends you do anything you can to show what you can do, HR people love that.
Paul graduated University of Chapel Hill with a Masters in Computer Science, but it was a summer internship that gave him the experience he needed to land a job out of college, stating specifically that he learned more in that one summer than his entire college career (but still recommends going to school!)
Tim simply had the connections, knowing people at Black Storm when they were growing their company and asked to help.
Ryan had to take a round-about approach, having no luck submitting resumes out of college with a Bachelor's degree in Computer Science. He took a job outside the industry in 3D hardware until he was offered a position at EA 11 years ago. Now he runs the local EA studio in Raleigh. He says to keep the passion alive and always have something playable ready to show.
Susan went to school in Raleigh for communication and was out at SIGGRAPH where she made contacts, including the president of a local Raleigh company. Upon returning home she sent in her resume and landed the job.
John took up Susan's story and reminded everyone that networking opens a door, a demo gets you in.
Resume tipsKeith: Mods definitely count as games that you've worked on and should be included.
Sandy: I spend about 10 seconds on a resume. "Will work for free" is a gamble, sometimes does not match your actual worth. There is a fine line between persistence and peskiness, do not nag companies for a response, but don't let them think you're uninterested. Don't call, email first. Then call once. Show your work!
John: The purpose of a resume is to get you a phone interview, not a job. Relevant information only, no longer than a page
Tim: Make sure your demo works! No HR or recruiter will bother trying to troubleshoot a demo that does not work
Paul: Experience is key. Internship at big company or group project during school. School is important but experience is more so
Ryan: Do not use acronyms the HR person can't understand from out of industry positions. In general, no acronyms at all. Wants to see completed work
Susan: Portfolio of a few really great images vs. pages upon pages of stuff is key
Artist demo reel tipsSusan: demo reel is good, website is better. Don't build portfolio in Flash - can't see on mobile phones.
Tim: List out what exactly you did when showing stuff from group work, don't take credit for the whole work
Breaking from testing into developmentSusan & Sandy: Make people aware of your abilities. Make yourself valuable to the company. Don't just do your job and go home - show initiative!
Sandy & Ryan: However, don't focus too much on that higher-level position that you forget that you're supposed to be testing a game and finding bugs - i.e. your current job!!
Paul: Keep up side projects so that Q&A doesn't degrade your skills over time
Interview tipsRyan: Tape yourself on camera and have someone interview you to see how you come off (he did it and realized how stupid he looks most of the time). Make sure you talk to people and look them in the eye, give a firm handshake
Go Procedural: A Better way to make gamesDr. Paul Slavini of Side Effects Software treated us to a one hour session demonstrating Houdini, a procedural node based 3D animation and VFX tool for film and games. The software offers a wide range of functions and has been used in over 300 films, and has won an Oscar 11 of the last 13 years for visual effects. Houdini offers a free license plus users can upgrade for $99 to a commercial license through their Apprentice program. Houdini also has been integrated with Torque technology from Garage Games.
First we were shown a demo of a user-created real-time bullet simulation including- fracturing wall particles, smoke effects, the works. Then he revealed the author of the demo was only 15 years old, which professed the technologies' ease of use.
When it comes to creating procedural content, they used the analogy of baking a cookie - instead of focusing on the end result (the actual cookie) the focus is on the recipe and the creation process. The first example was a demonstration of a fracturing cube. A cube was created and placed in the world, and then a node was attached to the geometry using the real-time node based modifier list which generated random points on the cube. Finally a fracture node was then attached to the list - after creating the ground plane in the world, and enabling collision detection they started the simulation the cube would fall and fracture on the specified points, like magic. Changing individual elements of the nodes and restarting the simulation would yield different procedural results.
The benefits of procedural content are numerous - smaller teams, higher throughput, higher quality, parallel work flow, more iterations, increased usability and ease of tool creation. The whole mantra of procedural content creation seems to be "explore, create, refine". The practical in-game uses for these types of procedural effects are also numerous with applications ranging from visual effects, level design/construction, to character customization and animation.
At the end of the presentation Dr. Salvini showed us real-world examples of the technology at work in "Killzone 2" for Playstation 3. Anybody interested in procedural content should jump over to the Side Effects Software homepage and see what they have to offer.
Practical Direct 3D 11 TessellationIn the wild and untamed world of Direct3D 11 rendering, it was nice to get a heads up from Dan Amerson and Jeremiah Washburn of Emergent Game Technologies in this one hour lecture. Emergent Game Technologies is of course the creator of the Gamebryo engine.
The talk started off with a quick overview of the changes in Direct3D. In Direct3D 9 (D3D9) the pipeline is relatively simple with Vertex and Pixel shaders. In Direct3D 11 (D3D11) the pipeline gets quite a bit more complicated. We have Vertex, Hull, Tessellation, Domain, Geometry and finally Pixel shaders. Confused yet? Yea, this is a sort of complicated topic so I'll leave the task of teaching D3D11 to somebody else, the main thing I took away from this session ( being a D3D9 developer with no D3D11 experience ) was the large number of technical and art-creation issues raised by this new technology. At the very least we have to change some core aspects of the D3D9 art creation pipeline to get satisfactory results from D3D11.
Ok enough of that, so what is with this whole tessellation thing? With the addition of the tessellation stages to the pipeline we can computationally add geometric detail to rendered models in real-time. I suppose the best example would be a cobble stone surface such as the ones used in the demo shown to us.
Using D3D9 we could do things like normal mapping, relief mapping, etc. to simulate depth/height of the surface. In D3D11 we can dynamically add more geometry in the shader to achieve this effect even better than before. The geometry is best tessellated using distance based level-of-detail, also the angle of the surface (EyeVector (dot) SurfaceNormal ) is useful to modify the amount of a tessellation applied to a piece of geometry.
When it comes to actual creation of art assets for use in a D3D11 game engine with tessellation, you'll need to take care to assure the mesh is "uniformly quadrafied", this is hard to explain, but because of the way the tessellation works, if you want consistent results, your geometry will need to be just about uniformly broken into evenly sized quads. Gridded input surfaces are the best food for the tessellation pipeline it seems. You have to be careful with the way the texture UVs are setup, there are certain limitations in addition to careful consideration of smoothing groups and removal of height displacement on geometry seams. Fading in the displacement map heights should go some way to alleviating tessellation artifacts. Some other tips include passing in the screen coverage of the primitive to the tessellator. When it comes to rendering the shadow maps, using the low-poly un-tessellated control mesh is recommended.
These art creation issues are compounded by a relative lack of D3D11-centric-tools. There is the Nvidia Bake Tool, also Crazy Bump to generate the displacement maps for the geometry, in addition to Parallel NSight, and Perf Shader but due to the infancy of D3D11 rendering, we still have a ways to go.
How to Qualify your Game PublisherJay Powell of Digi Ronin Games gave an interesting one hour talk about how to vet your game publisher. The talk was full of good advice for anybody seeking a publisher for one of their projects. The main points you want to be concerned with are the publisher's:
- Financial Stability
- Distribution Abilities
- Acceptance Procedures
You'll want to investigate the reputation of the publisher as well. Are there other developers coming back for repeat work, are they just working with publisher in "one-offs"? Do research into the public opinion and the press opinion of the publisher. Multiple titles per developer is preferable, as are favorable reviews of their games. In addition be sure to investigate the communities of the games they have published. Any kind words or references from fellow developers are key factors to determining a publisher's reputation.
The amount of experience a publisher has in publishing games in your genre is crucial. Determine if the games they have published in your genre are successful, and look for examples from other developers in your genre. An added bonus would be if they have produces with a pedigree for moving games of your genre onto the marketplace.
Distribution is a major aspect of a publisher's duties; you'll want to see a strong track record domestically, and overseas. Ideally a publisher will have localized versions of their hit games in most major territories. Spend some time going out to the mall, and to major retailers and check to see if their games are sold there. Sure they're going to tell you they can get your game into "Retail Store X", go to store X and verify yourself. Look into the publisher's history of cross platform launches or ports. If the publisher also handles digital distribution check the digital storefronts for games in their catalog.
The producers employed at the publisher are a very important part of your interaction with them. Before you get in bed with a publisher you'd want to make sure they have qualified producers, look to see how long the producers have been working at the publisher. Check how many projects have they been a part of, what were they? You want to gather as much information as possible. How many games does each producer simultaneously work on at the publishing company? You don't want to see producers taking on more than 3-5 projects at once. How many projects are the producers working on now? Where do they fit in on the "food chain"? These are all questions you should be asking a potential publisher.
Finally the acceptance procedures of the publisher need to be scrutinized. You'll want to see if your contact in the publishing company is the final decision maker when it comes to signing your product. If not, who is? Who are the other people involved in the decision making loop for the project. When it comes to milestone payments, how long can you expect to wait for feedback or approvals on milestones? What happens if a milestone is not approved? Ideally you'd want a 1-2 week turnaround on milestone payment approval.
Raising Capital to Build a GameThis was a panel of investors moderated by Bob Pickens of CED. The session lasted 1 hour, and included Glen Caplan, Justyn Kasierski of Joystick Labs who are providing early funding for game developers.
I really loved this session, unfortunately there were not many people sitting through this one - but the people who were there meant business. I'll try to cover some of the more interesting points from the notes that I took. The discussion started out describing the current state of the market - it's tough to secure capital but it's getting a bit better according to the panel. The traditional publisher model that has been around for so long is struggling in today's climate.
So, what can you do to enhance your chances of getting funding? Well you need to have a good business plan and a good team. If necessary fill in members of the team to make the investors comfortable. When it comes to securing money from venture capitalists, they like big bets. It helps to look at things from an investor's point of view; they want to invest in a company not just a single game. In other words they want to back management who knows how to run and grow a business. This is how you should think of things, not just "how can I get money to make my game".
The old saying around the racetrack applies to investors, they bet on the jockey not the horse. You need to find a "champion" investor within the group, somebody who believes in you and who will defend your position when you're not around. If necessary you many need to go out and find a 3rd party commissioned representative (i.e. somebody with experience raiding capital) to interact with the investors on your behalf.
According to the panel there are investment rounds of different scales based on the types of investor involved. For instance Angel Investors will do rounds of up to $1M, where Venture Capitalists typically invest a lot more, millions of dollars at a time. There was a brief discussion about the impact of geographical location on investor attitudes; for the most part the south east is about 20 years behind the curve of technological investments. The panel advised to look to the west coast for investment money, suggesting companies on the west coast will be quicker to get the joke.
In addition to investment money from Venture Capitalists and Angel Investors; there are also government grants available to small businesses. Some include NC Idea, Movable Code, and SBIR Grants. In order to secure a government grant you need a "catchy differentiator" to make you stand out from the rest of the businesses applying for the grant. "One North Carolina" is a grant specifically for first time entrepreneurs. Becoming a Qualified Business Venture ( QBV ) allow your investors to receive a tax credit for up to 25% of money invested into the company. This certainly can't hurt things; the only catch is that you'll need to generate a financial review statement for these kinds of benefits will usually cost a company around $10k a year. I asked the panel if this 25% tax break applies to funds you invest in your own company, the answer is no. Only people who are not employed at the company are applicable for the 25% tax break.
It's not easy, and never will be - but if you present yourself as a well managed business entity, and you can win over some investors in a group, your chances of receiving funding should be greatly increased.