Jamie DeRuyter  GDC report 2016

Post date: Apr 05, 2016 6:25:12 PM

VR

It’s everywhere.  The prevalence of VR at the conference was very surprising.  ALmost every booth had some sort of VR demo playing. Everyone’s trying to figure out what to do with it.  Nordhouser is using it in his live concerts. Unfortunately I didn’t try any of the demos. I was hit hard by some sort of cold bug the day I arrived.  Putting those headsets on in a public venue would have been inconsiderate of me. I heard the bullet Train demo and the whole room solution were best of the show.  Opinion - This may be a fad like 3D stereo.  Hardcore gamers and tech nerds are going to buy them but I’m doubtful they will be in everyone's home next christmas.

Substance and Nvidia

The very first session I attended may have had the most direct takeaways that are relevant to the classroom.  Allegorithmic showed off version 2 of substance painter.  It has some great new features; Iray rendering integration, smart masks, clone tool, smudge tool, post process effects. Nvidia is pushing for a standardized MDL (Material Definition Language).  They claim ‘shading is broken’ and materials don’t transfer well between platforms.  The MDL define how a material looks but not how to achieve it.  It’s a set of standard values.  The goal is consistency and efficiency.  I’m not convinced that a standard is achievable or useful.  There are so many factors that affect the final look of a surface in 3D; cameras, lights, maps, post process, the output device and finally the biology of how our eyes see.  Margaret Livingston teaches, “Vision is information processing, not image transmission.”  

This segues into another talk where the presenter exclaimed that PBR is the latest buzzword but we need to move beyond to more address the larger picture...

The future of lighting in games

Jeremy Vickery  Assistant Art Director at Ubisoft

Optically Based Rendering - is what he called it.  I'm going to butcher this but...I think he was trying to say that PBR isn't the whole picture.  It's narrowly focused on lights and materials.  Other factors like the lens, your eye...the biology of seeing....gestalt principle of proximity...all affect what we see.  PBR is a nice buzzword but we need to move beyond to what's next.  Again, I heard people walking out saying things like, "he's crazy".  I'd have to hear it again and ask some questions to really wrap my brain around what he was trying to say.

I do know this...Nvida and Allegorithmic are attempting to push an MDL as a standard and I'm skeptical.  

How are they really determining the roughness of concrete?

By the time you calculate viewing distance, camera (every attribute), wetness, light, post, playback device...does it matter what the standard roughness value is for concrete?  I'm going to have to tweak it to achieve a desired result.  A senior had a good looking chocolate material ball in the material editor but had to adjust the roughness based on the shot (camera angle relative to the light, light intensity).  It was too hot.

Substance painter VS designer

Brad Smith, alumnus ‘04, gave a great talk on Naughty Dog’s use of the Substance tools on Uncharted 4.  The single biggest takeaway for me was a clear way to think about the application of these tools.  He said that they used designer on environments, painter on more central assets.  Procedural tiling materials are very useful to cover a lot of surface; ground, walls, etc.  Painter is used more for custom 0-1 maps for individual assets.  This is a very useful way to approach these tools and has already helped me advise the students during the learning process.

Photogrammetry

This has become a hot new technique to create natural assets.  Dice gave a talk showcasing their use of the tech on Star Wars Battlefront.  More here.  Some interesting takeaways...they claim it reduced the production time on assets by half with the biggest savings on the high poly.  Artists didn't have to sculpt assets from scratch. Vegetation doesn’t scan well. They used world machine to generate terrains and free height data from opentopography.org.  In the end, I must question this technique.  The results are impressive so there’s no arguing with that.  I question their assertion that it’s more efficient.  They bought a bunch of expensive photo equipment and sent teams of artists to some extreme location around the world.  I doubt their calculations included this phase of the production.  I think and artist could sculpt a rock before the team even arrives in a location to find a rock to scan.  In the end they created an entire level file with just a handful of textures and 12-20 assets.  With such a diverse and simple kit, I have to ask was it necessary to go through the process of photogrammetry?

Visual effects and player agency in Just Cause 3

This was a great session revealing a lot about the process of creating very sophisticated visual effects. Some key points:

• 1. Parameters - attributes from the world driving particle attributes.  An example of this is a car on fire.  As the car increases speed, the fire should react to the increased wind with reduced scale and increased spawn rate to fill in the gaps.  Current junior Trent Sivek has figured out how to do this exact thing in UE4.

  2. They treated the explosions as characters, giving them all distinct looks and feelings. Explosions were placed into 4 distinct themes; electrical, fuel, concussion, bavarium (nukes).

  3. Explosions are hard to create! (they relied heavily on real world reference)

  4. Rigid body destruction was created in Havok

  5. Fire effects created in FumeFX for Maya _ WE NEED THIS TOOL!

Maximizing Critique

Jeff Hesser from harmonix gave a great talk on critiques.  Some takeaways:

• 1. Approach critique as more of a dialogue and less like a one way street.  Move beyond giver to receiver.

  2. Critique needs goals

     1. Align the work with the established art direction

     2. Leave room for creative problem solving

     3. Encourage team spirit and cohesion

     4. Develop a desire for feedback

     5. Become a valuable team member

  3. When giving feedback consider these tips

• • 1. Define the problems before you start offering up solutions

    2. Describe the reasoning behind any judgements

    3. Avoid judgement focused feedback

    4. Use more reason or experienced focused feedback

• 4. When giving crit it’s important to recognize your first response, reflect on that, then rephrase that response using reason and experience

  5. When receiving feedback ask questions to define the problems and reasons.  Frame the critique. Look for themes in the feedback

The Killer portfolio panel session contained most of the same comments I hear every year, all good reminders.

Presentation is key.  Great art in a poor presentation won’t get you the job.

Show in game art.

Make access easy. Right click save images.

All that matters is the work.  Experience might not matter. If the posting says 5 years experience, apply anyway!

Tools don't matter

You're only as good as your weakest piece - less is more!

Epic has hired env artist with two environments, 3 images each.

Tumblr blog is not a portfolio.  Latest posts at the top might not be your best work.  It’s great to have a blog to show process but you must have a portfolio site with your best work right up front.

Ok to do fan art but you have to back it up with other work.

Environments need to be in stills.

While you're looking for work you should be replacing student work with new work.

Keep applying.

You are not capable of critiquing your own work.

Concept artists need to show idea generation, thought process, reference, inspiration

Your entire online presence is fair game

Don't be a dick

Effects in high demand, ui and tech

Art tests can be intentionally out of scope to see how you handle it.