Cobbler & The Elves - Process Work

I've really been inspired by a bunch of work posted to a few art blogs I follow lately, especially when they detail their process work - in particular, Ralph Horsely and Sam Kieth's blogs are wonderful resources for inpiration, tricks and ideas. Ralph is an astounding fantasy artist, which I'm a huge geek for. Sam was the artist and author of The Maxx, a comic that had a huge impact on me when I was a teenager, and he continues to create amazing mixed-media pieces on a regular basis. You should check out their work.

In an effort to share a bit more of my stuff and talk about how I work, I thought I'd try and do some of the same, although obviously not with work near their level. I just completed this piece below as an illustration for a class and had a lot of images available to show the progression - I've taken to working on a new level any time I reach anything that feels like a milestone so I can go back and see how things are coming. It's helped a lot in keeping me on pace and on track.

The piece was an illustration from the story of the Cobbler and the Elves. We were given three stories and told to find scenes from them we wanted to illustrate and take to a finished state over a couple of weeks.

 I began with some very rough thumbs, trying to find compositions and images that told the story well and would be a fun challenge to work with.

I ended up selecting this piece. Now, this is where I ended up causing myself the most heart-ache. Instead of going in and doing a tighter pencil drawing, I just used my thumbnail. Foolish, lazy mistake. Every time I do that in an effort to save time I end up causing myself WAY more work. It's the same way with animation - if you don't put in the planning work at the beginning, it'll cost you a ton more heartache later when it gets a lot more complicated. Had I been wise I would have tightened up and resolved all of the shapes, the detail and the perspective here rather then trying to do it in paint in Photoshop. Live and learn, I suppose.

To start, I laid down a warm medium grey/brown tone as my base and began painting in my base values roughly to get a decent feel for the illustration and composition. Again, had my pencils been better, I would have been able to solve a lot of problems here, such as the cobbler's wife, the drapery and filling in the dead space.

Once I reached a certain point I turned off my pencils so I could begin working with the actual values. This is really where you know if a painting will work or not - does it read without the linework?

Here I began increasing contrast slightly, defining a little more of the forms and adding in some fill light, building up the shadow shapes and trying to figure out a bit more of the forms.

At this point my instructor gave me a great little paint-over on my piece, first starting with a line sketch over the Cobbler, helping to round out his shape and add better detail.

He also did a scrub-in of light and dark to add better contrast and form to the piece.

I selected my old cobbler and just blurred him out, giving me a decent base to paint over without getting wrapped up in the details.

Here I went in and started tying down the changes to the forms, adding in light and form to the wife, the cobbler's head, arms and beard. I also added the clothing for the elves he was working on onto the workbench and filled some of the dead space behind him with a shelf of some kind. Lots of defining shape and form.

More definition, more contrast. I should have worked from more reference of candle-light, there is far too much ambiguous and ambient light for a piece that has the single-point lighting of the candle.

Softening the forms of the shadow, defining the wife and the basket, getting a bit more work on the Cobbler.

At this point, I decided the lighting was still too ambiguous so I just did a huge rough lighting pass on a new layer, pushing the idea of cast light and shadow and increasing the contrast greatly.

I then blurred that layer and used its information to paint up new light and shape into the forms.

Softening the light a bit more, giving a warmer feel and adjusting the levels for clarity.

A final levels and contrast adjustment and the piece and I were ready to part ways.

I wouldn't say this is my strongest piece ever, but I learned a lot from the process and enjoyed working on it a ton. It was one of the first digital paintings I'd done where I felt like I was in control of the process and enjoying it.

I did nearly the entire piece using a single brush, one I gleaned from a tutorial video by a stellar game artist, Jesse McGibney. He demonstrates the brush, along with a number of his other techniques, in the following video.

The brush itself is actually pretty simple - a rounded corner square with the direction set to follow the direction of the stroke, with size, opacity and flow set to pen pressure. That's about it, but it is remarkable how much it creates a feeling of actual painting.

This was also my first piece working with the new painting tools in Photoshop CS5. I have to say, I really, really like them. The new color picker that lets you compare your selected color to your previous color is an absolute revelation, and I can't imagine working without it any more. Great stuff.

Anyway, I hope this post was useful to someone. If anyone has any comments or other process work/tips to share, I'd love to hear them.

Salvage Youth - Week 9 Progress Report

It never ceases to amaze me how quick things happen on this project - every week I have to go back and look at what happened since my last report, and every time I'm amazed that it only happened in a week.

Like what, you ask? Well, we've got a logo now. Makes things all official and stuff.

We've already used it in some branded artwork hanging on the projects boards here at school - great looking stuff Beau put together.

Alexi has been hard at work as always, modelling up a storm. We've got a few more gameplay props being pushed through the pipeline.

We also did a lot of work this week standardizing sizes and shapes for our tiled terrain system - the basic unit is complete, with a rough texture pass done for it. I already tested replacing one of the sidewalk tiles for another shape, the banked sidewalk that allows driveway access, and things worked perfectly.

I actually had time to do some texture work this week, too. I love getting to polypaint, it reminds me a lot of painting miniatures and traditional models, something I almost never get to do any more.

I also spent more time playing with particles, getting the water spraying out of the hydrant looking more realistic. I also got to do a little animation - the hydrant now rumbles and flaps as the water shoots out of it, and when you fix it, the lid floats up and slams itself down on top.

If you look closely, you'll notice another new feature we've got working in the game - real-time shadow-casting. We need to spend time dialing in the lighting, and right now it severely slows down our framerate (we're still averaging around 80fps, though), but that's an optimization issue - every item in our game is casting real-time shadows. In later builds, once we get the layouts more established, any static object (items that don't move around) will get shadows baked in, so that we don't have to render their shadows every frame. It'll save us a TON of computing power.

We've got some other new features working, too - we have patrolling enemies (the robots above, although the art used for them is placeholder art for the time being) with incredibly intuitive tools for setting their waypoints, thanks to Ryan. We also found a great plugin for Unity called iTween, a free system that allows us to animate or constrain objects and characters along splines and bezier curves. This lets us actually move objects along complex curve systems and control speed, direction, all sorts of things, in a perfectly predictable manner. The tool is pretty robust, and it's already found its way into our main menu system and some of our puzzles. It'll also allow us to move the kids between lanes much more naturally as well as make their lanes curve from side to side around objects, rather then the perfectly straight lines they were previously. Good stuff.

Alexi's been working on getting the first car up and ready for texture/sculpting, which is great - this is one of the props with a lot of character we've all been looking forward to seeing come together.

He's also got a great little mattress all finished, ready for texture and sculpt. Meanwhile, Zach's been hard at work getting our fast kid, Dustin, ready for putting in-game.

At this point, Zach is switching gears - instead of getting the zbrush sculpt done, we're going to get the Stu character concepted and put him in some drawings with Jenny and Dustin, making sure that we are thinking about the color schemes and designs of all three kids at once, rather then waiting until the end to do him. Zach has been rock-steady reliable for his contributions since the start, so this change-up shouldn't effect our estimated progress at all.

We're really moving along, getting tons of stuff in-engine and in front of testers, and it's really going great. We are all still excited and motivated about the project, but we all know we still have a long, long way to go. Hopefully we can keep our enthusiasm up through the rest of the year, but with a project this fun, it shouldn't be too much of a challenge.

See you next week!

Salvage Youth - Week 8 Progress Report

I missed putting up a report last week, since we were focused heavily on preparing presentations for our milestone, but I'm excited about sharing all the progress we've gotten done over the last two weeks because we've gotten a ton of work done on the game.

First of all, we've made a huge step - instead of having art assets for our game, we've actually got art assets in our game. This has been a gigantic milestone for us, and is starting to make it feel like a real game rather then just a demo of the Unity Engine. The Art Team is excited, seeing their art actually in-game, and the designers are having a great time working with the assets to make levels for testing that are much more reflective of real gameplay. It's hard to overstate how critical this stage has been for all of us.

The art itself has been chugging along wonderfully - adding Dan to the team has sped up the process greatly, as he's taking over doing all the high-res sculpts of the prop objects.

We've been doing a lot of planning for the beginnings of our UI stuff, starting with the Main Menu. We've got a neat idea we're hoping to get in-game soon that should tie in well with our backstory and help explain a lot about the world in an elegant fashion. Beau's concept work for it has been gorgeous, as always.

He's been cranking out great concept work for other props and world objects, as well.

We're getting more world props built and pushed through the pipeline as well. Alexi is a modelling machine, he's constantly getting props built and ready for texture and sculpt. Unstoppable.

Zach's got Dustin's concept art nailed down and has already started modelling him.

Randy's got the first rig test complete, but we were given a bunch of information on rigging for Unity by a former classmate, Micah Zahm, that pointed out a few issues we need to address, so we're going back and revising the rig to ensure we won't have any problems getting the character in engine.

I spent some time last week working on building particle systems and getting the designers familiar with the tools. At the moment we currently have a "starburst" particle effect that creates a UI reward element for any time the player does something correctly, solves a stage of the puzzle or gets a collectible. It's already helped with some of our playtesting, giving important feedback to the player as to the context and consequence of their actions.

This week, however, I spent a lot of time researching - I came across an incredible suite of tools called Substance Designer by Allegorithmic that allows us to create modular, dynamic textures. This lets us make "dirty" versions of all of the props we already have in-game and scale between the dirty and clean versions in real-time in engine, as well as creating a ton of modifiable dynamic tiling textures using set parameters - in other words, it saves us a ton of time in our art pipeline and adds variability to our world that we wouldn't have otherwise without devoting an absolute ton of programming time we just can't budget for. Awesome stuff.

We've already gotten successful tests in-engine in controlling these "dirt" passes, allowing them to scale in real time, so we can have Jenny fix an object and reflect her effort on the model by cleaning it up over time. Really, really neat. The Substance Designer toolset is incredibly powerful, and I'm just now starting to get in and wrap my head around it, but I'm excited to spend some time with it and get the results of the work in-engine.

We've really been moving these last couple of weeks, making up for lost time from a few weeks ago when it seemed like everyone was out sick. We were a little behind for this current milestone, but we put together a very solid plan on how to get back on pace for the upcoming milestone in a couple of weeks, and we're already racing to get everything back on schedule.

Wasted Youth - Week 6 Progress Report

It's been another exciting week of progress, with at least as much work on the back-end planning stuff then on all the pretty art that's so fun to post.

First of all, we've got a new artist on the team! Daniel Samuelsson has joined Scrap Metal Games as a support artist, he'll get to start working on high-poly sculpts and texture work on props this week. This helps us fill the last major gap in our pipeline, and just in time, as it frees me up to focus on animation right when we're getting our character rig completed. We're all excited to work with Dan, he's a great 3d artist, go check out his blog.

That's right! Our first iteration of the character rig is nearly ready! We should have a prototype of it tomorrow to start testing, making sure that the system will work in-engine and we can test the last major hurdle of our pipeline, getting character animations in and scripted to be fully functional, right away. As I've been helping some underclassmen with their animation lately as a Projects TA, I've been itching to get a chance to do some more animation, so nobody is more excited about this turn of events then me.

We spent a lot of time this week trying to finalize the most recent drafts of our GDD and TDD, and are getting a very solid roadmap of where we're headed with the project in the process. We got some really valuable feedback from professors on how we can further refine them and where we can put more attention, and we're well on our way to having an exceptionally thoroughly documented project - with a crew ten students working on this game, often in their own time at their own houses, this is a critical component to success, and helps us make sure we're working on the highest priority tasks we can be at all times.

We also wrote up a complete roadmap of our roadmap feature commitments this week, giving us a long, thorough overview of every critical feature of the game we're planning on implementing and when they're expected to be deliverable. It really is great getting to look this far ahead and see how everything is coming together - especially considering we're still right on schedule to meet our current deadlines.

We had a lot of new concept work and art tests done this week - Beau has been pumping out concept art for new props.

He also worked on testing some grass and foliage alphas to put through the "painter" feature in Unity that allows us to lay down broad strokes of plants and other debris on terrain objects. We're planning on using this feature to help scatter random trash and stuff throughout our environment, filling the world way faster then it would be if we had to place each object by hand. He put a bunch of this stuff down in-engine along with some test models just to see how things might look in-game, then dropped in a first-person controller to let us walk around the models and see how things look!

We got some new art assets textured and ready for engine, too.

Zach got Jenny all textured, too. There's a few final touches to make, but she's nearly done! If the eyes and mouth seem a little out of place, that's because they are - they're going to be applied later as an animated sprite texture, allowing us basic facial animation without having to deal with facial rigging, a major time-saving feature.

And last but not least, we may have a new name for the game! Wasted Youth was always a working title for us, as we liked the idea of it but it has some implications we're not particularly pleased with, so without further ado, I present you with our newest name:

Whatta ya think?

Wasted Youth - Week 5 Progress Report

Another busy week, and a whole lot of progress. We're all getting very excited, as we're on the verge of getting the first art assets in-game this week, which will prove out our asset pipeline.

A lot of this week was spent on getting art assets ready for import into the game - we have a couple of props complete now with normal, specular and diffuse maps.

Our Jenny character has gone through the ZBrush sculpting phase and is now moving on to texturing. This is thrilling, as it means we can get our character rig built and ready for testing in the next week. Plus, she just looks awesome. Zach's done a great job on her.

Randy's been doing great work on the tank, too. The high-res mesh is complete, with placeholder colors to demonstrate how each plate is a separate object that can be colored on its own.

He also did some rigging experiments with the treads. We may use a path deform, like in the video below, or we may just animate the UVs for the treads - we haven't decided yet.

Alexi got the first house unwrapped and ready for texture tests, and we should have diffuse maps for it in the next week or so.

The house uses four different texture maps - one for the roof, one for siding, one for trim pieces and one for accessories like the porch, windows and chimney. This allows us to swap out a different roof style, for example, to create a ton of variety with very few assets.

Beau and I spent part of the weekend working on writing the script for the opening cinematic and promotional materials, and we're nearly ready to create the art for it. We're really excited about this stuff, but we don't want to give away too much yet. There will be more to show soon.

In terms of design, we've spent a ton of time working on improving the gameplay feel based on testing feedback, we've got placeholder UI in place and are working with the particle systems, greyboxing out the first few puzzles and getting the interactivity scripts such as scripted physics events and such working. We also spent a bunch of time finalizing our most recent draft of our GDD, which came out to nearly 60 pages! There's still a number of things we want to refine further, but it's been great having such a detailed blueprint for the project.

A lot of our meeting this week was spent looking at our task list and making sure everything was up to date. We're right on our expected schedule, and we may be bringing in another artist to assist with high-res sculpts and texture work, which is awesome. We've got a very busy couple of weeks ahead of us, but we're all motivated and working very hard to get stuff in-engine and get the game looking good. We should be able to get character animation tests in in the next week or so, and that'll be the last major pipeline test we need to resolve - if that works as well as we hope, there's nothing stopping us from cranking out art and gameplay through the rest of the year.