"release" from Unity Developer Magazine

I've had a few questions about how I did this, and thought I would answer them here and share it in the showcase for people who don't have the magazine.

So, one of the things that was a challenge throughout building the first issue was getting screenshots of everything to really look good -- obviously the technical considerations when producing artwork for a game are different than those for print. But, I knew that Unity could produce really beautiful and interesting stuff that would be appropriate for a printed medium if realtime playback wasn't a consideration, so I really wanted to do a feature in the magazine where people could pick a topic and create sort of a visual essay and do the kinds of things using Unity that you wouldn't normally get to do.

I contacted a few people but timelines didn't work out and so the spot was just sort of sitting there as we approached the deadline. It was the end of a rough week -- right before going to press, and all of the text in the magazine had reflowed when I made an adjustment, and I had spent the day leaving voicemail messages with a printer trying to determine how some changes would influence printing costs... a bunch of really non-creative stuff. So I said "you know what, I'm going to give this one hour to try have some kind of creative release, and I'm going to take a screenshot of what I have at the end of that hour, and it is going in the magazine."

Back up to when I was first learning Unity -- there was a thread about creating atmosphere stuff -- day/night cycles and clouds, and I had been experimenting with ways to get high performance clouds based on that. What I had found was that by starting with a plane and deforming it so that it was slightly convex - like a lens towards the user, and by layering a series of these with adjusted rotations, you could create some incredibly believable clouds using modified particle shaders. The whole thing was low poly, used few textures, and was able to have a very cool visual effect when point lights were placed among the clouds -- it was a very believable lightning. That project file was sitting on my hard drive, and I opened it and started there.

I obviously adjusted textures and lights on things -- I was venting a good deal of frustration and was feeling rather "dark". I was also experimenting with DoF from the shader replacement for 2.1, and adjusting values so that "clouds" at various depths had different degrees of blur to them. I don't know how that came across.

Anyway, after an hour, I built the project adjusted the camera and took a screenshot. That screenshot, without any post-processing is what I used (attached here). I really hesitated putting it in at that point, because it was expressive enough that I knew that I wasn't a good gauge of whether it was any good or not, but I got some quick feedback that consisted of "well, it doesn't look like anything rendered in a game engine" and since that was ultimately what I was after, I went for it... I also figured that in the worst case, it set the bar low for anyone who wanted to give it a try for issue two :wink:

 

That's what I thought about it. I also like it. However, my impression was that it probably took a lot of work, and would have been created faster with other tools. What are your thoughts on that? If this could be used dynamically in real-time, even, say, as a kind of lava lamp you could project onto a wall of a club or something, I would be all for it. But using Unity for static images, that I don't quite get.

 

It took an hour, but it was an "art" hour rather than a "production" hour -- time seems to pass very differently when you are just creating something open-ended than when you have an intended end state.

Ironically, even though my final thoughts were to create something that wouldn't be limited by real-time considerations, it was incredibly high frame rate to navigate around (it is based on real-time clouds after-all). But, since I was staging a scene there were things like clipping that I didn't have to worry about -- if it were to be viewed from different angles, you would want a different scene, but the technique would stay the same.

Ultimately, I think that part of the goal of using Unity for static images is that when you use tools to create things that they weren't intended for, you end up discovering new things about them. It is something I first noticed while doing motion graphics for a video project a few years ago -- After Effects failed to launch (repeatedly) and I couldn't find the install disk. I also had to have something to show, and so I improvised with Flash and Photoshop (this was before Flash could do cool things like blur text or have different blending modes). I created my animated text layers in black and white in Flash and exported them as a series of stills and then used batch processing and scripting with Photoshop to use each still as a mask for creating the effect I wanted. Over lunch photoshop crunched away at about a minute worth of single frames and that afternoon I sent a finished video to the client. (and, a few hours later, found the install for After Effects and kicked myself for not looking in that specific drawer.) The whole process took more time but it gave me a new respect for the software -- a new understanding of technique and ultimately made me better at using it on a daily basis I think.

 

That's a really cool image. Nice work

Your comment about the "art hour" vs. "production hour" is so true...and profound.

Never underestimate what you can learn by experimenting, especially when using tools for things other than their most common purpose. I fully support what you're doing! Keep it up!

 

Nice! That is my new wallpaper. I will have to sign up for the magazine soon.