Continued from Page 2

Once all of the parts have been arranged as needed, I then render out a series of images to use in the instructions. I use a sort of cartoony lighting scheme with the ambient light cranked up as far as possible, with a distant light slaved to the camera. The resulting effect is to eliminate shadows and make it feel more like an illustration. I could do photorealistic renders, but the whole point of doing them this way is because I want that cartoony, illustrated feel for the step-by-step shots.

I also do the color scheme previews or parts manifest snapshots at this point, for the sake of visual consistency. I then save the images in Photoshop PSD format, and I do three things with them: I use them at full size as print resolution images for the instructions, I shrink the color previews and parts manifest images to 400x300 pixels for the gallery images, and I also create 200x150 thumbnails from the color previews and parts manifest images for the product pages. I use Photoshop's neat automation ability to make actions for those tasks, so it's often just a matter of clicking a few buttons and then wandering off for a cup of coffee while Photoshop does the work.

The next step is to prepare a scene for the model's "box art". This image is used as the cover for the instructions, the catalog thumbnails, and the product page blurb. This is something that I've wanted to do since day one, but it wasn't practical until I actually had all of the necessary props, tiles, and whatnot. Back then, if I had done it, I'd have had to make the diorama props available too, or I'd never hear the end of it in e-mail. So, it'd have required me to spend time working on freebies that don't pay the bills, or I'd have to somehow make a full retail set out of whatever diorama props I used, which basically made the whole thing pointless from an efficiency standpoint.

With the 2007-2008 product line being packed full of props, tiles, and stuff, that pretty much made using diorama-themed box art a snap, so I gleefully jumped on the opportunity to do so. In this particular case, I arranged some tiles, crates, and crate rows around the model, and then played around with camera angles and distances until I had the shot boxed in just right. For the box art renders, I went full-hog with Carrara's photorealistic rendering capabilities.

This part takes forever. The scene is illuminated by a HDRI map, which is basically a fisheye panorama image that stores much more intensity information than a normal image, and the effect of the HDRI map is to make the subject of the render appear to have been illuminated by that environment. Put simply, if you had a fisheye panorama photo of your living room, and you created a HDR image from it, then you rendered a chrome sphere in a scene that used the HDR image of your living room, you'd see your living room reflected in the sphere, and the sphere would appear to be realistically illuminated by the lights in your living room.

The scene also makes full use of global illumination and indirect lighting, which means that reflected light is bounced around the scene, and nearby objects illuminate each other with their own bounced light. In normal ray tracing, this subtle object-to-object lighting contribution doesn't occur, which is one reason why a lot of rendered images look glaringly fake and harshly shadowed. Rendering the box art image sometimes takes upwards of an hour or two. Break time!

With all of the necessary image assets finished, the next step is to actually organize and write the instructions. I use OpenOffice.org and Photoshop for this task, placing the necessary PSD image files into each page, and typing up the text. This part usually goes pretty quickly, since it's basically just a matter of putting the puzzle pieces together into a coherent whole. I take a break when I'm done, then I review it again with fresh eyes to make sure I didn't overlook anything.

Once finished, I export the instructions to PDF format, then I collect all of the product files together into a folder. This folder is zipped up, and I also zip up the instructions separately so people can download them before buying. I can't put 1600x1200 pixel close-up images on the website, but they can be put into a PDF, where you can zoom in as close as you like.

At this point, I add the product information to E-Junkie's seller admin page, and I then upload the instructions to the website's downloads folder, and the product file is uploaded to E-Junkie. The last step is to add the product's page to the site, and that gets uploaded at the time of release.

There you have it: the 12 steps it takes to get a model from my brain to your printer!

Back to Page 2