Frank's Sketchbook Entry of the Day

"Is that some sort of evil magic orb you're holding? What're you going to do, zap me with it? There's nothing you can do, you're just a stat--ayyyeeeeeeeeeeeee!!!!"

OK, I've kind of fallen a little behind in my pictures and this one was from October when I was visiting friends in the Boston area. Tom and I went to the deCordova Sculpture Park. When I saw this work (Eve Celebrant by Marianna Pineda from 1991), I looked at it for a few moment, looked around and then sighed, because I knew what I had to do. I handed Tom my camera and said take some pictures when I tell you. She was doing the perfect Street Fighter or Mortal Kombat move and SOMEONE had to get blasted by it. And there was no one else around but me who knew what kind of blasting was required.

So we staged a few tries with me leaping off the steps. But I'm not completely stupid. I wasn't going to fling myself back onto the hard ground which had very little grass on it. So we did a couple and they didn't look great. For the last series, I started lower, so I could throw myself onto the ground, but also twisted so that I could use my arms to help break the fall. And that worked, though I did hit a small rock below my knee and that fucking hurt for a few hours.

But there weren't enough frames to do a good animation.

It took me a few months to get to these pictures, and after reviewing them I figured a single photo could work. But of course it'd have to be a doctored photo. One of my poses was good enough to be used for me flying through the air, but that means I'd have to erase myself from one of the other photos, which means filling in the background from some of the other-other photos (turns out 3 of them). I'm sure AI could do it, but fuck AI.

So I had to create a composit photo by hand. In theory, it's pretty straightforward, just use masks to select the desired pieces from a few photos. In practice, since the camera wasn't on a tripod, the camera moved slightly with each photo in a sequence, and between each sequence the camera had moved even more significantly. And to add another challenge, each had a slightly different rotation too (since people don't hold cameras perfectly level).

But that all can be fixed by brute force stupid trial-and-repeated-error. Thus, I was able to construct a Frank-less photo. And then I cut myself out from another one and did a rotation. And then there was the tedious construction of the mask to get rid of all the parts that were not-Frank just at the n-F/F boundary.

Then came the fun. Since it was one image, rather than have a fireball that was flying towards me or had hit me, I decided lightning would be better, since there can be some continual zapping. I had wanted to add sparks, but the photoshop tutorials I saw suggested the way to add sparks to a picture was to find some sparks and add them to the picture (I didn't like that approach). All other Internet search results were on how to add sparkles. So lightning it was. I tried using two methods (generating with 'difference clouds' then adjusting levels, and selecting real lightning from photos and warping sampes as needed), and after doing both, I liked the cloud version better, even though it looked more like a magic blast than proper lightning, but in context, that fits better. I added a green glow to a few places on the statue and me, and an outer glow to the lightning, and thus it was complete. Simple...

And this is the full size image.