I ended the last blog with “simple?..” I can now attest that the next stage of the process is in fact not “simple”.
So when I finished the last blog I was in the midst of cutting out hundreds of tiny figures to prepare for shooting a stop frame animation of them. After setting up a little scene for the figures to dance in. I realized I had two major problems to contend with. Firstly matching the camera angles and movement in the source footage seemed to be nearly impossible, and secondly replacing the cut-outs physically in the space with the accuracy needed also seemed to be nearly impossible. To cut a long story short, I developed several mounts and spent ages trying to measure and recreate camera movement all to pretty poor effect.
<iframe src=”//player.vimeo.com/video/84604458″ width=”500″ height=”281″ frameborder=”0″ webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe> <p>Untitled from <a href=”http://vimeo.com/wilkiebranson”>Wilkie Branson</a> on <a href=”https://vimeo.com”>Vimeo</a>.</p>
Here the accuracy needed to make it work was just not possible for me to achieve. In addition to that, the amount of time it took and the fact I’m very reliant on natural lighting, meant that throughout the shooting the lighting changed to drastically. This would have been setting me and the computer up for a colour grading breakdown (mental) further down the line.
I was despondent. For several hours I just sat looking at all my hard cutting-out work, kind of pretty and interesting but now redundant never to be brought to life by the magic of stop motion.
But I didn’t give up.
It was clear that, not only was my current work flow going to take me the best part of spring to realize, but that also it wouldn’t work. Plan b. was to try and do the whole thing digitally. Originally I was quite keen to do it physically, so that it had an intrinsic realness to it. But if that wasn’t going to work and I wasn’t going to give up it had to be the computer.
So my new plan was to jump back in the work flow at the rotoscoping stage and do it all again more accurately (as unlike before there was no longer going to be a subsequent stage where I would cut and refine this). Once everything was cut out on the computer I was left with the outline of the dancers that I could put in front of any footage. Next I went back to the little scene I created in the house and tried to recreate the camera movement of the source footage in real time with my camera and two pieces of slidey broken mirrors. I placed a couple of markers down so I could later track my camera movements in After Effects – these later proved redundant but it looks like I know what doing with little blue crosses on the set.
Next I brought this footage into the computer and got after effects to tack the movement of my shot. I also tracked the original camera moves from the source footage of the dancers, and kind of swapped them over in a manor of speaking. The effect of this being that the net camera movement of both sets of footage is together and we don’t have dancers too obviously floating off into to space where they shouldn’t be.
That’s basically as far as I have got. It remains to be seen whether grading and postproduction will be able to give the dancers the look that they are physically there in the space. But I will keep going and see how if I can make this work. In the mean time here is some test footage from the rotoscoped attempt.
<iframe src=”//player.vimeo.com/video/84604459″ width=”500″ height=”281″ frameborder=”0″ webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe> <p>Untitled from <a href=”http://vimeo.com/wilkiebranson”>Wilkie Branson</a> on <a href=”https://vimeo.com”>Vimeo</a>.</p>