We've been off for a while now with life getting in the way, but we are back and have some amazing content to boot! What is that content you ask? It's my second attempt at a hyper lapse video. Some of you might remember the first one I created and are probably wondering what makes this one different.
For one, it's Christmas themed. Second, I learned a lot from the first attempt and there are some noticeable improvements. That doesn't mean it's perfect and I plan on getting better with each attempt.
The whole thing was shot over the course of three days in traditional time lapse/ hyper lapse style. What does that mean exactly? It means that I took over 5,000 photos in the process of making this latest video. Handheld and without a tripod.Only about 4,600 were actually used in the end, but there was no video recorded during the whole process.
Some of you reading might be wondering why I didn't just shoot video using a fancy stabilizer or gimble like the Came-Tv Single. That answer is pretty simple, I don't have one. That and even if I did, Disney doesn't allow hand held gimbles large enough to accommodate my Sony a7S II and lenses. Even if Disney did allow them, getting a walking path free of people inside Disneyland is next to impossible. Some of the shots had zooming for added effect and smoothly zooming all the out manually is a feat I have yet to see done by anyone without the assistance of a focus puller.
"Anyone could get similar results from a $300 DSLR with a bit of tender love and care"
It was all fairly simple. I would get to the park and look for interesting/iconic landmarks or things with a bit of that Christmas feel. Before I continue, I must note that I used the best equipment I had at my disposal. Anyone could get similar results from a $300 DSLR with a bit of tender love and care. Now, lets discuss the gear I used during the making of the video. I am fortunate enough to own a Sony A7sii. I have reviewed the previous model and was absolutely in love with it. The advantage with the A7sii is its incredible low light performance and 14 stops of dynamic range. Combine that with in body stabilization and I can use slower shutter speeds to squeeze a little more light into the image. Along with the camera, I would bring along 3 SD cards with 64 GB's of space on them to make room for the 14 bit uncompressed RAW photos. Let's not forget the 4 extra batteries because these battery life is terrible. Then I'd decide between my 70-200mm f/4.0 lens or my 50mm f/1.4 Canon lens with a Commlite adapter. Turn my camera on, turn on the digital level and guides and get to work. Selecting a specific reference point on a structure and keeping everything level works best. Then I would inch my way picture after picture until I had between 150-200 pictures to get 6-8 seconds of video at 24 frames per second. Something that should be mentioned is that having a straight path marked works best. I would use the patterns in the brick paths, rails, walls and other things present in the park to help me along. Those were helpful for about half of the shots, but other shots required me to guess what a straight pass looked like. Something I discovered was that sideways pans/movements are much more forgiving in post production even if you didn't use a path while taking the pictures. Segments shot without the use of a path ended up a mess and were unusable.
Putting everything together was fairly simple. I imported everything into Adobe Lightroom. Edited one photo and applied the changes to the rest in a sequence. I then took all of those and brought them into Adobe After Effects where I would stabilize the footage. Some shots, especially the sideways pans, were immediately stabilized by the warp stabilizer effect. The more pesky shots needed to me stabilized manually which was a pain, but necessary. Ran the manually stabilized sequence through warp stabilizer and I would have a mostly stable clip. That's when I imported all of the fixed clips into Adobe Premiere and I just plopped everything into place. The entire folder where I'm keeping the project is taking up around 300 GB of space with RAW,high-res Jpegs, and lossless video.
Pretty troublesome considering that there are apps that will simulate hyper lapse by recording video, taking frames out and then stabilizing the footage automatically. The only issue there is storage, flexibility, the choice of editing RAW doesn't exist and video is output at 1080p when I would like something higher res. Not only that, but not having optical zoom makes it much harder to get the shots that I want. Most of the hyperlapse videos produced on those apps tend to be made by people moving their phones at normal walking speed and it gives off the effect of video on fast forward and not really time lapse.
This second attempt was definitely more successful than the last and I plan on learning more. If any of you have tips or suggestions, please comment below and let me know how I can make my third attempt even better.