I am almost done. I think.
To be honest this is the first time i have put much effort into making any sort of a timelapse film. I have made lots of test reels, so this is a bit new to me and my workflow is not really that efficient.
As I was shooting I would stitch the images into a timelapse after each sequence for QC, but mostly I was just watching for flicker, movement, any jumps or skips. with exception of the eMotimo i used in a couple clips (only one which made it into the final film), I had designed/built all the moco gear and control systems myself, it was not an off the shelf turn-key package, so it was a very unproven system. I occasionally lumped some of them together here and there to mock up some ideas for a finished result.
So I had started to re-render a lot of the videos using the kens burns effect on the stationary camera shots to add a slight bit of motion. It took a while to get the effects I was looking for, the main problem being I was having a hard time getting full rez images to show through when nesting them. I would then have to blow the image up a bit to get it to make the movement, which reduced the overall image quality.
It is funny how frustrating Adobe can be if trying to figure out something on your own. The answer is not always that obvious. I am quite good with photoshop, I am getting better with Pre-Pro, and i really do not know AE at all. After struggling on how to keep the images in the nested sequence at full res for 5-6 hours over the course of a week, I found a tutorial that showed how to do it, it was so fast I almost missed it. Hard to believe it could have been so simple. Its amazing how sometimes even the smallest and easiest thing, if unknown, can cause delays of weeks.
Now that stumbling block is out of the way, Istarted re-processing all the images. The next major issue was dealing with varying color tones from one clip to another. I was shooting with 2 different camera’s and non-native lenses. I re-rendered about 15 clips from scratch and tried various methods of color correction. I was not overly enthusiastic about the results, and after stitching the rendered videos into a single flow, it was clear the color shift work was not going that well.
I pulled one of the images up in photoshop, and started working on trying to adjust the white balance which is not the easiest task at times. Also, I had found the action/automation process to be a bit clunky. The problem is i am dealing with JPEG not RAW images. Damn you Adobe for not properly supporting Fuji RAW images. However, being shot in the studio, I have ultimate control of lighting, so the extra leeway provided by Raw is of little or no benefit. And about 80gb of Jpegs would be about 150gb of raw files.
After a short struggle I figured out how to convert the image to a smart object, then i could use an ACR filter on it! Yes! This is whatI needed! I always loved using ACR as it was very simple to get very good results.
A few tries with automation to get that workflow down, and then run batch edits. …. on a total of 13,000+ images
So about 3-4 hours a day for several days, and I finally had all the color correct, fairly consistent exposure, and everything is really starting to match and look correct from one clip to the next.
So now I have gone through the same clips, rendered each one at least 6-7 times, I have edited over 13,000 images, corrected color, exposure, balance, etc. Holy shit this has been a lot of work. I seriously can not wait to get this god forsaken video finished and online so I never have to see it again.
One thing I have learned, I much prefer writing code, designing circuit boards, and playing with cameras and moco gear far more than i like to edit. But this was a first effort, I suspect next time things will run smoother.
For my next project I think i will buy another eye-fi card and i can just edit as i go, and work on the master sets rather than waiting for everything to be done. maybe I can get away with less editing. This has been one heck of a learning experience, and I now have a far better grab on how to run with my workflow. I can also avoid a lot of unrequired work now that I have learned a few more tricks in Adobe CC.
Last night I finally cut the first draft using the master sequences. And Icould barely stand to watch it. I am just so sick of looking at these images, I shot myself in the foot by waiting until i had everything done. I should have worked on this sooner. However I have got some excellent clips that I think people are really going to like. Some of the older clips did not make the cut, and I can clearly see a difference in overall quality from my first clips to my last. Everything about the later clips are better than the first ones. However I am NOT going to re-start this all over again. I am finished filming and very close to being done with this project. I have put a ton of time and energy into this, and I am not confidant it will be all the way up to my own personal standards, but this is the first actual timelapse film I have ever done. Not to say I am unhappy with how it looks at all. I have some excellent footage and cannot wait to share it with everyone.
I will take the lessons learned from this into the next project, which i think will be a bit more abstract. I want to create some very bold and stunning imagery, and take a bit of a departure of trying to make it look too natural.
So all I have left, is to review the draft cut, make some notes, go in and do some tweaking. I also want to audition some more music, the song i have right now is only 3 minutes long, i would’nt mind finding something that is more around the 3:30 mark. I suppose i could simply slow it down 10%, i doubt anyone would notice. The 3:00 is a little cramped and i want to get as much of the footage in as I can.
I have seen a lot of timelapse photographers stitch together various timelapse sequences and stick them to music, without putting a lot of effort into the flow, resulting in a 5 minute video that you get tired of watching halfway through. I want to make sure I get all the content I want into this video, and let it gracefully end before the viewer gets bored and stops the video. I think 3:30 is a good length.
Music right now is a big concern, im not that happy about the song i have been using, mainly because i have heared it so many times im sick of hearing it. I would not mind finding something refreshing. I refuse to use anything i cannot licence, while i have tried to seperate Biolapse from The Chronos Project, I would have a hard time defending the link between the two in court if somebody decided to sue my company for the use of thier music.
Plus using any material from somebody without thier permission is a dick move.
Hope to be done soon, and I hope to show you all something you may not have seen before.