Organisation, synchronisation and transformation: With the footage safely held on many memory cards, it was now time to carefully transfer it to my computer, so I could begin the editing process…
Being faced with such a large task of preparing over 12 hours of footage from 6 different camera angles to something that could be watched on DVD required a logical approach to prevent any serious headaches later on. In total, the footage taken from the cameras on the night came to 58.9 gigabytes. As I knew that I may be up against it trying to handle this in Blender, I decided to split the task into two projects: first half and second half. Firstly, I filed the footage into folders based on the camera’s location and whether it was first or second half footage. I also renamed each video file in the same way. I was left with a more organised folder system consisting of: Audience, Centre, Left, Platform, Rails and Rear. I would use these words to define each individual camera during the editing process.
I had already prepared my Blender project beforehand, so I was now ready to load the footage and begin. I imported the first video file from each camera angle into Blender and positioned them on separate channels.
I was to have them all run together in adjacent windows so I could cut shots as appropriate. For the purposes of editing and while the recorded audio was being prepared by Matt, I would use audio from one of the cameras as a guide. In this case, I split the audio from the centre camera and placed on the bottom channel, underneath all footage.
With the footage imported for the first half, I faced my first challenge in the editing process. Obviously, the cameras all showed the same concert from different angles, but each camera began recording at separate times. I therefore had to align each camera angle with the same point in time as the centre camera, because this corresponded with the audio. Ideally on the night, I should have prepared a reference point for this, like a single light flash (such as a camera flash). Without this, I had to choose my reference point and align the footage to the nearest 1/29.97 th of a second (to the nearest recorded frame). This took nearly an hour because I had to make sure the footage was aligned correctly; there was no point going any further until this was done.
Here is a short time-lapse of me aligning the camera angles for the second half:
Transforming all the footage into something that was continuous, watchable, entertaining and focused on the best of the action made up the bulk of the editing that had to be done before DVD production. It was a multi-stage transformation process, passing through, reviewing and editing several times over.
Stages of editing:
- Initial cut
- Secondary cut
- Shot transitions
- Manipulating zooms, pans and other camera effects
- Final cut including removal of obsolete footage
- Testing and enhancement
I won’t go through each of these stages in great deal as most are self-explanatory. For the cutting procedure, I used frame markers and soft cuts in Blender to move each selected shot to it’s correct position on the final channel. Each and every cut, shot, fade, pan and zoom plus all camera motion was individually chosen and implemented. Though there could have been more cuts if there were more camera angles, there were still over 500 shot changes in a 2 hour final cut. Though there was more testing towards the end of editing, throughout the whole process I was constantly rendering out partial tests to check shot flow and timing.
Part 3: Final touches and DVD production