London Small World was posted last week here on the blg. The 360 VR timelapse processing has been very tricky. Some result was on the blog earlier this month, but it was a complicated workflow and the result has raised some questions and eyebrows.
The main problem arises from shooting scenes with different zoom settings. At first the idea was to zoom right in to the mirror to allow for as much resolution of the focused 360 area as possible. This has become less and less important because of the fact that there is enough pixels anyway for HD use with 5mp capture settings and I also realized that the frame and the background, usually the sky around the mirror make for a good backdrop. The first trial was run on the bases of a cutout version of the mirror set in a white space. The HD dimensions did not really matter, later with the requirements of the background the dimensions and the proportions became important or rather defining.
Image by UrbanTick
Having this variety of zooms it makes it virtually impossible to get an insinc version out, using the current scenes. Solution, going out a shoot some more. It has to be seen as a go experiment and some conclusions have to be drawn from it. So, for me this means, think about the workflow and requirements of each step of the work flow first! And then go out and take pictures. Well it wouldn’t be fun anymore if we’d known everything beforehand, would it. The experiment is the exiting bit, isn’t it.
Anyway some more tests with a more elaborated workflow using FinaCut or AfterEffects to adjust and aline the image content.
The conclusion is basically that I will go out again and capture some more material. I will be using a more strict set of camera settings to make sure the raw material is easier to compile.