There are numerous method for capturing high dynamic range images for use as CG environment maps. I recently acquired a new method and I thought I’d test it against more tried and true methods.
Prefer to see a video? Here you go…
I recently purchased a Ricoh Theta S as an speedy way of capture reasonably high resolution HDR environments. This blog article covers my thought and comparisons with the cheaper ‘mirror ball’ setup and the more expensive ‘Nodel Head’ option.
After getting the Theta S I started looking at auto-bracketing apps on the Google Play Store but didn’t realise that the official Theta S app includes the ability to bracket shots by setting up multiple custom exposures. Moments later I was up and running and took images from all three setups to compare.
Left: Mirror Ball, Middle: Theta S, Right: Nodal Head.
These are tone mapped version so as to quickly see the dynamic range captured. The Theta S appears to have quite limited latitude in the sky, I believe I wasn’t using it’s fastest shutter speed, 1/4000 instead of the maximum 1/6400, I shall be re-testing in daylight soon. It feels like the fidelity of the image is definitely reflective of the price of each setup but how do they compare when used to rendering an object for both GI and reflections.
The top sphere in each image is a photo of the mirror ball for comparison, the second and third ball are CG spheres to show reflections and diffuse colour. Other than some exposure and white balance differences I’d be hard pressed to see such a difference in fidelity as before. The Theta S doesn’t have the hot highlights of the others but, as previously stated, this could be my error.