Virtual Realality - Stitching
The starting point for stitching are the eight images obtained on location as described on the previous page. After loading the raw format images in Lightroom, I check the exposure to ensure consistency across the eight images, and adjust the highlights and shadows to ensure the exposure across the full range is balanced and accurately reflect the scene. I then convert the eight raw images to jpeg format with 100% quality. I could use lossless tiff format instead of jpeg, but tha would increase processing times considerably without any noticeable image quality benefits.
The eight images illustrated above form the basis for the next stage in processing, stitching the virtual reality. On close inspection, you can see that these images have a correction to the exposure, and are brighter and show more detail in the shadows.
Show two images to compare adjustements.
Loading the images into PTGui.
Generating Quicktime output.
Screen shot from PTGui use.
I use Pano2VR to extract the cube faces embedded within the Quicktime format to enable me to remove the tripod and tripod head from the nadir as illustrated on the next page. As illustrated on the PTGui website, you can also warp an equilinear jpeg to enable removal of the tripod and tripod head, and reverse the warping once done before producing the final Quicktime. However, this will resample the image twice and will, therefore, reduce image quality and overall sharpness.