Week 3 Overview (July 27th, 2012)
Last week the team shot the images 42ft from the wall. Our final processed images lost detail and we couldn’t see any of the cracking we were looking for. This week we decided to shoot 20ft away and compare the differences in quality. We shot at F8 and ISO100 and changed shutter speed as necessary.
After creating our measurements and placing dots along the wall, we placed the chalk line 20ft from the wall. We took our horizontal shots, making sure we overlapped 60% of each frame. When we got to the portion where the wall curves, we stayed 20ft from the wall but moved along the curve every 15%.
As we began to lose sunlight, it became necessary to light the wall with our 400 watt HMI Digimole Par arc lamp. Shortly after setting up the arc lamp it began to rain. Prepared for the worse, Greg followed Megan with an umbrella, shielding her and the Canon EOS 5D from the rain. Dale followed us down the wall and hand lit every shot with the arc lamp.
When we began our processing of the images in PhotoScan, we processed all 108 images at once at high quality. Each jpeg was around 15MB. We used a MacBook Pro with 8GB RAM, a 2.4GHz processor and the processing time was over 40 hours. The software froze on us, so we came to the conclusion we needed more memory to process that amount of information. Our next test will consist of discovering what resolution is needed to get the information we need. This will also be a good test to see what distance we need to be from the wall to see the wall cracks in the final mesh.
The end of the week took us inside, to the Georgia O’Keeffe Research Center to do one RTI capture of “Church Steeple” by Georgia O’Keeffe. We filmed the entire capture process and will be releasing some great tutorials soon!