Posted 04 August 2009 - 04:32 PM
Posted 04 August 2009 - 05:28 PM
I recently had a new idea on an old project. What if I did three, color-wheeled monochrome exposures of each frame with my 4.5K Bayer scanner and jammed them together in post. How would separate 4.5K, RGB put together and what would my max possible res end up being? 13.5K? Would this be a viable way to stretch my DIY scanner into higher res results?
If it is in monochrome and you use a color wheel you will get a "real" 4.5K scan not 13.5K meaning that each RGB color will be 4.5K like a Arriscan or Northlight, etc. FotoKem has a scanner called the bigfoot (made by Imagica) for Imax scans which is 12K and that means 12K for each color RGB not bayer mask.
Posted 04 August 2009 - 05:44 PM
It's difficult to mathematically calculate exactly what this would give you.
If you could get a 4.5K non-masked CCD and did three (or six or nine, for HDR) exposures with filters, I would not consider the resolution of the CCD to be a limiting factor for even quite high end work. The rest of the rig, particularly optics and registration, is entirely another matter.
Posted 05 August 2009 - 08:53 AM
If I might get your best guess: Would it be better to get the RGB filtered image as monochromes then reassign the colors to them before blending? Or do it HDRI style and grab the RGB images as color images with the associated color wheel bias?
I've done some pixel offset experiments digitally. Wouldn't it be easier to do it digitally than mechanically? It's verlayed pixerls either way. Or is my gourd not thinking right?
Posted 05 August 2009 - 05:26 PM
It's verlayed pixerls either way. Or is my gourd not thinking right?