Hi all, Roger is right. The width/height ratio of the computer display can simply be changed by the manual controls below the screen. When sending pictures, or placing pictures on a website, one cannot be sure that the addressee or visitor sees what one intended him/her to see. However, the dimensions should be correct when measured in pixels.
Now back to perspective correction. Let's say I take a (digital) photo of a vertical dial up a wall. The location forces me to take an oblique shot. Let's also say I am sure that the dial (or its frame or setting) is rectangular, with horizontal and vertical edges. Now I would like to reconstruct the view, as if I took the photo straight in front of the center of the dial. That way I would be able to measure the substyle angle, check the correctness of the hour lines, etc. Is there a procedure that enables me to make such an exact reconstruction? My graphic program Paint Shop Pro X has a 'perspective correction tool'. I tested it as follows: I drew (in PSP) and printed a square of 150 mm sides and a rectangle of 100 mm wide and 200 mm high. Then I took oblique photos. The perspective correction tool produced rectangles which were always too wide. The error varied between 2 and 7%, depending on the amount of cropping applied beforehand. If anyone would like to give it a try, the oblique photos in their original size (1944 x 2592 pixels) can be found here. Only the contrast has been enhanced. square 150 mm (16 kB): www.fransmaes.nl/sundials/perspective/square-150.gif rectangle 100 x 200 mm (19 kB): www.fransmaes.nl/sundials/perspective/rect-100x200.gif Best regards, Frans Maes Roger Bailey wrote: > > What we see is determined by where we stand. This is a reality in so > many aspects of life. > > This applies to sundials. We were all misled a while back by the > distortions from perspective on an excellent sundial design. This was a > useful experience. I have recently be trying to correct for perspective > to analyze a historical sundial and and wish to share this useful > experience with you. The normal photo editing tools can correct for mild > distortion from perspective but generally fail to mathematically resolve > the distortion. We rarely know the input parameters, where we stood and > how we set the camera. Some lines can be made parallel and perspectives > skewed but are these true? Usually not. Faced with this > problem, I acquired new pictures of the sundial from known points of > view with minimum perspective correction. In the end this worked but in > the meantime I discovered that my view of the world through my computer > was distorted. > > I set the display at 1024 x 768 as this gave me reasonably sized print > and icons. When analyzing these sundial pictures I found that they did > not display correctly. Measuring my monitor I determined that the height > to width was 80%. But 768 x 1024 is 75%. Everything I saw on my screen > was distorted by perspective. An alternative computer system gave me > 65%. Changing to 1024 x 1280 gave me 80% and I could then go on with my > analysis but it caused me to think "Is what I am seeing reality". I was > fooled in this simple case. What about the rest of what we perceive > about our world? We can be fooled by perspective. > > Choose where you stand carefully and be aware of your perspective. There > are many distortions to reality. > > Again everything I need to know about life I learned from sundials. > > Regards, Roger Bailey > > > ------------------------------------------------------------------------ > > --------------------------------------------------- > https://lists.uni-koeln.de/mailman/listinfo/sundial > --------------------------------------------------- https://lists.uni-koeln.de/mailman/listinfo/sundial
