NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Which diameter of the sun in digital photos ?
From: Marcel Tschudin
Date: 2009 Aug 21, 21:19 +0300
From: Marcel Tschudin
Date: 2009 Aug 21, 21:19 +0300
George, you wrote: > A discrepancy of 8% is hard to accept. Something is wrong, as Marcel > suspects. And such a big discrepancy should be easy to check for. Yes, the discrepancy is big. I thought I would have a reliable scale from my sunset observations. This scale is now put in question. It turns out to be not so easy, that's why I start to ask around, hoping that someone may have done a similar experience and may have even found the reason for it. > I suspect the angular calibration of his camera Perhaps he should check that > more simply, in terms of the angle subtended by an object a few metres away, > rather than by relying on distant topography. Near enough to that he can > check dimensions with a tape, yet large enough so that the camera focus > doesn't need to change from infinity, and also so that uncertainty in the > position of the effective optical centre of the lens is unimportant (within > much better than 8%). One reason could be distortion, but I wouldn't attribute it to barrel or pincushion distortion for the following reason: These optical distortions are important when referring to all of the image. The 6x zoomed pictures of the sun or the geographical feature (an island with a well defined shape) use only a very small part of the full picture size and they are usually near the centre. With the resolution used for those images the full size of the pictures is 958 x 1278pixels compared to the size of the island about 60 x 210pixels or the diameter of the sun about 105pixels. Within this small range these distortions are likely to be negligible. It doesn't seem to be a distortion of the computer screen when measuring the size of the geographical feature with GoogleMap. In order to verify this I made the measurement also with GoogleEarth which has a tool for measuring distances; the result was the same. > Alternatively, even though Marcel considers that the Sun image is not > overexposed, perhaps it is, and he is looking at a bright halo of light > around it. The scale derived from the sun is a statistical result from over hundred measurements, having a standard deviation of a little bit less than +/-1.5% or pixels. Indeed, I do also have overexposed photos, but they don't enter this statistics. The colour of the sun's circumference is a good indicator. Any suggestions for possible reasons are welcome! In the meantime I try to get a better estimation for this difference. Marcel --~--~---------~--~----~------------~-------~--~----~ NavList message boards: www.fer3.com/arc Or post by email to: NavList@fer3.com To , email NavList-@fer3.com -~----------~----~----~----~------~----~------~--~---