NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Camera distortion of sky images. was Re: NG's "Midnight Fun
From: Marcel Tschudin
Date: 2010 Jun 16, 22:52 +0300
From: Marcel Tschudin
Date: 2010 Jun 16, 22:52 +0300
Regarding George's assumption: > I presume that before the analysis was made, predicted altitudes of the > Sun's centre had already been corrected for dip and refraction,.if we're > looking into such fine distinctions. After we realised that the measurements depend on the personal techniques and the personal skill, all the measurements were made then by Greg alone. In the photos the sun's edges and in this case also of the apparent horizon were measured. The corresponding angular height was calculated knowing the position of the observer and the time when the photo was taken, considering also refraction and dip. I actually hesitated first to use this procedure for calibration because refraction and dip introduce additional errors. When Greg made the photos he had therefore to verify that the horizon was "normal" in particular that Santa Ana was not blowing. Personally I preferred to do the calibration only using the sun's diameter as reference and by integrating over the differential scales. Since the calibration obtained from the sun-horizon photos showed to be sufficiently good and the final equations very easy to use we didn't continue further. The calibration graphs which I previously sent do actually not show the real amount of distortion. The almost linear relationship is rather a coincidence. I therefore attach here an other figure which gives a better impression on the amount of lens distortion and how it was measured for obtaining the differential scale (moa/pixel). This example is from the 200mm lens. Each data point corresponds to the mean value of 6 diameter measurements which results from differences of 6 measurements of the pixel position of the right limb and 6 measurements of the left limb position. The data at y≈1350 Pixels corresponded to the centre line. The black lines are polynomial fits to the data in the corresponding figure. The red data refer to the results from the fit to all data in the x/y-plane which here in the centre agreed better than at the outer borders. Marcel