Upload & Sell: Off
Tobin - interesting set!
Cal - great colors and the B&W of the car for a moment had me thinking Cuba...
Very nice photos Charlie, Ron, Kape, Tobin! Great work!
I have a question regarding MTF if anyone cares to shed some light on the difference between sagittal and tangential measurements. I had the impression that sagittal graphs were directly related to the lens sharpness and resolving power while the tangential graphs were mostly relevant in determining lens aberrations such as CA or the lens rendering of OOF detail especially the way the the tangential graphs diverge from the sagittal ones. Am I correct and what is the true story here?
Edward, there was a detailed paper about reading MTF curves by Nasse of Zeiss, but I can't find it on the Zeiss site, so I've uploaded my copy, which you can access here (PDF).
He gets into sagittal vs. tangential on page 19.
The way I see it is as follows:
Sagittal values represent the ability to resolve lines that radiate from the center of the frame while tangential values represent lines at 90 degrees to sagittal lines at a given point in the image. If you were to have a small square object at some point in the image, whereby its edges aligned with the sagittal and tangential orientations, and both the sagittal and tangential MTF values were very similar at that point, you'd have an accurate representation of that square, with equal sharpness/contrast on all sides. If the sagittal and tangential values were significantly different, then the edges of the square that align with the higher MTF values would appear sharper/higher contrast than the sides that align with the lower MTF values, which would appear somewhat softer and lower in contrast.
This is probably an over simplification, but is how I visualize it.
As for OOF rendering... Nasse states on pg. 33:
"MTF only makes statements about the focal plane or its immediate surroundings. And in that case, a circular point spread function is indeed an advantage, because it reproduces small details in a way which is as faithful to the original as possible, with the best trueness of shape. This is important for the legibility of writing, for example.
It is not possible to use MTF data to draw conclusions about the brightness distribution of the strongly defocused point spreads, however. There are lenses with nicely parallel tangential and sagittal MTF curves but which are spherically strongly overcorrected. This correction state causes annular defocused point spread functions, which are visible as rim-lights and as pairs of lines and produce a restless-looking background.
This unpleasant characteristic cannot be deduced from the MTF data."
Edit: the link Derek posted is a really nice example with the spoked wheel illustration!