Chris_Malcolm
Edinburgh Art College viewed from Westport 2
This image is part of a focusing experiment. It is deliberately left largely unedited, as it came from the camera, with the exception that to clarify tiny differences in detail it has been upsampled to twice its original size. For more details see the description of the previous image.
I allowed the camera to decide all the settings for the previous image on full "camera" auto. The lens was set at its widest, 15.1mm (24mm 135 equiv). The camera chose for the previous image f4 and 1/250th, focusing on the middle windows of the college.
For this one I tried to improve depth of focus, and reduce chromatic aberration, by stopping down to what I hoped was a good aperture of f9. That gave me a shutter of 1/50th of a second, which I was confident I could hand hold with enough care for sharpness on this wide a lens.
Knowing the R1's habit of autofocusing on rather short declared distances on the wide angle end of its zoom, and regarding the horizon detail as the really critical part of the image I wanted to aim at, I set the lens on manual focus at infinity.
Comparison with the previous image shows that the only thing to have improved is the chromatic aberration at the edges of the image. Even with extra chromatic aberration the ribbing detail on the green dome on the skyline is clearer in the previous image, and the faint line of the upraised torch-bearing arm of the statue of the Golden Boy atop the dome can more clearly be seen in the previous image. There's also clearly more detail in the foliage on the tree on the skyline to the left of the dome, and on the spires of the building to the left of that.
Coming to the closer parts of the image, the window frames are distinctly sharper in the previous image, even in the near round window at the top right of the college.
Since nothing in this image is even as sharp anywhere as in the previous image, I conclude that either infinity as labelled is a slight overfocus (not uncommon in zoom lenses I discover), or else at this level of extreme detail I can't hold a wide angle lens steady at 1/50th sec.
Although both these images were taken from camera jpgs with sharpness on the lowest setting, after upsampling to double size I sharpened them with just one click on Irfanview's conservative very local sharpener. That only had a slight effect, because the detail wasn't in general quite sharp enough (at double size) for the sharpener to get its teeth into. It was also clear when sharpening this second (f9) image that there was noticeably less sharpening done, because there was less detail in the image to be sharpened. Consequently the first image had had a bit more sharpening applied to it simply because there was more sharpenable detail in it,
I then tried the experiment of viewing both these large images with autoresize to display the entire image on one complete screen, using Irfanview for this. It was noticeable that although it was no longer possible to put my finger on specific details which were present in one image and absent in the other, that even at that much lower display resolution (probably 1280 horizontal) there was a subjective impression of greater window frame sharp clarity from the first f4 image. That could of course have been due to heightened edge contrast still present at the reduction, and caused by the extra sharpening applied to the first image because it had more detail in it for the sharpener to bite on. Note too that although the camera was set to lowest jpg sharpening, it can be seen that at that setting it is still doing some sharpening.
More experiments later :-)
Original DSC04339
Edinburgh Art College viewed from Westport 2
This image is part of a focusing experiment. It is deliberately left largely unedited, as it came from the camera, with the exception that to clarify tiny differences in detail it has been upsampled to twice its original size. For more details see the description of the previous image.
I allowed the camera to decide all the settings for the previous image on full "camera" auto. The lens was set at its widest, 15.1mm (24mm 135 equiv). The camera chose for the previous image f4 and 1/250th, focusing on the middle windows of the college.
For this one I tried to improve depth of focus, and reduce chromatic aberration, by stopping down to what I hoped was a good aperture of f9. That gave me a shutter of 1/50th of a second, which I was confident I could hand hold with enough care for sharpness on this wide a lens.
Knowing the R1's habit of autofocusing on rather short declared distances on the wide angle end of its zoom, and regarding the horizon detail as the really critical part of the image I wanted to aim at, I set the lens on manual focus at infinity.
Comparison with the previous image shows that the only thing to have improved is the chromatic aberration at the edges of the image. Even with extra chromatic aberration the ribbing detail on the green dome on the skyline is clearer in the previous image, and the faint line of the upraised torch-bearing arm of the statue of the Golden Boy atop the dome can more clearly be seen in the previous image. There's also clearly more detail in the foliage on the tree on the skyline to the left of the dome, and on the spires of the building to the left of that.
Coming to the closer parts of the image, the window frames are distinctly sharper in the previous image, even in the near round window at the top right of the college.
Since nothing in this image is even as sharp anywhere as in the previous image, I conclude that either infinity as labelled is a slight overfocus (not uncommon in zoom lenses I discover), or else at this level of extreme detail I can't hold a wide angle lens steady at 1/50th sec.
Although both these images were taken from camera jpgs with sharpness on the lowest setting, after upsampling to double size I sharpened them with just one click on Irfanview's conservative very local sharpener. That only had a slight effect, because the detail wasn't in general quite sharp enough (at double size) for the sharpener to get its teeth into. It was also clear when sharpening this second (f9) image that there was noticeably less sharpening done, because there was less detail in the image to be sharpened. Consequently the first image had had a bit more sharpening applied to it simply because there was more sharpenable detail in it,
I then tried the experiment of viewing both these large images with autoresize to display the entire image on one complete screen, using Irfanview for this. It was noticeable that although it was no longer possible to put my finger on specific details which were present in one image and absent in the other, that even at that much lower display resolution (probably 1280 horizontal) there was a subjective impression of greater window frame sharp clarity from the first f4 image. That could of course have been due to heightened edge contrast still present at the reduction, and caused by the extra sharpening applied to the first image because it had more detail in it for the sharpener to bite on. Note too that although the camera was set to lowest jpg sharpening, it can be seen that at that setting it is still doing some sharpening.
More experiments later :-)
Original DSC04339