#52. "RE: Soft Focus - How About A Standardized Test ?" In response to In response to 0 Wed 06-Feb-13 06:24 PM by elec164
This thread has surely taken it's twists and turns. Perhaps I'm being a bit pendantic here, but I believe camera bodies don't soft focus, lenses do. So first we need to know if its your intent to determine lens copy variation or body copy variation.
So let's assume you are referring to focus accuracy not lens resolving ability. Trouble with checking for focus accuracy is that AF systems in cameras are tolerable accurate not precise. So as I believe has been already stated, even with the same body with a fixed subject distance, if you focus, defocus then refocus multiple times you will most likely come up with a slightly different focus distance. You will also see a variation if you come from infinity inward or minimum focus distance outward. I also believe that wide angle lenses will show more of a variation than longer focal lengths.
For a good read on the trials and tribulations of AF testing there are two good articles over at LensRental.com located here and here. Particularly interesting would be the second article where they plot the variation and show how far off a bad copy would plot from the scatter of good copies.
There are commercial devices you could buy such as LensAlign or there is the free much maligned 45 degree angle test charts. But the trouble with self testing is in controlling variables and setup. Personally when I wanted to test a recently purchased lens for focus accuracy I downloaded and printed a standard resolution test chart, mounted it to a wall, aligned my camera as best as I could, lit the chart with an old photo lamp that my farther had from many years ago (insuring sufficient lighting and a good contrast target), pressed the shutter half way 5 times to obtain the best AF the camera could provide then observe the results in an editor at 100% view.
I noticed that the lens seemed to not be a good match to my D7000. So I then used Live View to focus, then switch to viewfinder and half pressed the shutter button and observed the distance scale. I then adjusted the fine tune until I no longer noticed any movement between the two focus systems. The results of my testing can be seen at my Pbase site.
On a Yahoo Forum I participate in one of the more esteemed participants suggested we use the moon as a target when comparing camera/lens ability. After all, no matter where you live it receives the same light and is for all intended purposes about the same distance providing the same perspective. But I don't know how well that would fit in with your request.
Bottom line is that optical testing is not for the faint of heart, and there are numerous variables that can give erroneous results. As such a comparison on forums such as this would be difficult at best, if not down right impossible.
Hope this helps.
Edited to add:
I was wondering about this and finally stopped being lazy and download the cylinder shot to look at the EXIF data. And to my surprise what I found plays into what Stan just posted.
You took that shot using manual focus!!!
So again, what are you referring to??? An image can be in focus yet soft (usually due to lens aberrations), but a missed focus image will be soft in your intended subject area, but sharp elsewhere (actual focus plane). Also being a camera with a AA filter the end result will be soft and need sharpening. And to that end different models will have different pixel densities and different amounts of strength in the blur of the AA filter making direct comparisons difficult.
The in focus areas of your example appear to me to be very sharp, but then with an in-camera sharpening setting of 9 that’s no surprise!! So again with the thin DOF of that shot you are going to get the results you did, and if the plane of focus is not where you intended while using manual focus it’s not the cameras fault.