Originally Posted by Grif
But unless you design bad test set ups, the errors introduced by your test equipment are smaller by a, or several different orders of magnitude.
As mentioned in post #9, testing a lens requires a bit more than a test chart.
What we can do however, the next best thing, is record images on film, and examine those.
That introduces lots of different, extra ways to change the result, besides lens performance itself. And unless you can subtract those from the result...
... you're left doing relative testing. You can test a lens against another one.
You could declare one lens to be your bench mark, and see how others do compared to it.
Or you could test one lens at different apertures, with and without filters, etc.
Even then, you will have to make sure that, for instance, errors in focussing, variance in developing the film, etc. do not mask the thing you want to test.
The more we stray from direct testing, the lower the probability that what we measure actually is what we wanted to measure.
In optics, flare is non image forming light bouncing around eventually ending up 'on' the image, creating a general lowering of contrast.
Originally Posted by Photo Engineer
The thing that reduces resolution, the thing you test with charts, is spread. The phenomenon that an image of a point (or line) source never is a point, but that instead the light that should all have ended up in a single spot in the image is spread around the place it should have been in.
The result of that is that an abrupt transition from light to dark in the subject space is rendered as a more or less gradual transition from light to dark in image space. If the detail is smal enough, the light to dark gradients of adjacent detail mixes (what you call fill/bloom, PE), and detail gets lost in a gray blur. The end of resolution.
Diffraction is what is responsible for that.
Two quite different thingies. Despite chemists calling (according to you, PE) the second "Lens Flare".
Well, flare lowers contrast but it also causes spread and fill if you look at it objectively. The light is spread and unwanted areas in. It therefore lowers contrast.
Flare does not cause spread.
Flare is a different phenomenon entirely. Flare doesn't blur the edge of light-dark transitions. Flare doesn't change a 'block wave' into a 'sine wave'. Just lifts the amplitude of both trough and peak equally.
Flare doesn't make fine detail disappear (except in areas close to overexposure, due to the additional exposure). Flare does not reduce resolution.
Flare is not what you test using test charts like the USAF thingy.
Last edited by Q.G.; 09-30-2010 at 06:35 PM. Click to view previous post history.
Reason: Spelling of "trough" corrected (i hope)
Flare is defined very well here: http://en.wikipedia.org/wiki/Lens_flare
Basically, it places light where you do not want it. You are right. It reduces contrast. I should clarify then that flare "can" produce the effect of spread and fill if it is distributed properly by using the chemical definition. I am crossing terms that I should not cross.
Spread and fill in optical terms are different.
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
I had quite a few requests for the manual that outlines the old Modern Photography lens testing procedures. The size was too big to upload to APUG so I separated the portion that deals with the descriptions of procedures, plotting field curvature and the simple astigmatism test from the ratings' charts. The charts are the old magazine ranges of ratings for different formats/focal lengths that range from acceptable to excellent. I'll upload them separately for those interested... If I already sent the original to you, these are a bit easier to read because I bumped the resolution a slight amount and you might find these more pleasant to view.
I had a lot of fun with this test kit when I first got it many years ago (25?). I had kind of forgotten about having it but pulled it out when I happened into a number of duplicate optics when buying some Pentax 67 and 645 systems a few years back. I used it to cull the lesser examples and was surprised at the variation in the samples that I had. I've also used it to confirm a suspicion of a lens that's not as snappy as expected. My Konica optics (slr and rf) have been quite consistent and a number of medium format RF optics from Mamiya and Bronica have all tested at the high end of expectations, too.
One odd thing that I've noticed is that some of my favorite optics don't necessarily test out at the very top of the charts. They're quite good and fit into the high "good" and low "excellent" charts but deliver a look and snap that are very pleasing.