Good morning all.
I love doing sports, but it doesn't mean that I'm good at them. For instance, it takes a lot of balls to play golf the way I do.
Senior citizens are so wise, right?
As Parallax advised us in the Jokes Thread - don't wait until we're old and feeble to tell people how we feel about them.
By then, our arthritis may be too bad to let us raise our middle finger.
This week, I want to talk about how to glance at a company's MTF charts.
What quality can we expect when we buy a lens? Since all products are subject to manufacturing tolerances and variation, about all we can really know before hitting 'Add to cart' is how they're supposed to be. Online 'reviews' typically only test a single copy of the lens, and sample photos can be less useful than we think. Often, the skill of the tester is being measured more than the lens itself.
I actually think it's very handy to look at the MTF curves for the lens if available, then see what else is known on the web about it.
These are what the manufacturers themselves use for design and reference. They're the theoretical curves generated by their CAD programs for the perfect specimen of their new lens (Zeiss and Leica actually use real rather than theoretical measurements, they have so much confidence in their factories' quality control).
A curve for the Pentax FA 35mm f2 is below, a fine lens, small with a 49mm filter size, that I own myself. For years I have enjoyed it on Sony as well because it has a manual aperture ring.
It's for wide open aperture. This is most commonly chosen because we see any lens at its worst, which is very handy. It's what we need to know. Closing the aperture will improve almost all aspects of performance until diffraction softening really kicks in, typically from f16 and beyond.
The blue lines refer to how sharp a lens is contrast-wise, the solid and dotted lines being measurements taken at right angles to each other. We want the lines to be as high up the graph as possible, and for them to be together. Where they diverge, it is possible for aberrations (including chromatic) to be prominent, and for bokeh to be ugly.
The red lines are similar except they refer to how sharp a lens is with its details, its so-called 'microcontrast'.
The bottom of the graph, the horizontal axis, represents how far across the frame from the centre we are. You will notice that on full frame (22mm in the corner) there are the usual drop offs meaning softness and vignetting.
We can see the advantage of using a full frame lens on a cropped sensor camera, because that sensor extends only 14mm into the far corners, the best part of the lens is being used, so the image quality is still good.
As a note, the FA35 is interesting for such a small lens. It does not have severe humps and dips like some Zeiss and Leica wide angles do. Those humps can reveal themselves as field curvature, where some parts of the image seem sharp, others soft, even at the same subject distances.
It is possible to approximate these tests yourself with software like IMATEST, but your lab testing skills really come into play there. Focus must be absolutely spot on, the target must be well-illuminated without glare returning into the lens and the set up must be completely perpendicular. Experts like Digitalis will use an optical testing bench, often set at infinity focus, those at home or an office would have to settle for a certain distance, and the mount must be completely free of all vibrations, probably a rig not a tripod. Very difficult to do well unless you are rigorous and have experience. There would be very few Youtube vloggers I would trust to do the process without stuffing it up.
Many lenses don't have published curves - I think Canon do best at this, and they do so at different apertures, too ... not needed, but nice. We can quickly see what a lens is like and then see sample pictures and experiences of others afterwards as corroborration. We can even identify bad copies. The Cameraville Guy took down his original DFA*70-200 review down because viewers were able to tell him it was obviously decentred and not up to spec.
I really don't think any of the above is forbidden knowledge, ordinary photographers should be able to glance at published MTF curves and not be ashamed to understand what's going on.
To finish with, there's the story of the 89-year-old man stopped by the police around 2 a.m. and asked where he was going at that time of night.
The man replied, "I'm on my way to a lecture about alcohol abuse and the effects it has on the human body, as well as smoking and staying out late."
The officer asked, "Really? Who's giving that lecture at this time of night?"
The man replied, "That would be my wife."
Find the rest of the series here:
Clackers' Beginners Tips (Collected) - PentaxForums.com