Cal -
I've got a buddy whose got the GW690 with an EBC Multi-Coating. I think this weekend I might do a compare and contrast between these four different lenses/formats:
Nikon F3 with Nikkor 35mm AIS F1.4
Leica M3 with Zeiss ZM 35mm F2
Fuji GW690mk III with 90mm F3.5
Fuji GL690 with 100mm F3.5
My buddy never really understood the differences between a 135mm frame and a 6x9 frame so I originally wanted to test that and show how these Fujis render. But now I also want to test the single coating versus the multi-coating.
If you were doing a test, what kind of tests would you run and how would you run it?
Say have these three scenarios, all shooting Portra 160:
1) Looking out my back porch, lots of green and muted yellows from the houses, focused at infinite, sun within the shot.
2) On the street, more saturated colors, sun from behind, focused at infinite.
3) On the street, muted colors, sun from behind, focused at infinite.
Each scene will be metered and then bracketed with three shots.
So I suppose the variables I really care about changing within my scenarios are:
- Color rendition
- Flare handling
- Contrast
- Image draw
Any suggestions would be helpful!
Jeremy,
The problem I see with your venture is too many variables. You are trying to compare too many things, and for any test to be valid one has to minimize differences.
As far as comparing multicoating verses single coating it is probably best to wait till I get a working lens back from Frank that has the AE glass installed. Then basically the only other variation might be the calibration of the shutters inside each lens (highly unlikely that the shutters will be perfectly the same). The idea here is to limit variations to be more scientific, otherwise data gets kinda not definitive.
In the French post I'm not sure how the results were guaged or if somehow processing or post processing of the images equalized the end results. Somehow with any evaluation some form of standards needs to be maintained to create an enviornment for fair judgement.
In my example above I am not using different cameras and I would use the same body, but even so the difference between the shutters could effect exposure and already that could be an unintended second variable that could skew results from being just about lens coating.
Another factor could be haze where one lens has been recently service and the othe might not be so prestinely clean. This could also be another secondary variable that will effect testing accuracy.
Then how do you measure accurately any difference, In the French post was the image scanned in the same exact manner with the same exact parameters? How was the image optimized? How big was the image printed if it was (Large prints show differences better than small prints).
In my experience focal length has an effect on contrast as well. To me longer length lenses concentrate light and shorter focal lengths tend to have a broader range of contrast. Think of a telescope concentrating light. A longer lens is a step in that direction.
Then there are differences in multicoating where Ziess lenses are known to be very contrasty. The point is not all multicoating is the same. Nikon for example continually advanced/changed their multicoating over the decades.
Testing is not easy to do. I used my Monochrom to compare B&W 2X yellow filters against Heliopan 2X yellow filter and found a dramatic difference. I used the same lens, same camera, same exposure, same scene, same lighting, same FOV, and used the histogram and 1% clipping indicators to measure the results.
My results were that the Heliopan filters worked better on my Monochrom by far margin.
Then I compared a Heliopan 2X yellow marked "DIGITAL" against a regular Heliopan 2X yellow. The point here is the only variable was the "Digital" factor between the two filters. I found another vast improvement as indicated by my histogram (no post processing direct out of the camera).
Later I would learn that Heliopan filters marked "DIGITAL" have additional IR and UV filters built in and are specifically made to notch out these bandwiths of light that otherwise would be processed as noise (unwanted signal that deteriorates IQ).
Unfortunately tests that cannot be controled experiments with only a single variable do not really or necessarily present scientific proof. Then there is the uncertainty principle that any experiment one concieves can actually can create a result that can be contradicted. Basically there is no absolute truthes and scientific thought is more like religion or philosophy (a system of beliefs).
Perhaps I have too much education; that's probably why I get confused. LOL. In the end people believe what they want to believe whether truth or lies.
Cal