For those of you who are
really curious about this Schneider 90mm Apo-Digitar, I did some more testing and loaded up some raw files. As usual, even simple testing is fraught with “error bars.” Basic focus errors are impossible for me to completely eliminate at this level, so please keep that in mind. All images include a focus ruler so you can check for yourself. I can see focus shift (or operator error?) even between f/8 and f/11, so viewer beware and have a look at the ruler.
This is the usual boring bookshelf subject. I apologize for that, but it is the best controlled environment I have.* The test goes basically like this:
Equipment:
Schneider f/4.5 90mm Apo Digitar (SAD)
Rodenstock f/5.6 90hr-sw (HRSW)
Alpa STC
IQ3100
RRS 34L tripod w/leveling base
Arca Swiss Cube
Shooting process:
Tethered to C1
Electronic shutter w/ 2 sec delay
Wide open down to f/11 in 1-stop increments
No filters
No LCC
Focus on LensAlign
Shoot, judge focus accuracy, reshoot if necessary
White balance on LensAlign
Shoot two series: One on-center with no movements, a second shifted horizontally 18mm, which equates to a 98mm image circle at the image corner. I moved the camera in order to keep the shifted scene roughly the same as the unshifted scene.
I usually shoot between f/8 and f/16. Even with C1 Diffraction Correction enabled, diffraction at f/16 is pretty clear, so I didn’t bother shooting above f/11. There is one image at f/13 with the SAD in order to see if there was any further improvement at the extreme edge, but I couldn’t detect anything significant. Note the SAD aperture runs in ½ stops from f/5.6 to f/32. There are two positions wider than f/5.6. Wide open is f/4.5, so I assume the detent between f/4.5 and f/5.6 is a 1/3 stop increment.
Results:
If you are in the market for this lens, I suggest you download some of the raw images and make your own assessment. Again, please pay attention to the LensAlign ruler. I do think a few images are slightly front or back focused. Here is an example at f/11. HRSW (left) looks back focused, SAD (right) looks front focused.
In general, the unshifted images are essentially identical at f/11. I am surprised at how similarly these two lenses render in both contrast and color. If you hide the image names, I wouldn’t be able to tell the full-scene images apart, aside from knowing the HRSW is very slightly wider.
Here is the shifted image with image circle diameters (not radius) written in to give you a reference of how it was shifted:
At lower apertures, my previous post summarized the results:
At f/11, The SAD is excellent out to about 72mm, where it slowly falls behind the HRSW. By 98mm I’d call it very good. The HRSW is excellent everywhere. Considering the official image circles are 92mm (SAD) and 120mm (HRSW), I’d say the SAD does remarkably well.
For me, this is where the rubber meets the road:
This is at the extreme edge shifted 18mm, so mid-90mm IC to 98mm. This is the difference between 0.9 lbs / $3000 and 2 lbs / $9000. If you need to shift this far and beyond, AND there will be important stuff in that area of the image, then the HRSW is clearly better.
Raw files are here:
https://www.dropbox.com/sh/kb4wc91hojt33ei/AACDtrdSqX3Q3YQIuNBHXsOva?dl=0
(well, they will be soon - 5 more to go; I’m uploading from a DSL connection…)
Naming should be straight forward:
Example: Cntr_Sk90_11_9200
Cntr / Edge refers to the image without shift (Cntr) and with 18mm horizontal shift (Edge)
Sk90 / Rodi90hr refers to the lens
f-stop is next
Image number
Dave
*Please don't judge my audio equipment; that little thing hasn't been hooked up in the 17 years we've lived here!