Schneider Digitar lenss?

[Dave Carter]

HI All,
I would really appreciate some input on this as I need to improve the resolution of an image.   I would like to capture more fine detail in the file.  Cost is not a big issue.

Currently I am shooting a specific landscape shot with a Nikon D2x with an 80 - 400 mm zoom lens set to 150 mm.  I am using f16 because of depth of field issues.  The camera and lens are on a very solid platform and I use mirror up shooting.  I process the file in Nikon Capture and CS2.

With all the discussion around the article by MR with respect to maybe the lenses are now limiting the new digital backs - (The Nikon D2x has 5.5 micron pixels).

I am wondering if you think a Schnieder Apo-Digitar  5.6/150 mm lens with a Schnieder Helical mount mounted precisly at the correct distance from the chip face would provide a better starting file from the D2x?

IF you think it might, I will be happy to do the experment and post the results.

Or, any other suggestions?

Thanks, Dave

(Sorry for title mispelling - but I do not see how to edit it).

[BJL]

HI All,
I would like to capture more fine detail in the file. ...
Currently I am shooting a specific landscape shot with a Nikon D2x ...
I am using f16 because of depth of field issues.
[a href=\"index.php?act=findpost&pid=62301\"][{POST_SNAPBACK}][/a]

Open up to at least f/11 (the limit suggested by Nikon author Thom Hogan for the D2X) or maybe f/8: smaller apertures like f/16 are producing diffraction limitation on the resolution capability of the D2X sensor and its 5.5 micron pixel spacing. The rough rule seems to be that aperture ratios higher than twice the pixel spacing noticeably reduce sharpness through diffraction effects.

P. S. It will not change the diffraction/DOF trade-offs at all to change to something like a 5D with its 8.2 micron pixel spacing: you could then go to about f/16 before suffering from diffraction limitation, but the need to use focal lengths larger by a factor of about 1.5x gives the same DOF at f/16 on the 5D as with f/11 on the D2X. In other words, trade-offs between the aperture size effects on fine detail from DOF on one side and diffraction on the other depend only on the angular field of view, not the format used.

[BernardLanguillier]

Ultimately, the only 2 solutions to have both a lot of apparent DoF and a very detailed image are to use the optimal diffraction aperture of the lens (f11 on most DX Nikkor) and:

1. use camera movements (view camera or T/S lenses),
2. do DoF stacking with softwares like CombineZ5

Both have limitations though.

Cheers,
Bernard

[Dave Carter]

Thanks BJ and Bernard,
I will test a few shots at f11 with my current lense.  Even at f16, I stack pictures for the depth of field.  So, moving to f11 or 8 will only mean stacking one or two more images.

I have looked at using "Helicon Focus" for auto-stacking, but currently I just do it very carefully in CS2 with layers, etc.  That way I can very carefully choose the mask locations.

I have also ordered the Schneider lens.  I do want to see if there is any difference for my application.

Thanks again, Dave

[madmanchan]

Thanks BJ and Bernard,
I will test a few shots at f11 with my current lense.  Even at f16, I stack pictures for the depth of field.  So, moving to f11 or 8 will only mean stacking one or two more images.

I have looked at using "Helicon Focus" for auto-stacking, but currently I just do it very carefully in CS2 with layers, etc.  That way I can very carefully choose the mask locations.

I have also ordered the Schneider lens.  I do want to see if there is any difference for my application.

Thanks again, Dave
[a href=\"index.php?act=findpost&pid=62401\"][{POST_SNAPBACK}][/a]

You might also consider stitching multiple images if your particular landscape situation allows it (i.e. relatively stationary subjects).  Like all other techniques, there are limitations to stitching, but resolution isn't one of them.

Eric

[dazzajl]

I am wondering if you think a Schnieder Apo-Digitar 5.6/150 mm lens with a Schnieder Helical mount mounted precisly at the correct distance from the chip face would provide a better starting file from the D2x?

Without a doubt the quality of the Digitar lens will exceed the Nikon one. If you can find a workable set up for your camera then I think you'll be most impressed with the resluts you get.

Perhaps you'll let us see the before and afters too.  

[Gary Ferguson]

IMO the Schneider Apo-Digitar is not the best way to go. I've no experience of that particular lens, but I've plenty of experience with Rodenstock Apo-Sironar Digital lenses with a P25 back.

Firstly, I'm sceptical that even the very best large format digital lenses materially exceeds the performance of the best 35mm format lenses. And secondly any performance advantage they may enjoy certainly reduces as focal length increases. By the time you are at 150mm you have some gold standard optics at your disposal that already come complete with perfectly fitting mounts and built in helical focusing.

[bob mccarthy]

Another potential issue is the Digitar's massive overcoverage of the DX sensor will flood the light box with plenty of excess light to bounce around. The digitar lenses are no doubt spectacular over the surface of a "large" sensor. Over  the area of  DX, Nikon's best should outperform. Nothing wrong with a 80-400, but its a "good" lens not a great lens. There are better choices in this FL (150 ish) that would equal or more likely better the large format lens.

Bob

[BJL]

I'm sceptical that even the very best large format digital lenses materially exceeds the performance of the best 35mm format lenses.
[a href=\"index.php?act=findpost&pid=62470\"][{POST_SNAPBACK}][/a]

I am inclined to agree. Though I have not seen MTF graphs for Nikon lenses that can be directly compared to the new wave of "digital view camera lenses" from Schneider and Rodenstock, I imagine that Nikon DX lenses at least are comparable to Olympus 4/3 lenses, and comparing 60lp/mm MTF curves for those Olympus lenses and for the digital view camera lenses, there is no advantage to the latter ... and obvious disadvantages, in loss of camera-lens communication and automation.

P. S. One consistent feature of the Schneider Digitar 60lp/mm MTF curves is evidence that diffraction limits f/11 to lower resolution than at f/8 or f/5.6. F/11 consistently gives lower MTF than those lower f-stops, and that looks like diffraction limitation. Since the D2X (and even the D200 with its 6 micron pixel pitch) seem to resolve to 60lp/mm and beyond, that supports the idea that maybe f/8 rather than f/11 might be needed to get optimal in-focus detail from the D2X. Which is what the sadly missed Didger said of his D2X.

[BJL]

I just discovered very interesting web page on diffraction limitation in relation to pixel size and f-stop: http://www.cambridgeincolour.com/tutorials...photography.htm
One quote: "the Canon EOS 20D [with 6.4 micron pixel spacing] ... begins to be diffraction limited at about f/11". And by their calculator and guidelines, the D2X does so at about f/8.

[bob mccarthy]

Question: Did slow fine grain film have a different diffraction limit than did medium speed/grain or higher speed/grain film?

bob

[BJL]

Question: Did slow fine grain film have a different diffraction limit than did medium speed/grain or higher speed/grain film?

bob
[a href=\"index.php?act=findpost&pid=62507\"][{POST_SNAPBACK}][/a]

First answer: yes, if you enlarged and viewed images to see the all the detail recorded on the film, so that the finer grained image was printed larger if viewing from the same distance (or equal sized prints were likely to be viewed from closer when scrutinizing the finest recorded details.) Because then the greater enlargement would make diffraction more noticeable at the same f-stop.

Second answer: no, if you compared prints from both films with equal print size and viewing conditions, like the "norm" of viewing distance equal to the diagonal length if the image.

(The diffraction limitation computer at the site I mention above include prints size and viewing distance in its determination of whether diffraction produces visible effects.)


In this thread, the OP was clearly seeking maximum image detail, which would including sufficient enlargement and sufficiently close viewing to be pushing the limits of sensor resolution, diffraction effects and out of focus effects. So my first answer seems more relevant.

To put it another way, about 5 to 8MP seems enough for normal viewing, so any significantly larger pixel count is only relevant for either for closer than normal viewing (prints bigger than viewing distance) or cropping. In either case, the same viewing distance requires a greater degree of enlargement than for "normal viewing", so increasing visibility of diffraction effects, and requiring a lower limit on f-stop to make diffraction invisible. The enlargement limit is set by sensor resolution for one thing, and it makes sense to say that diffraction is limiting resolution and sharpness when diffraction spots are significantly bigger than the photosites, and that the OOF effects are limiting resolution and sharpness where the circles are significantly bigger than the photosites.


This reminds me of the fact that standard DOF guidelines are not appropriate for very large, sharp, detailed prints intended to receive close scrutiny: those DOF tables and markings on lenses assume about 5"x7" prints viewed from 10".

[Dave Carter]

Gary or Bob, do you have a recommendation as to an excellent 150mm in 35mm format.

And yes I agree, I think you have a very good point and I am concerned about the amount of light that will be flying around off sides of my adapter.  I plan on using a black velvet on much of the inside of the adapter as well as baffels in several locations down the light path.

(Just as a side note:  I went to a local store and bought a 1' x 1 yard of 4 differnt black velvets. I then set up a table in my office with light shinning at the velvet sample from about 30 degree angle.  I then measured the light from the black velvet at about 30 degrees from the oposite direction.  I had a baffel vertically above the velvet so that my light meter did not see the light directly.
There was an amazing amount of difference in the amount of light reflected.  In fact, one of the velvets was worse then a flat black painted panel I included. Also, it made a difference as to direction of weave. )

Thanks again for your comments, Dave