PDA

View Full Version : Reduced DOF with Scanning Back--Lens Suggestions



prohtex
12-Feb-2011, 23:15
I recently moved from an 8x10 Cambo Monorail with two lenses (240mm Nikkor and 360mm Caltar) to a 4x5 with a 90mm F8 Fujinon, a 120mm F6.8 Schneider Angulon and a 210mm F5.6 Rodenstock Sironar N. I'm working with a BetterLight scanning back which picks up very minute changes in focus. My DOF is effectively reduced. When I was shooting film I always had great razor sharp images with the Nikkor, but am finding the wider lenses are not nearly as sharp, even in the center of the image.

For my current project I'm shooting a set that is 11' deep, from a distance of about 17'. I need for the entire set to be in very, very sharp focus. Can anybody recommend a lens? I can't get much farther away so will probably need something in the 180mm range. I have been testing these three 4x5 lenses without stellar results. Below you can see some example images which show the soft focus issue. (Use the controls to zoom all the way in and inspect focus)

Thanks in advance!

http://www.harkawik.com/betterlight/?x=BetterLight_0-crop-soft
http://www.harkawik.com/betterlight/?x=BetterLight_15-corr-soft
http://www.harkawik.com/betterlight/?x=BetterLight_3

Frank Petronio
12-Feb-2011, 23:46
You could try bracketing the focus -- one shot where you loupe it on the ground glass, the next two with a tweak forward and backward.

It sounds like you're using older lenses - the Angulon has to be at least 35 years old, the Sironar-N could be 25 easy. They maybe fine for general use but they may have a little haze and the newer coatings would probably lessen the "glow" you are getting. The older lenses like the Angulon are pretty soft wide open regardless, they need to be used at f/16 or smaller to look like a modern lens, otherwise they make "arty" images ;-)

Certainly the latest Rodenstock or Schneider lens will be the sharpest lenses available... $$$

Also, use a bellows lenshade and more careful protection from flare...

The other thing is the stability of your platform, are you using a really tight camera and sturdy tripod on a stable surface? You might be picking up some vibration.

And finally, this is where I don't know, but maybe the BetterLight is more UV sensitive to your light source? That's for the real BL users to answer but what light are you using?

Brian C. Miller
13-Feb-2011, 00:11
If the Betterlight back is sensitive to UV, then a cheap UV filter will cure that.

The third photo, with the lens, seems to be in focus. I wouldn't describe the other two as having "glow". Prohtex, I think you should create some kind of step wedge or something and find if the real focus of the Betterlight back matches your ground glass.

Ken Lee
13-Feb-2011, 03:06
According to the catalog Rodenstock Lenses for Digital Professional Photography (http://www.rodenstock-photo.com/en/main/products/lenses-for-digital-photography/)

"Digital photography makes... different and sometimes much higher demands on the lens because of the special technical requirements of the sensors (e.g. regular pixel grid, planar sensor surface and a 2 mm thick protective and filter glass plate in front of it). Only by meeting these requirements, the theoretically possible quality increase is to be realized in practice.

The resolving power and the contrast must be at their optimum even at large apertures (f/8 to f/11, when used with the relatively small area sensors even from f/5.6) to ensure that diffraction and color noise do not impair sharpness.

Contrary to roll and sheet films, the correction of curvature of field has to meet the highest demands because of the virtually perfectly planar sensor surface or scanned plane. Furthermore, the lenses must not generate any color fringes or any visible distortion."

To spend on a digital back and use ordinary analog lenses may be of questionable wisdom, no ?

Nathan Potter
13-Feb-2011, 10:27
prohtex, the depth of field of the lenses you are using is the same whether using film or a digital back. DOF is a characteristic of the lens aperture only. As Brian points out, you need to confirm that the sensor plane is exactly where the GG surface is - so a standard sloped ruler test is in order. That will also compensate for any delta due to refraction in the sensor surface protective glass which I believe the BL back does employ, albeit very thin.

In fact the superior planarity of the sensor makes it easier to capture the plane of focus compared to a typically slightly warped sheet of film in the standard film holder.

However the physics of image capture with a sensor is fundamentally different than with film since the sensor pixel is a distinct defined area for each RGB color (except in Foveon sensors) whereas with film RGB couplers are in layers behind each other. This among other factors will yield a quite different feel between the two image capture methods but I don't think that is the source of your DOF problem.

With apologies to Rodenstock assertion that "contrary to roll and sheet films" curvature of field ought to be eliminated at the film plane for both film and digital and is not limited to digital sensors alone.

In fact the last two quotes from Rodenstock that Ken has relayed hold for film as equally as for sensors.

Nate Potter, Austin TX.

erie patsellis
13-Feb-2011, 11:25
Betterlight backs are very, very sensitive to every part of the imaging chain, it's something you will have to come to terms with in your own way. I use a Dicomed FieldPro, the predecessor to the 6K, and you need lots of light. Also, use the focus tool in the Betterlight application, it will help you attain precise focus on the exact object you want in focus. The ground glass is a great tool for basic composition, but there will always be discrepencies due to tolerances and such.

Movements can be your friend, so it's time to cowboy up and start learning them.

prohtex
13-Feb-2011, 18:06
Hi everybody, thanks for your replies. I just don't think I'm going to be able to afford these "digital" lenses. I'm looking for something in the 150-180mm range. $500 ballpark. Suggestions? Is there really an appreciable difference in sharpness due to the coating of a Sironar-N vs Sironar-S?

prohtex
13-Feb-2011, 18:10
Also, as DoF capability increases with Focal Length decrease, will I get noticeably greater DoF with a 180mm lens vs. a 210, assuming all other factors remain the same?

prohtex
13-Feb-2011, 18:39
The other thing I was thinking was maybe I should use a 120mm lens that has coverage for 8x10 to get a 4x5 image comparable to using a 180mm lens, like a Nikkor SW 120mm?

Nathan Potter
13-Feb-2011, 19:28
Well I think you're more concerned with the Depth Of Field (DOFld) that you have to work with. That general relationship is DOFld = 2u^2NC/f^2. where:
u = subject to lens distance
N = f number of lens
C = diameter of an acceptable circle of confusion
f = lens focal length

This just derives from simple geometrical optics and to plug in numbers keep the units all the same - inches, centimeters, millimeters or microns, (f/no is dimensionless) in case you're not used to these simple equations.

So we can see that the DOFld increases by the square of the focal length so it really pays off to use a shorter focal length lens.

Whoa! The amount of light (intensity at the film plane, mW/cm^2, EV, etc.) depends only on the aperture setting. f/8 on a 120mm lens yields the same intensity as f/8 on a 210mm lens.

Notice from the equation above that the 120mm lens has almost 4 times the depth of field as the 210mm but with the added caveat that the image will be much smaller on your film. To get objects the same size you'll have to move in closer, which changes the value of u above. Notice that U goes linearly with DOFld whereas f goes inversely to the square of the focal length so there is a net gain to going to shorter focal lengths.

Also notice a secondary effect of change in perspective. Since your subjects have depth the spacial relationships of those in focus will be changed between longer and shorter focal length lenses.

Nate Potter, Austin TX.

prohtex
13-Feb-2011, 20:01
Hi Nate, thanks for your thoughtful explanation. Most of this stuff I know intuitively but it helps to have it explained like this. That is what I was wondering -- it seemed like a better idea to get as far back as possible, but then I would be using longer focal length lenses. So I will try to find a happy medium, maybe a 150. Still though, it seemed like there was some advantage to using 8x10 lenses, maybe because the image was all contained in the sweet spot of the lens where there is less barrel distortion and chromatic abberation etc...

prohtex
13-Feb-2011, 20:05
Also, in reading up on these "digital" lenses, it seems the primary advantage is the very short focal lengths, which allow you to do wide angle stuff with a CCD back, which has a tiny image plane. This is not the case with my scan back -- it has a 4x5" image area just like the ground glass. So I don't really see how one of these lenses would help my DoF issues or my sharpness issues...

erie patsellis
13-Feb-2011, 20:15
Actually, your scan back has a 7cm x 10 cm image area, significantly smaller than 4" (100mm) x 5" (125mm).

rdenney
13-Feb-2011, 20:15
For my current project I'm shooting a set that is 11' deep, from a distance of about 17'. I need for the entire set to be in very, very sharp focus. Can anybody recommend a lens?

Except for the Angulon, the lenses you have will exceed the capabilities of the BetterLight back. The 6000 model currently for sale has a sensel density of 83 spi. (This is the only back that provides that figure that I could find on their site. I do wish they would simply tell us the size of the sensor area rather than the unhelpful "how big a print you can make at 300 dpi"--sheesh!) At that density, any decent modern lens will give you what there is to get, if used properly.

But your above statement demonstrates what might be a misunderstanding. Depth of field does not make things in focus--it just makes things appear to be sharp according to a particular standard. The only thing that is in sharp focus is the focus plane itself.

So, "very, very sharp focus" is not helpful. A print size requirement would be much more helpful, because it would allow one to work backwards to the needed resolution. Now, we can find use for that statistic provided by BetterLight. If we have 8000 pixels horizontally, and printed at 300 (too low a number for optimal results from an Epson, but typical for a Canon or HP printer), we can make a print that is 26 inches wide. That is about a 7x enlargement from the BL6000 back. When viewed closely with the unaided eye, we might require a standard in the print of 5 lines/mm. With a 7x enlargement, we only need 35 lines/mm from the lens to achieve the full capability of a print that size. That's no problem with your Fujinon and Sironar lenses, and probably not an issue even for the Angulon if you stop it well down.

And that implies a circle of confusion standard of 0.015mm (figuring the c of c at the size of half that line pair, and about the size of one sensel on the scanning back--a very tough standard of sharpness in this situation). According to DOFMaster, a 90mm lens will provide sharpness to a circle-of-confusion standard of 0.015mm at f/32, from a distance of 13.1 feet to 24.3 feet, when focused at 17 feet. That's 11.3 feet within the depth of field, as you specified. Your specification cannot be attained with the 210, unless you can acceptably adjust the focus plane using tilts and swings. At f/32, there will be some effect of diffraction, but not enough to matter to the resolution of this back, I don't think.

Film can record more resolution than the scanning back, so if a lens can produce a satisfactorily sharp image on film but not on the scanning back, the lens is not at fault and the problem is either the limiting capability of the scanning back or somewhere else in the image-making chain.

Others have suggested checking the focus accuracy, squareness, and rigidity of the camera very carefully. That is good advice. The scanning back might not be sitting in the camera just the same way as your film holders, the camera or some of its movements may be moving when attaching the back, or the camera/tripod may be vibrating as a result of the scanning motion. All that must be carefully checked.

Rick "running the numbers" Denney

erie patsellis
13-Feb-2011, 20:21
Rick, the 6K uses a 6000 element tricolor array from Kodak, the image sensor's capture width is 7 cm (or 70 mm if you prefer) for a sensor pitch of .0116 (6 repeating) mm.

ic-racer
13-Feb-2011, 21:43
You have five LF lenses. You really need quite a few more than that to get any good images.



NOT!

rdenney
14-Feb-2011, 06:00
Rick, the 6K uses a 6000 element tricolor array from Kodak, the image sensor's capture width is 7 cm (or 70 mm if you prefer) for a sensor pitch of .0116 (6 repeating) mm.

Yeah, that's what I determined from what specs they did publish on their web site, though I rounded a bit.

Rick "6000x8000, 96mm wide" Denney

Ken Lee
14-Feb-2011, 06:30
The other thing I was thinking was maybe I should use a 120mm lens that has coverage for 8x10 to get a 4x5 image comparable to using a 180mm lens, like a Nikkor SW 120mm?

"Comparable" lenses for 8x10 are twice the size of those for 4x5. You need to stop down that longer lens by 2 f/stops to get equivalent depth of field.