IQ of digitar lenses on 6x7 film back versus film lenses

[paulr]

Weihan, it might be worth it to ask about all this on the getdpi.com forums. A lot of guys there have technical cameras and all the related lenses and toys.

[Weihan]

Weihan, it might be worth it to ask about all this on the getdpi.com forums. A lot of guys there have technical cameras and all the related lenses and toys.

I'll check out the forum there; thanks very much for the information! I'll have a browse through their offerings and see what is being discussed in the way of technical cameras etc. It's always good to hear about additional sources, and I wasn't aware of this site at all. Cheers!

[Weihan]

[QUOTE=Nathan Potter;992355]

Weihan, I recently started using the Schneider 47 XL with my TK45s'. Had to resort to a bag bellows of course but also a recessed lensboard (20mm recess I believe). Works fine but yes the focusing is tedious. For a number of scenes I simply set a hyperfocal distance based on a 20 µm COC at f/22. Typically yields everything from about 8 ft. to ∞ in focus or if you can stand a 40 µm COC at f/22 everything from 4 ft. to ∞ will be in focus.

If you really want to push the DOF you can choose a COC of 40 µm at f/45 and get everything from 2 ft. to ∞ in focus.

No matter what screen you use you'll be limited to critical focus within the central region of it because the peripheral rays are so far off axis that even a tilting loupe is of little value - at least that's what I find. As you say, it's a great special purpose lens for 4X5.

Nate Potter, Austin TX.

Those are very interesting figures, Nathan, and your values seem to reflect my own experience even though I haven't actually measured the COC. You've precisely understood the focusing issue in the corners of the GG, which is problematic even with a loupe like the Silvestri tilting one. There is very little tolerance for any movement with the 47mm XL, but it does deliver astonishing results. I do have the wide angle Maxwell screen, which improved things considerably, but the process is just as you've described - tedious. Do you have any images posted using the 47mm? Thanks for all your help!

[genotypewriter]

Has anyone ever compared the IQ of shots taken with a view camera + a 6x7 / 6x9 FILM back using a Schneider or Rodenstock digital lens versus a 4x5 film image shot with a traditional analog lens?

Haha wondered that a long time ago myself. Thank goodness I'm over it

Most of the Rodenstock ones have image circles of 120 to 150mm so they will cover most roll formats with no problem.
But be aware, The Rodenstock HR Digaron-S series and the former Rodenstock Apo Sironar Digital HR lenses require a correction plate mounted to the rear of the lens if these lenses are used with film. This is because the last element in the formula of these lenses is the cover glass over the sensor on digital backs. If the corrector plate is not used then the image will be slightly out of focus on film.

I think that's mostly true if you're using a sliding back. The flat piece of glass should only shift the focus... same happens when I take the clear-filter off lenses with drop-in rear filters. Same also happens when the IR/AA filter is replaced on digital cameras with specialised ones of different thicknesses. If focusing on GG as you'd do for film, I can't see how the lack of such plates can be an issue.

Based on experience with both 6x9 and 4x5, though, and assuming the same film for both 6x9 and 4x5 and appropriate working apertures in each case, I doubt very much that you're going to get "much higher resolution on film" with the 90 A-S-D on 6x9 compared to, say, a 135 Apo-Sironar-S on 4x5.

My thoughts exactly. There's always the issue of chunky grain with smaller formats... regardless of how infinite the optical resolution can be

[Bob Salomon]

"I think that's mostly true if you're using a sliding back. The flat piece of glass should only shift the focus... same happens when I take the clear-filter off lenses with drop-in rear filters. Same also happens when the IR/AA filter is replaced on digital cameras with specialised ones of different thicknesses. If focusing on GG as you'd do for film, I can't see how the lack of such plates can be an issue."

Then you are more then welcome to use the lenses without them. But don't complain when you find that the lens is not up to spec with film since you want to leave out the last element when used with film.

[Emmanuel BIGLER]

The purpose would be to shift the image back very slightly, most likely.

Yes, but not only. Of course the image as formed through a piece of flat glass is shifted by an amount of about 1/3 of the glass thickness (depending on the refractive index of the glass, the actual formula is : shift = t.(1-1/n) ; this yields t/3 when the refractive index n=1.5).
Hence as mentioned, the first advantage of the plane compensating glass is that if your lens is mounted on a calibrated helical with infinity stop, you do not have to change the mechanical setting of your infinity stop either with a silicon sensor or with film, since the total distance between the last lens vertex and the image, including the glass thickness, will be the same whether you use a digital sensor with cover glass (behind a "glassless" lens), or "bare" film (behind the glass-compensated lens).
The fact that this compensating glass can be located either near the image plane or just behind the last lens vertex is a non-trivial issue, and a purely paraxial analysis, found in all good textbooks of geometrical optics, yields a paraxial focus shift equal to t.(1-1/n) wherever the glass plate is located (this is an elementary analysis, valid only for a very narrow beam of rays at very small incidence angles).

This elementary paraxial analysis does not explain aberration issues at high angles of incidence, as briefly discussed below.
So, even if it is very convenient to be able to swap from a silicon, glass-covered sensor to a "bare" film detector without changing the infinity stop of the helical mount, this is not the real issue for high performance lenses delivering the maximum possible image quality. The problem is that introducing a flat glass inside a converging beam of rays introduces some amount of spherical aberration & other image defects. This has been known for microscope lenses for at least one century and so-called 'biological' microscope lenses are internally compensated to see through a 0.17 mm piece of covering glass of a standardized type (mostly, a given refractive index and a given 0.17 mm thickness). At high magnifications, light rays entering a microscope lens are very oblique, and without the internal compensation of the lens for a given, standardized covering glass, spherical aberrations generated by the covering glass would blur the images. This is not really visible with a low magnification microscope lens like a 5X, but with a 50x or more, this is really visible. Metallographic microscope lenses are not compensated for this and should directly look at the surface of the object without any glass cover.

So far, this effect of additional aberrations when a flat glass is introduced through a converging beam has been neglected in most photographical situations e.g. when introducing a filter behind the lens. People in the past used to care for the paraxial image shift ~ t/3, not for the degradation of image quality.
A well-known example is the [OFF-TOPIC] infamous plane glass introduced for the Rolleiflex TLR in the 1960s. This plane glass applied against the film itself was supposed to improve film flatness, but nobody cared for additional aberrations. Actually, Rolleiflex users cared for additional dust and parasitic reflexions, but this is another story.[/OFF-TOPIC]

[Brian K]

A few observations, and maybe some answers for some of the wide range of questions posed here.

First, MF will never equal LF, no matter the lens, except it will come closer at smaller reproduction sizes. But still compare an 8x10 contact print to an 8x10" print made from a MF camera and you will see a difference in the smoothness of the gradation.

The lenses on the Mamiya 7II are by far the sharpest lenses I have ever used. And I have done head to head resolution testing between them and my Rolleiflex glass, hasselblad glass, Fuji GX680 glass and my 3 dozen view camera lenses. The film being examined under a microscope. However I can well understand that people do not like the look of the Mamiya lenses as they can appear harsh and have poor OOF qualities.

I can't recall if the original poster talked about the final out put of the image, silver or inkjet. Inkjet prints can enlarge vastly higher, and keep what seems to be much better image quality at much greater magnifications. I have images shot MF that I would never consider enlarging on silver past 24-30" that can go up to 70" on pigment. So one solution for the original poster might be digital output.

As for roll film backs, I own a few Sinar Zoom and Zoom II backs. They are bulky but they do hold the film flat and have the advantage of not requiring that the GG be removed from the camera, not a minor point when working in the field. They also have the advantage of being able to switch format sizes on the fly. One has to make certain though that their camera back can hold both their thickness and their weight. I used to use them with a Canham 45 DLC and the weight of the RFH would tilt the back out of focus. Not an issue with a more rigid camera though. On my Sinars and Linhof Technika MT 3000 there are no issues.

I have found that the highest image quality with the smallest dimensions and weight combination is a Technika and a sinar zoom II. The technika is super rigid, folds down quite small, has no protruding rail, uses small lensboards, and the zoom back is not much bulkier than 2 4x5 holders. But note that I prefer to shoot 6x12 cm, if you prefer a more squarish format, 6x9 or 6x7 proportions then the zoom will have to be set accordingly and now you are using 3 1/2 inches of film width instead of 4 1/2". The IQ of this is quite high, I have 40" silver prints from this and from the look they could go 50 or 60". For a digital output these images could easily go 80-100"

[Nathan Potter]

Emmanuel, I think you have identified the concern I have about the use of a sensor cover glass with any lens not designed for cover glass use.

As you point out in microscopy objectives with higher magnifications and short working distances have far off axis rays and the path length through a cover glass is longer than rays on axis. Thus there will result a focus shift with a radial dependence (spherical abberation) in the image due to, I believe Snells Law, as you point out. This can only be avoided in camera objectives if the optics are infinity corrected at the image side. Do I read you correctly?

I apologize for what I suspect is a poor en francais translation.

Nate Potter, Austin TX.

Emmanuel, je pense que vous avez identifié ce qui m'inquiète à propos de l'utilisation d'un couvercle en verre du capteur avec n'importe quel objectif n'est pas conçu pour être utilisé couvercle en verre.

Comme vous le soulignez dans les objectifs de microscopie avec un grossissement plus important et courtes distances de travail ont éloigné les rayons axe et la longueur chemin à travers un couvercle en verre est plus longue que celle des rayons sur l'axe. Ainsi, il en résultera un décalage de focus avec une dépendance radiale (aberration sphérique) dans l'image raison, je crois que la loi Snells, comme vous le soulignez. Ceci ne peut être évité dans les objectifs de caméra si les optiques sont corrigés à l'infini sur le côté image. Dois-je vous avez bien lu?

Je apol; ogize pour ce que je soupçonne est une mauvaise traduction en francais.

Nate Potter, Austin TX.

[Bob Salomon]

"A well-known example is the [OFF-TOPIC] infamous plane glass introduced for the Rolleiflex TLR in the 1960s. This plane glass applied against the film itself was supposed to improve film flatness, but nobody cared for additional aberrations. Actually, Rolleiflex users cared for additional dust and parasitic reflexions, but this is another story."
A better example was the glass plate used in the Hasselblad Polaroid back to get the image plane closer to the roll film image plane.

[Brian C. Miller]

My thoughts exactly. There's always the issue of chunky grain with smaller formats... regardless of how infinite the optical resolution can be

Personally, I'd love to see the SPUR (Adox CMS II 20 or Rollei 25 tech pan) film shot on MF with a "digital" lens and then compared with LF. The 35mm shot they had didn't hold up, but it would be very interesting to shoot it in 645 and 67 formats and compare those with 4x5 and 8x10.

One of the main problems with small-format ultra-high resolution film is the enlargement. The smaller the film, the more difficult it is to preserve the details in an enlargement. Also, if the technical film is compared to a film with a different curve, that will also throw off the results.