The LL Digital Field Camera Experiment has Ended...

[Eric Leppanen]

It appears that Michael Reichmann's experiment with his digital field camera setup (see http://www.luminous-landscape.com/reviews/cameras/digital-view.shtml and http://www.luminous-landscape.com/essays/counting-ants.shtml) has ended, and his kit is now for sale:

http://luminous-landscape.com/forum/index.php?showtopic=18617

Note Gary Ferguson's comments (which Michael says he agrees with) regarding the real world limitations of such a setup:

Like 4x5 it's slow and cumbersome, but unlike 4x5 you're not rewarded with a unique viewfinder experience or gold standard quality. You're squinting at a 37mm x 49mm image, trying to make precise camera adjustments that will be later scrutinised at much higher enlargements...And rather than slipping a cheap film holder into the camera, you're gingerly attaching, and then detaching, a very expensive and delicate digital back before each exposure. So check your insurance, scrutinise the weather forecast, and brush up on your sensor cleaning techniques!

So it appears that practical digital view camera options for field use are limited to architectural cameras (such as the Cambo Wide) which support rise/fall but not swing/tilt; stitching several "lower" resolution MF digital shots together using a standard 4x5 camera (as suggested by Richard Sexton at http://www.luminous-landscape.com/reviews/cameras/digital-ebony.shtml), which sounds good for architecture but problematic for shooting moving objects or dealing with rapidly changing light; using tilt/shift lenses on SLR's, which have limited coverage and focal lengths; or using a scanning back (slow, bulky and again difficult to use with moving subjects or changing/low light).

I doubt any of this will surprise those who have been following digital technology, but it does mean that (excluding for the moment the argument about digital versus film image quality) digital has yet to introduce a practical, functionally equivalent replacement product for film-based field cameras.

I take no pleasure one way or the other in any of this; I am pretty agnostic regarding the film versus digital debate. But it does appear that some of us will be holding onto our film cameras longer than originally anticipated (possibly much longer). I just don't see myself stitching together an entire matrix of digital SLR captures to compensate for a lack of camera movements: compensating for wind motion effects, lens distortion or falloff, light changes, depth-of-field limitations, etc. would take most of the fun out of photography for me. After working in the computer industry for 20 years, I don't want to spend still more time sitting in front of a computer.

So let's keep working on buying more film and keeping at least a few film processing labs healthy, right?

[Sheldon N]

I had figured that era had passed as soon as I saw that he was using the H2 system instead.

I saw the camera in use on one of his DVD segments, and it was even more complex than using a 4x5 view camera. That Linhof weighs 9lbs and is quite cumbersome, so you might as well shoot 4x5 instead.

I would hate to try and achieve absolutely perfect movements/focus in the field while composing on a tiny little section of ground glass.

[Doug Dolde]

This is his usual pattern...gushing over a new equipment acquisition, detailing the myriad reasons for acquiring it, calling it "simply the finest blah blah blah I've ever used"....then 6 months later dumping his latest love for a new girl.

Whatcha wanna bet he dumps the H2 for the new Hy6?

[Sheldon N]

This is his usual pattern...gushing over a new equipment acquisition, detailing the myriad reasons for acquiring it, calling it "simply the finest blah blah blah I've ever used"....then 6 months later dumping his latest love for a new girl.

Whatcha wanna bet he dumps the H2 for the new Hy6?

No need to bet, it's a sure thing. If it's not the Hy6, it'll be something else at the bleeding edge of technology and stratospheric pricing.

He uses new toys and reviews them, just to fund his continued acquisition of other new toys.

I know I'd do the same thing if I could pull it off. :)

[paul r w freeman]

Whoops, I think that they've thrown out the baby with the bathwater!

Michaels test was only one way of skinning this particular cat.

Its taken me a year, but I now have a useable VC solution for the field. You just have to understand the limitations and advantages of each tool for the job.

For wide angle and architectural shooting I use the Cambo wide. For personal work I am using the Ebony 45S + Cambo stitching back with the Aptus 75.

Its taken me quite a long time to understand how to use this gear in a way which benefits me. As with all things digital, the whole workflow has to be considered.

The final piece of the puzzle fell into place just a few weeks ago with the release of Photshop CS3, specifically its new improved Photomerge facility which can now take three raw files and stitch them automatically into a 77 megapixel image.

For moving subjects, obviously this is not a good solution. But it works fine for landscapes and my selected subjects. Its a new kind of photography, more akin to painting. If you shoot say, 10 frames and select features from 7 to be composed into the final image its a different way of representing reality, but no less valid than a single slice in time. Actually it answers David Hockney's objection about photography that it is limited in the amount of information that can make up a picture.

CS3 makes this work feasible at commercial speed due to its ability to accurately 'auto-align' frames.

The Ebony 45S has certain limitations. I made my own laminated adapter for the Cambo sliding back for the Ebony. This, if refined could be maybe 2mm thinner. However with this setup I can comfortably get a lot of movements with the Schenider digitar 72XL in a recessed panel. It could probably be made to work with the Rodenstock 55 apo sironar digital but with less movements.

Stitching makes up for the relative lack of wide angles. Using the Cambo back I get roughly the following equivalences using the Aptus.

72XL equivalent to 50mm lens
90 equivalent to 65
105 equivalent to 72
150 equivalent to 90

Thats in MF digital terms. Roughly this means my widest 'virtual lens' in 5x4 format is about 90-110 mm. Which is pretty fine for most photography.

One advantage of the Cambo stitching back is that effectively you get a large viewfinder. It has a built in loupe. This is particularly nice to use with the 105 f2.8 Xenotar but is still very usable at f5.6. Lenses with max openings of f8/9 are harder work.

People can argue about the quality of a 33 megapixel digital file. Practically I find it competes very well with 5x4, though probably not with 5x4 at the level that people on this forum can achieve. With a stitched 77 megapixel file then it easily supercedes what I can achieve with 5x4.

The exciting discovery for me was that with stitching, I can now take an image with digital with the 'LF look' rather than an MF look. Yes, this kind of photography feels more like 'LF photography' ie: slower, more cumbersome etc. But then the results I think will be worth it.

This setup easily fits into a small rucksack, taking up less room than if I had to include film holders and all the other paraphenalia. Two large batteries allow shooting all day.

Into the bargain, throw in some 5x4 film holders, some type 55 and a polaroid back and you can shoot 5x4 at will as well.

Lets see if I'm still using this in a year. If I am, then the debate will be over for me. I have been 'through hell' to get to this point, but now its all working. The biggest problem was realising that the MF back did not have correct chip alignment.

In the end, I think Reichman was barking mad to try to use a monorail as a piece of field equipment, but then there are probably people on this forum who regularly use giant cumbersome cameras and get fantastic results... so maybe it just takes a tougher and more determined photographer than he was to get digital LF to work.

[Marko]

Paul,

Aside from Photomerge, the new layer align and blend fetures in Photoshop CS3 bring about another, less obvious but still equally powerful possibility. Shooting a number of frames of a same scene using a stationary camera, effectively a mini time-lapse sequence, allows for relatively easy combining of those frames in such a way as to elliminate any moving object from the image altogether. The most obvious way of using this technique would be for taking an image of a busy street with people and cars always present but rarely stationary.

[PViapiano]

Paul,

I'd love to see some examples of your work...

[gregstidham]

Blog entry #234. www.iwannabeadigishooter.com

"The weather turns and a light rain starts. Clouds form and create a wonderful palette of greys and blues providing a perfect backdrop for the building I am photographing. Too bad I had to pack up my camera so I didn't ruin my $30,000 digi back. It would have made a nice image. Clear sky tomorrow I think - I hope."

I joke, test was interesting reading anyhow. :)

[Eric Leppanen]

Paul,

If I understand your post correctly, you are using a solution similar to Richard Sexton's set-up (documented in the LL article in my post above), only instead of shooting a 2-3 frame panoramic (as I understand Richard does), you shoot a matrix of as many as 10 overlapping shots (maybe more?), is that correct?

I'm very interested in how well your approach works for you during the coming months. Unlike Richard (who did not need more than a 16Mp back for his work), you are using a top-of-the-line MFDB. So theoretically one would think you could achieve amazing resolution assuming you have the need/value for doing so, and are not lens limited. Do you eventually plan to purchase some (more?) digital lenses to improve image resolution if you end up happy with your solution? Also, do you foresee an eventual need for longer lenses (Richard was quite happy using a 300mm lens via an extension tube) for your shooting style?

I'm not a big fan of spending a lot of time sitting at the computer stitching images, but if Photomerge takes most of the pain out of this task then I might reconsider.

Thanks!

[paul r w freeman]

Paul,

Aside from Photomerge, the new layer align and blend fetures in Photoshop CS3 bring about another, less obvious but still equally powerful possibility. Shooting a number of frames of a same scene using a stationary camera, effectively a mini time-lapse sequence, allows for relatively easy combining of those frames in such a way as to elliminate any moving object from the image altogether. The most obvious way of using this technique would be for taking an image of a busy street with people and cars always present but rarely stationary.

Yes, I tried this and its really fantastic if you want to compose a scene from multiple elements. I actually combined two architectural shots made at slightly different perspectives and it transformed one layer so it matched the other. Incredibly useful in a number of ways.

[paul r w freeman]

Paul,

I'd love to see some examples of your work...

Glad to oblige

my personal work, type 55 based is on the Ebony gallery

http://www.ebonycamera.com/gal.contrib/freemanp/index.html

my commercial work is on my website, mostly shot on the Ebony (film) or on the Cambo Wide camera (Aptus 75)

http://www.paulfreeman.net

I haven't got anything yet I'm greatly proud of from the Ebony + stitching back, well there is one shot so far. I'll post some more where I have them

[paul r w freeman]

Paul,

If I understand your post correctly, you are using a solution similar to Richard Sexton's set-up (documented in the LL article in my post above), only instead of shooting a 2-3 frame panoramic (as I understand Richard does), you shoot a matrix of as many as 10 overlapping shots (maybe more?), is that correct?


No, I am pretty much doing the same thing as he has. Thanks for the heads up, I hadn't seen his article but I should give it some more attention. I long ago worked out that lenses of a short FL like the 35's were pretty impractical on a view camera. I could only forsee these being easy to focus with 'live video' view in tethered mode. The problem is that the tiniest misalignment in the front or rear standard would result in part of the image going out of focus. This is very challenging even for a geared camera like the one used by MR. Pretty impossible with an Ebony.



I'm very interested in how well your approach works for you during the coming months. Unlike Richard (who did not need more than a 16Mp back for his work), you are using a top-of-the-line MFDB. So theoretically one would think you could achieve amazing resolution assuming you have the need/value for doing so, and are not lens limited. Do you eventually plan to purchase some (more?) digital lenses to improve image resolution if you end up happy with your solution? Also, do you foresee an eventual need for longer lenses (Richard was quite happy using a 300mm lens via an extension tube) for your shooting style?


I'll keep you posted. When I bought the Aptus, it was specifically for my commercial architectural and interiors work. I couldn't bring myself to sell the Ebony and lenses, so I've been searching for a way of using it. I've always admired the 8x10 photography of people like Joel Meyerowitz, Stephen Shore, Simon Norfolk et al, and now I see this technique as a way of reaching something approaching the techical quality digitally. I'm guessing that I'm achieving roughly the resolution of 5x7, though I have an advantage in the small size of the platform and its being easier to use in windy conditions than an 8x10 camera, lower vibration etc.

I still have a G-Claron 240 which I still have to test. My first tests suggested it was a wee bit soft, but this may have been due to the focusing fault with my first Aptus back. If its still too soft I'll be looking for another 240 or possibly a 300. My subject matter would typically need a wide angle or normal field of view.



I'm not a big fan of spending a lot of time sitting at the computer stitching images, but if Photomerge takes most of the pain out of this task then I might reconsider.


In the case of the sliding back I was having to manually work on the images for about half an hour or more because as the back slides from left to right there is a little vertical movement, as a result the successive images are rotated by one or two pixels. This is probably just due to the tolerance in the sliding mechanism and the weight of the Aptus back. Really, without CS3, this would not be a practical approach for me at all.