Rolleiflex SL 2000 with digital back „SWVS“ (2017)
This took me a loooong while…
Apparently, the striking idea to be able to use several kinds of film simultaneously in one camera did not appeal to so many photographers after all. Or it proved to be too expensive, like in the case of the Rolleiflex.
I always had an ambivalent relationship with this camera. As convincing as the feature list reads, the camera is a most unwieldy brick.
is a milestone and one that points towards the digital age, even if
foresee that at the time. When they realized, the showed us a digital
1987, alas with only a tiny ½” sensor, with
laughable 0.3 Megapixels and with a
floppy disk drive for storage.
Still, 5 points to Gryffindor for trying…
Oh, if everything had turned out differently…
If Rollei had not become insolvent, or if its new owners had set different priorities, we might have seen an autofocus version of the 3003 with lenses from Carl Zeiss (and Zeiss might even have refrained from sailing away with Yashica, to new autofocus shores).
And eventually, we would have been given a digital back to fit the “3008 AF”, because this Rolleiflex is the single camera where such a back makes sense.
Even today, this would be an intriguing camera system, featuring the (relatively low cost) option to upgrade ones ageing camera kit to newer digital sensor technology. Sports shooters would own the newest body with lightning fast autofocus, landscape photographers would own the highest resolution sensor but maybe an old SL 2000 body – and people like me, who love to shoot an occasional slide film would just do this, with their never ageing Rolleiflex.
But then, being able to raise a hand and say, “Well, I’ve got one…” would still be nice, so I could not get the idea of such a digital back out of my head.
Even more so after Sony gave the photo world the NEX cameras, incorporating a relatively big APS-C sensor in a very small and rather cheap body. Should I tinker with that? Could I fit the digital innards of a NEX camera into the housing of a SL 2000 film magazine, thereby retaining the camera’s ability to also work with film?
Even from looking at the NEX camera it became clear that this was no mean feat, because the shutter sitting in front of the NEX sensor would have to go and make room, in order to move the sensor into the focal plane (where the space in front of it is occupied by the SL2000’s own shutter). That way, the digital sensor would capture the image - focusing and exposure would remain the task of the Rollei.
The first Sony NEX camera (a NEX 3 with 14 Megapixels) became my victim in 2014 and I enjoyed taking it apart. It is remarkable how few components the camera contains, how „integrated“ it is.
Bringing the sensor far enough forward to reach the focal plane proved to be the most difficult task of the whole undertaking – unfortunately an essential requirement for sharp pictures…
To be able to position the sensor, its frame needed to be machined down, from a little over 24mm to a little less – and the material is very hard, some kind of ceramic.
This also means that a “full frame” version of this back (using a 24x36mm sensor from e.g. a Sony A7) is out of bounds. Any sensor of this size would have at least a very narrow frame around and a protective glass in front of it, making it impossible to move the sensor far enough to the front.
There is not much room anyway, even though the actual sensor area is only half of 24x36. The flexible wirings are rather short, so there is not much space to rearrange the components.
I did a lot of filing and then I damaged the sensor and I threw it all into a box and out of sight.
Then I tried again with another camera with the same result…
Meanwhile another tinkerer had successfully tried something similar, using an old Konica rangefinder (https://frankencamera.wordpress.com/). Unfortunately, his kickstarter follow-up project to convert a Leica M3 went not so well, at least not for those who invested money in him.
This guy here (https://digitalfilmcams.wordpress.com/) managed to pack a Nex 3 camera onto a Nikon SLR. It is fun to read how he ran into similar problems like myself.
Well, I stopped short of wrecking yet another Sony camera body and turned to a Samsung NX 1000 instead, similarly cheap and with impressive 20 Megapixels, which turned out to be easier to work on and which also suits the Rollei well, given Samsung owned them once. The Samsung sensor has the same (APS-C) size and its frame also needed machining, the material however is much softer and therefore fits my abilities better. Note that I did the grinding with a machine, for the first time…
Below pictures show the sensor already mounted in the housing. Its mounting plate (supposed to press the sensor against the film rails) is made of ordinary cardboard, a compromise between stiffness and elasticity. Later, the actual position of the sensor had to be corrected quite a bit (using stripes of film and tape) – it seems the absence of the AA filter caused a significant change of the focal plane, compared to what my measurements were in the beginning (or I simply miscalculated…).
The initial idea was to position the Sony display under the camera (the NEX display is tiltable). Since the Samsung display is not and the flexboard is short (and since the camera does not work if no display is connected), I had to fit the display inside the housing as well. This looks a little bit weird but at least allows to change settings when the Rollei battery back is removed.
Yet one more dial had to be fitted into the housing – ISO, needed by the camera to calculate the correct shutter speed. The original magazine bears a stepped resistor and a dial on the left side, which I removed and replaced by a variable resistor, at first.
Eventually, I went all the way and transplanted the original parts to the other side of the housing, replacing the knob that once moved the film holder. The roll film symbol now serves as the ISO index.
Functionally, this was the plan:
- Switch on Rolleiflex and digital back
- Focus and select aperture on the Rolleiflex
- Release the shutter on the back („B“ or e.g. 2 seconds)
- Compose the frame and fire the Rollei shutter.
All Rolleiflex shutter speeds can be used, as long as these are longer than the speeds set on the digital back. During the Samsung's exposure, photons are collected by the sensor - it just happens that those only come through when the Rolleiflex shutter is open. Just which part regulates the light quantity - the Rollei shutter or the idle Samsung shutter in the magazine housing - does not seem to bother the Samsung image processing at all.
If exposure correction is needed, this can be done on the Rollei or (faster) using the exposure lock. If the ISO setting is changed on the digital back, the corresponding value must be set on the magazine housing.
It would have been nice to be able to fire both „cameras“ at the same time. This may indeed be possible, both cameras offer electrical releases – I may look into that another time.
Such things would be no-brainers for a camera manufacturer – but they are complicated for me. Eventually, my only chance is to make both cameras believe everything is well under control and there is no reason to show an „error“. I have no way of influencing how the Samsung electronics tick.
Since the APS-C sensor is smaller than 24x36, the finder(s) show more than what is on the picture, the actual frame is considerably larger than the grey “meter” field on the focusing screen.
Here are APS-C sized markings I have put on the focusing screen (not quite perfect, I admit).
So no big deal, in theory. In practical terms, there were a lot of components that waited to be fitted into that small housing.
And then, very exciting, the first test image. This could well be the first digital picture taken with this camera type, this side of the 1987 Sony Mavica back.
More body panels waiting to be fitted:
I have made a cover to protect the sensor when not in use:
The blue „Anti Alias“ filter sitting in front of the sensor had to be removed, obviously, it would have collided with the Rollei shutter. With that, the dust removal function of the Samsung went as well (a tiny ultrasonic actuator makes the glass vibrate).
Since the sensor is pretty exposed now, any dust can be wiped off easily (don’t worry, there still is some glass in front of the photo sites). And dust indeed has to be wiped off often... WIth the filter gone, dust particles now gather much closer to the focal plane and the photo sites, and are visible tack sharp on the pictures.
If photographing in RAW format, the color shift caused by the missing blue glass can be entirely corrected. While the Samsung also allows for correction of jpeg colors in camera, its range is not sufficient to balance the missing filter.
However, the blue glass is not only an AA filter but also one that cuts out infrared light.
This test image already gives a clue of the things that can happen. Where the white balance is acceptable on the whole, the purple cast on the left comes from incandescent light in the adjoining room.
Outdoors, things become very strange…
Light reflected from green plants contains a large share of infrared frequencies, invisible to us mere humans but not so for the digital sensor. If the infrared cut filter is missing (as in this case), the color reproduction gets completely off-balance, like here:
Luckily, there are separate infrared cut filters available (B&W 486) – though not for free - which cure the problem and make the green appear again.
In principle, there should also be occasional moirée on fine structures – we will see.
This is the completed back:
Weight is 275 grams, 50 grams less than the film magazine, even if it includes an obsolete focal plane shutter, a battery and a display. Looking at it, the film magazine would not have to be that heavy, it is - in small scale - a symbol of the demise of Rollei.
The magazine includes such a stunning number of individual parts, and it is so obvious that cost was no issue for the design, no wonder if was so expensive, it cost more than a complete body or two from other makes (and the image only shows the parts I have not used further!).
To compete with Samsung or Sony, Rollei would have had to learn more than autofocus and sensor technology…
As a happy ending, here are some Sunday-morning-in-Cologne pictures:
Here are some pictures taken during a nightime visit to the "Landschaftspark Duisburg Nord", a major industrial heritage site.
One open questions remains: What does „SWVS“ stand for?
"System With Variable Sensors“?
Fabricating a common shutter release for the Rolleiflex camera and the Samsung back proved to be more complicated than I had thought. Both components can be triggered electronically, the Rolleiflex via its remote control socket on the left side, the Samsung through its USB port.
USB remote controls for the Samsung can be had for a few Euros, the plug however sticks out quite a bit, which is both ugly and impractical. While trying to shorten the plug, so that the wires would run inside the housing, I found that the plug contains a small printed circuit with SMD components (the remote control handle contains nothing but a switch).
I needed several attempts until I finally managed to solder wires to the plug itself, to lead to the small print board. Not something I would want to do all day…
Triggering the Rolleiflex shutter was easier. Once I understood which contact switches the meter and which the shutter, the camera happily rattled on when the cables were connected. It still needs a two stage shutter release switch though. Meter and shutter can be triggered at the same time, this however means that the camera fires immediately if one of its own shutter buttons is pressed only half way…
And well, the shutter switch would also have to be bipolar because both 'cameras' work off their own energy supply (and with different voltage).
A two stage bipolar switch however does not exist, at least not a sufficiently small one…
After some failures, I managed to fabricate something from two separate bipolar switches mounted in sequence. The first one (pushed from below) puts on the Rolleiflex meter, and - when pressed further – actuates the second switch, which fires both shutters. Luckily, the Samsung shutter delay is less than that of the Rolleiflex (inside which a mirror box takes its time to operate), so that both shutters can be released at the same time. When the Roleiflex has finished raising the mirror and opening the shutter, the Samsung sensor already waits for photons.
The rest was no big deal. The cryptic SWVS lettering had to go, but at least the whole thing is not much bigger now than it was – and a lot easier to operate.