Olympus Pen F: solving problems no one had?

The Olympus Pen F series of cameras is unique. And “unique” has many connotations in English, some not so complimentary. The Pen F, FT, and FV are variants of a half-frame SLR body that takes a variety of lenses. They deviate from conventional SLRs in ways that are apparent – and maybe not immediately so. These cameras are lovable but also irrational.

Size and weight: zero sum

It’s not really small. The first thing that people need to get out of their heads is the idea that a Pen F-series camera is smaller than a 24×36 camera. Not by a long shot. The Pen is far larger than a Contax T, a Minolta AF-C, or a Rollei 35S. In fact, it’s only 10mm narrower (side-to-side) and 4mm shorter (top-to-bottom) than a Canon VL2 rangefinder with a 35mm f/2 lens that shoots a frame double the size of the Pen’s “single frame” and has at least six times as many lenses that fit it. A Pen is, however, smaller than typical 35mm SLR.

Why is the size not smaller? What kind of gets lost in all of this is that the major dimensions of a camera, especially an SLR, are dictated by the frame size and whether it has a mirror. A “half frame” is half the area of 24×36, but its linear dimensions are only smaller by 30% (18×24). The picture below provides some scale; that width difference is indeed just 1cm.

Also not shocking is the fact that even with the frame scaled down, the 35mm cassette and takeup mechanism are the same size as a 24×36 camera. Smaller 35mm cameras manage to get the canister and the takeup immediately to either side of the film gate; the Olympus, weirdly, has this (because, oh, yeah, we forgot – a rotary shutter has to retract to somewhere).

No, the lenses aren’t really that small either. Compared to a 35mm SLR, you could make this argument. Compared to a 35mm Leica thread-mount camera, this is false. a 35mm f/2.5 Voigtlander pancake has the same diameter and half the height of the 38mm f/1.8 Pen Zuiko (in no small part because the VC lens sits partially inside the camera body). The 90mm Pen lens, likewise, is very similar in size to a 100/3.5 Canon LTM lens.

Pornoprisms and the rotary shutter club

Porroprism. The Pen uses a porroprism (no, not pornoprism) to deliver the viewfinder picture. Porroprisms are derived from binocular design and allow an optical path to be twisted on its axis to get around corners. You might know this same general idea as a periscope. The Pen uses this to get light from a side-flipping mirror, up the right side of the camera across the top deck, and into your eye. The byproduct is that the top of the camera is flat. Olympus penned this as “revolutionary” and derided “old style pentaprisms.”

The reality was that the porroprism was the only way to solve for a situation where the reflex mirror flipped to the side rather than flipping up. This, in turn, was the byproduct of a frame that was 24mm high and 18mm wide: if you reflected straight up into a pentaprism, the mirror box of the Pen would have been just as deep as a full-frame 35mm SLR – because 24mm worth of mirror height would have to flip up (because the image height is the same as a full-frame camera.

The porroprism system is not bright. Sorry. I have a bunch of Pens, and even the FV has nothing on a full-size 35mm SLR. Why? The porroprism system has many more parts and many more air-glass surfaces. For light to get into the eyepiece on a Pen, it passes through:

• A main reflex mirror (1 front-surface)
• A focusing screen (2 air to plastic surfaces)
• A lower prism (2 air to glass surfaces)
• Another front-surface mirror (on metered models, with a 50% light loss)
• A 2-component condenser lens (2 air to glass surfaces)
• A second prism (2 air to glass surfaces)
• An ocular lens (2 air to glass surfaces)

On a conventional SLR:

• A main reflex mirror (1 front-surface)
• A focusing screen (2 air to glass)
• A condenser (maybe)
• A pentaprism (2 air to glass surfaces)
• An ocular lens (2 air to glass surfaces)

On top of the insanely complicated porroprism system in the Pen that unavoidably degrades brightness and contrast, a most conventional SLRs do not compromise the entire viewfinder brightness for the sake of a meter. Some feed off the roof of the pentaprism and some look though semi silvered spots in reflex mirror. But it’s rare that there would be a first-surface mirror that had 50% reflectance/50% transmission.

The porroprism concept would not have worked on a full-size 35mm SLR, nor would it actually work on anything that did not have a “portrait” frame orientation relative to film travel. It might have done something to mitigate the awful size of 6×4.5 SLRs, but no one actually tried that. But at the end of the day, the porroprism was an artifact of an arbitrary choice of film format. It was the one way that Maitani could design a small half-frame SLR, and it is not portable to most other formats.

Rotary shutter. Olympus did not exactly invent this; the rotary shutter was long in use on cine cameras and even some full-sized 35mm, like the Unisex, er, I mean, Univex Mercury. This is a questionable innovation, and its use is really driven by the body-consuming porroprism path. Most 35mm SLRs are built around something like the Copal Square S shutter. The shutter is a flat unit; the mechanical drive is a box affixed to one side. None of this impedes the optical path, which goes straight up into the pentaprism. On the Pen, you can’t use that space because part of the viewfinder is there. The rotary shutter hardware in a Pen sits fairly flat. The advantage of a rotary shutter is that it can synch at all speeds; the disadvantage is that it can only spin so fast. Consider also the sensitivity of something whose timing relies on sheer speed. A conventional metal SLR shutter runs mechanically as fast as its synch speed. So it may sweep the frame at 1/125 of a second, but the effective speed is 1/1000 because only a fraction of the frame is being exposed at a time. This is one of the reasons why the Copal was one of the most popular focal-plane shutters of all time. It isn’t high strung.

The optics of optics

Lenses. One of the big draws for the Pen are its tiny lenses (compared to full-sized 35mm SLRs). That said, it’s a fairly closed universe. Lenses run from 20mm (~30mm in 35mm format), and conventional primes maxing out at 250mm (~375mm). Most primes take 43mm filters, which is convenient.

But there is nothing that wide-fiends would recognize as more than a moderately wide lens. The two zooms are constant aperture but do not really save much space.

In retrospect, Maitani’s decision to put the lens mount lock on the lens (and not the camera bayonet) was a poor choice. In theory, you should be able to fit any SLR lens to something with a register distance within .15mm of a Leica M. But the locking mechanism in the rear apparently has made this unattractive. So your choices are adapting M42 lenses (of which super-wide is rare), or Nikon or Olympus super-wides, which are generally so big that you may as well use them on full-size cameras.

Perhaps most infuriating is the inability to mount Leica lenses. The register distance of a Leica thread mount lens is 28.80mm; the register of a Pen is 28.95. It should have been possible, in theory, to put most (but not all) LTM lenses on the table. In theory, though – Leica-mount lenses still have rangefinder cams that project behind the lens mount, and some wides have protruding rear ends. So at best, maybe 50mm and up, plus retrofocus wides.

Those gripes aside, the Pen lenses are fantastically sharp. Having tested quite a few of these out, here are some short takes:

20mm f/3.5 – awesome wide-angle (like a 28mm or 30mm).

25mm f/4 – a sleeper of a wide, sold for very little, the equivalent of a 35mm lens. Fairly dim viewfinder image on the FT but passable on the F and FV. Unlike the 2.8 version, this is tiny, about the size of the 38mm.

38mm f/1.8 – the base lens on the FT and the normal bundled lens with an FV, this lens is small and sharp. It does not give up much in terms of speed to the 40mm f/1.4, and it is a touch shorter.

40mm f/1.4 – the “upgrade” on the FT, this is actually super-bright on the FV. It does not have great bokeh, and it does have lanthanum glass that will need bleaching every now and then.

100mm f/3.5 – surprisingly sharp and cheap 150mm equivalent. Not much smaller than a 35mm rangefinder 100mm lens.

50-90mm f/3.5 – a strange and wonderful fixed-aperture zoom, this actually has remarkably low distortion. The rotating front is not great for grad or other filters that themselves rotate, but this deserves far more play than people give it.

Replacing some old problems with new ones

The Forever War Roll. The Pen series actually exacerbates one problem: the roll of film that never ends. This was originally sold as a feature. In the long view, it is probably actually a bug. It is pretty much a given that you lose interest in a roll of film when it is about 80% done. Search your feelings. You know this to be true, no matter how long or short the roll. On a 24×36 camera, that might mean frame 19 on a roll of 24 or frame 29 on a roll of 36. Unfortunately, when your roll of film could be 72 frames, as on a Pen, that means you could be in the pain zone with fourteen frames to spare.

Scanning. Half-frame was never a favorite of photofinishers; even in the minilab era, 18×24 wreaked havoc on printing to normal sizes. For scanning, Pen negatives are manageable with a Pakon x35 series scanner (since you can tell it an arbitrary frame size and it does all 72 frames in one continuous process). They are not so manageable with higher-resolution film scanners because very little scanning software is designed around anything outside of “mainstream” formats (24×36, 6×4.5, 6×6, 6×7, 6×9).

Optical printing. One thing to bear in mind is that optically printing half-frame usually benefits from shorter enlarging lenses than 50mm; for 18×24, they are typically 40mm. You should also be aware that most brands of enlargers don’t make (and in many cases never made) negative carrier inserts in 18×24. That said, without the urge to pixel-peep, you might actually enjoy the prints.

We still love the Pen F

None of this is to say that the Pen F is a bad camera – in fact, it is an extremely cool platform that is fun to use and actually manages to make very sharp, well-exposed images. It is, however, a system whose marketing hype outruns the utility of its innovative features. In the end, if this piece sounds like razzing, just bear in mind that we only do it because we know the Pen can take it!

Taming the Durst AC800

Ok, you bought a Durst AC800 or AC800 Elite for black and white work – so now what?  The good news is that your unit’s memory is probably blank.  The AC800 series used a soldered-in Nicad to hold its memory settings.  More than 20 years after production, it’s unlikely that your unit will hold its settings unplugged for more than a day or two. But since it’s probably unlikely that you (or the previous owner) ever correctly understood the Allemainglish directions, you’re not losing much.

This is best I could come up with via trial and error.  Use calibration data at your own risk.

What the hell is an AC800?

An AC800 was pretty much the last thing before minilabs.  It is a 6x9cm enlarger with two color densitometers (one reading negatives and one positives) that was capable of self-calibrating and auto-correcting color prints from negatives and slides.  It cost about $15,000 new.  It was usually paired with an auto-advance roll-paper easel so you could really crank out the prints.  The AC800 Elite was the same thing, with autofocus.

Most AC800s that did not hit the scrap heap are being pressed into service printing black and white photos.  For this purpose they are overkill – but it is far easier to get an AC800 than an M805 or any other quality 6×9 enlarger.  Dursts have solid metal chasses and are pretty much immune to misalignment.  They make everything else look like a cheap, crudely-engineered toy.

The AC800 uses a different theory of metering, namely offsets in terms of density units (30 units = 1 stop).  In essence, you can translate real-world density adjustments directly from an external densitometer or step wedge directly into the machine.  Although this may seem strange at first to people used to using conventional timers with conventional enlargers, it is actually a lot more intuitive.  The AC800 is designed to get you a proof-quality print on the first try, no test strips.  From there, you fine-adjust the filtration or change the contrast to suit.  The AC800 needs to know what aperture it was calibrated at; from then on, if you tell it the aperture you are using, it will auto-compensate the exposure for that aperture.

Behind the scenes, the AC800 uses additive filters (R,G,B) to approximate subtractive (C,Y,M,K) filtration.  It uses its dichro filters both for color and neutral-density purposes.  For this reason, it is important to foget about the use of these filters to approximate multigrade filters.  In color mode, the machine will expose different colors for different times – something that just doesn’t work for multigrade paper.

Preliminaries

Don’t worry about expensive Durst COLAMP bulbs.  Get the cheap GE 24V/250W equivalents.  For b/w work, you don’t need color-calibrated bulbs.

The original carrier for this enlarger is COBINEG with a glassless BIMANEG carrier top and bottom (top has square sides, bottom has beveled, though you can use either in either position).  You can also use a BIMAGLA-AN on the top and/or a BIMAGLA on the bottom.  The AN glass of the BIMAGLA-AN will eliminate the need for most dust-spotting.

Since COBINEG is just a BIMANEG (M805 carrier) minus the masking blades, you can use a BIMANEG.  Just make sure that the masking blades are retracted when doing any type of calibration.  They can also skew the exposure if they impede too much into the frame.

The 6×9 mixing box is all you need.  You can get the 35mm or 6×6, and pay a lot for either, but your exposures will become painfully short. When used with 35mm film, the 6×9 box also helps diffuse things a little more, overcoming the inherent 35mm dust problem.

For lenses, you should use a 50, 80 and 100 with the AF version.  For the plain AC800, you can use 40s instead of 50s and 90s instead of 80s or 100s. I would recommend getting some hot glass, since this is a diffusion enlarger.  Lenses go on flat flanges for 100mm, semi-recessed for 80mm, and recessed for 50mm.  You can use the circular disk boards common to most Durst enlargers.

Understanding the memory of a machine

You must remember that this enlarger has four memories (well, you wouldn’t remmeber this because the instructions are terrible at explaining this).

CALIBRATION – this is long-term.  This is the baseline for each channel, and it stays in place until changed by the user or until the power is cut and the battery backup dies.

PRODUCTION – this is short term.  This treats the calibration value as the “zero” point and adds on.  This will re-zero between prints unless [HOLD] is lit.  This is the value that is shown on the 3-digit display when you are making prints.

BASE – this is medium term. If you hit [ENTER] + [SETUP] in production mode, the Production correction (on the numeric display) will be added to the calibration (temporarily), a dot will appear on the numeric display, and the numeric display will go to zero.  This is useful for changing to a paper that is, say, a stop slower.  Base memory gets erased if you change channels or turn off the enlarger.  This is useful if, say, your total density correction exceeds 99 units.  You could put up to 99 on the production, move that into base, and add another 99 to production.

HEAD – under the little sliding cover on the side is a fourth, hardware memory – this is the basic calibration to cancel the orange film base with color film.  Do not mess with this, ever.  Not only is it pointless to do so with b/w printing; you could also break the filter mechanism.  This is adjusted with the strange little ferrule tool that sits in a hole in the enlarger base.

Basic b/w calibration for multicontrast paper

First, get it out of your head that you are going to use the built-in dichro filters for multicontrast b/w work.  It just won’t work.  Those filters are designed to automatically deploy for color film, and the filtration values will not hold constant to recreate your favorite polycontrast settings.

Turn on the unit with a carrier in but no negative inserted.  You will see an R-G-B test sequence.  This allows the transmission densitometer to take a baseline reading.  At the same time, the reflection densitometer (on the side of the head) takes a baseline reading of an 18% grey reference (make sure you have a piece of grey card under there).

Switch the unit into b/w by pressing [ENTER]+[CHANNEL].  In this mode, the enlarger will use only yellow light to expose paper.

Here is the calibration method.  This will illustrate grade 0 paper assigned to channel 0.

1.  Make a 0,7 density negative (base plus exposure).  If you have the color calibration negs for this enlarger, they are not useful for this purpose.  Bracket a roll of film of a blank white wall, -2 to +2 exposure, in as small steps as you can.  Zero a densitometer (like a Getrag) on the film base, and then pick the negative that reads 0,7.  Note which negative it is.  The negative should be no smaller than 6×4.5.  The enlarger only reads the center 24x24mm, but you should have some margin.  Put this negative in the carrier.  You only need to make one negative for all channels.  You could conceivably also use a representative negative.

2. Cut a piece of the #0 filter to fit on top of the negative carrier.  Insert the negative carrier.  Set the lens to f/8 (“working aperture”)

3.  Press the [HOLD] button on the head.  This will assure that any corrections you will make will not disappear after your exposure.

4.  Expose a piece of paper.  Develop it.  Compare it to a grey card.

5.  If it is lighter than a grey card, press [DISP] until you see the D value and increase it.  30 units is one stop.  Likewise, if it is too dark, decrease the d value.

The unit actually has a closed-loop self-calibration (expose, show the print to the reflection densitometer, repeat until the D number is stable), but you will drive yourself up the wall trying to get it to work.  This method is simpler, is a one-time operation, and gets you results just as close.

6.  When you have the D number that gets you the right print density, record that number on a piece of paper.  Then press [+] and [-] simultaneously to zero the D setting.  This last part is important.

7.  Press setup. Using [+] and [-], input the lens value.  Then press [ENTER].  Then input your working aperture (see #2) and press [ENTER].  Then input the D value you recorded in #6.  Then press [ENTER] + [SETUP].  The enlarger will read the negative and lock that D value into the memory for that channel.  Press [SETUP] until the setup light goes off.  You don’t need to set XCP, LN and DN because b/w negatives and paper don’t have the same reciprocity issues that color does.

8.  For subsequent channels and grades (1-3), you want to achieve exposure times roughly equal to the exposure time for grade zero.  So on Channel 0 with the 0 filter in, press the [LIGHT] button on the enlarger.  Press [DISPLAY] until you see the Ey value.  This is the exposure time in seconds.  Mark this down.  Press [LIGHT] to turn off the lamp.

[Light] turns on the head with filtration activated.  [W. Light] provides no filtration, just what is set in the HEAD adjustment.  There is no practical difference.

9.  Next, put the #1 filter in.  Hit [LIGHT] and take a look at the Ey value.  Press display and move the D value up or down until Ey for #1 is the same as Ey for #0.  Mark this down.

10.  Next, do  the same for #2.

11.  Next, do the same for #3.

12.  For #4 and #5, try to get to values that is about double the Ey of grades 0-3.  This gets you into the right ballpark for the speed loss associated with grades 4-5 (usually half the paper speed).  Multicontrast filters affect film speed – it is not like using the 2-filter method with a dichro head.

Why are we trying to get to even exposure times?  Because you are using a color lightmeter to measure a monochrome negative plus a colored filter. The procedure above helps filter out the metering errors caused by using pink and yellow filters when you are using only the yellow channel to meter.

13.  Plug these values into the setup sequence for each filter (#7).  Make sure that the right filter for the channel is above the negative when you program these in.  And make sure the production D is 0 before initiating the setup sequence.  Otherwise, that value can screw up the calibration.

At this point, you should be able to run a series of test prints across grades that will be a consistent grey to within about 10 density units (1/3 stop).  You can agonize further and fine-tune the calibration numbers until they are all perfectly visually identical, but the reality is that changing paper grades will almost always necessitate changing the production D setting by about that much.  And the two emulsions of Ilford MGIV appear to be very slightly different colors, so there is a point of diminishing marginal return. 

14.  Once you get this whole process under control, go through each channel in [SETUP], hitting enter through the values until you can write down the D value for each channel.  If you do that and experience memory loss, all you need to do is repeat step 13.