The SEI photometer

Being the first part of a series of entirely nonfictional articles on light meters which will not fit on hotshoes.

The SEI Photometer is an astonishing device. Produced in Salford from 1948 until the 1960s, it looks like something that should be used while wearing a lab coat, a spacesuit or perhaps in a submarine. And so, of course, it was. It was what would today be called ‘dual use’: the Geiger-counter attachment (now very rare) was available for purchase by members of the public, but certain other attachments were not. It is best, even today, not to discuss these attachments in a forum not all of whose readers may have the appropriate clearance and I will therefore not mention them further here, but describe only the photographic applications of the device. I cannot comment on whether I am in possession of any of these attachments.

It is worth mentioning, however, that should you acquire one of these you should certainly check it for radiological contamination, and be aware of possible chemical traces. Fortunately mine appears quite safe and the traces of what appeared to be blood around the eyepiece washed off quite easily.

The SEI meter was famously used by cinematographers: they were much sought-after long after production ceased and may still be so. The rumours that one is still sitting in the wreckage of a crashed spacecraft on the Moon are unconfirmed.

Solving an interesting problem

Photographically the SEI Photometer is a narrow-angle spotmeter. Designing such a thing presents an interesting problem: the amount of light that reaches the meter goes as the square of the angle, so a narrow-angle meter sees almost no light. Unless the light-detecting element is very sensitive, which it was not in 1948, this is a serious problem. The simplest solution is to give the meter a very large light-gathering area. A refracting design rapidly becomes hard to lift. A reflecting design is both shorter and considerably lighter, but the infamous Roberts meter of 1946, with its 12 inch primary mirror was not exactly portable, even by the standards of large-format photographers. The mirror also needed frequent resilvering and was physically fragile.

The SEI in contrast is relatively tiny. Larger and heavier than a small 35mm camera of course, and needing two hands to operate, but at least relatively practical to use. And it’s a half-degree meter: the Pentax spotmeter (see a later article) is one degree, and so sees four times as much light. It’s a lot later than the SEI, and it’s informative to compare the sizes of the objectives.

How does the SEI work? It does a brilliant trick: it uses the human eye. The eye is completely terrible at measuring absolute levels of brightness, but it’s really good at detecting step changes in brightness between two adjacent areas of an image. So the trick the meter does is to light up a tiny spot in the image to a controlled, adjustable, brightness. You then place this spot next to, or over, the thing whose brightness you want to assess, and adjust the spot’s brightness until it vanishes, which it does in a most magical way. This trick was discovered by Bunsen – he of the burner – in the 19th century, and it’s called Bunsen’s grease-spot photometer (also the name of an avant-garde rock band, of course). His trick was to make a grease-spot on a bit of paper, with a light shining through it. When the light illuminating the paper was the same brightness as the light shining through it, the grease-spot would become invisible.

The trick only solves part of the problem. The other part is to know how brightly the spot is being illuminated: unless you know that the whole project is doomed. The SEI solves this by using a bulb whose brightness can be controlled by a rheostat on the base of the meter. This bulb shines directly on a photocell, the current from which drives a galvanometer on the top of the device. You adjust the rheostat until the galvanometer reads a standard value, at which point the bulb has a standard brightness. The calibration depends on the photocell, but it has the bulb shining directly on it: it doesn’t need to be very sensitive. The calibration doesn’t depend on the photocell being linear. To illuminate the spot the bulb shines through an adjustable ND filter: two wedges which slide over each other as you rotate the bottom of the meter. That’s purely optomechanical: there are no potentiometers to get noisy. It’s just a lovely design.

And you can tell this as soon as you handle the device (do remember to check for contamination): it’s made like certain German cameras, except rather more nicely: a column of machined aluminium held together with machine screws with some antique thread standard. And it’s only incidentally a tool for photographers: it feels like what it is: a scientific instrument built for scientists. It is unsurprising that one Bernard Quatermass worked for SEI for a time. What he did there is still classified, but I like to think he worked on the Photometer.

Things that can go wrong

The meter uses a D cell, which can vomit battery acid in the usual way. I’d avoid one which is too horrible inside. The cell also goes into the meter ‘upside down’: with the positive end towards the base of the meter. Nothing electronic cares about this as the battery just lights the bulb, but it won’t make contact if it’s the wrong way around. The battery in mine read 0.6V open-circuit and was in backwards. But it hadn’t leaked at all. With a new battery and some minor contact cleaning it sprung to life immediately. The manual (certainly written by people wearing white coats in offices aromatic with pipe smoke: it cites academic papers) advises you to take the battery out when the meter is not in use and that’s good advice.

The rheostat that adjusts the bulb brightness can get cruddy. But it’s wire-wound and frobbing it a bit scrubs off any nastiness pretty effectively.

The bulb can die. But it’s run way below its rated voltage so it usually doesn’t. If it does it’s basically just a 1.5V incandescent bulb. The only interesting thing about it is that it needs to be frosted glass. That’s probably something you can arrange by painting some suitable translucent coating on an ordinary bulb. More likely is that the bulb has just got very old and dim, and the rheostat will no longer make it bright enough. It needs replacement at that point.

The lurking nightmare is the photocell used to calibrate the bulb: the whole meter depends on this. It is in theory replaceable. But it is right at the top of the battery/bulb compartment and I’ve not even tried getting access to it. Fortunately what kills selenium cells is light, and particularly short-wavelength light: the one in the SEI only sees light when the bulb is on, and even then it has little short-wavelength content. It’s probably as well to keep the base of the meter on, and definitely do not arrange for sunlight to shine on the cell. If replacing it the meter would need to be recalibrated. Again, mine seems fine: with a new battery the first thing that happened was the galvanometer hit the end-stop with an audible click.

A problem which I have not completely solved is that the photocell is in the rotating part pf the column while the galvanometer is not. So the output of the photocell has to go through a sliding contact which inevitably gets dirty. Worse, it’s not accessible without more disassembly of the thing than I was willing to attempt. I just live with having to repeatedly adjust the column until the reading is stable and then adjusting the rheostat to calibrate it. You don’t have to do this every time you use the meter of course.

I’m sure the galvanometer could die: mine hasn’t.

The optics in mine are dirty, but usable. I think cleaning the internal optical surfaces would be hard.

Using the meter

To use the meter you first calibrate the bulb by pressing the button and then adjusting the rheostat on the base until the galvanometer sits on the red line. On mine this is fiddly because of the contact problems.

You then look through the little telescope and match the spot against what you want to meter. The telescope shows things upside-down of course: the kind of photographer who might use the meter will be used to that. You can adjust the dioptre by moving the back of the telescope in and out, and also move the front of it out for very close-up work. There are two little click wheels: one lets you set one of three brightness scales by putting filters in the optics, the other makes the spot the right colour for incandescent light. The telescope is not really usable with glasses, but there’s enough dioptre adjustment for me that this was not a problem.

Once you’ve picked the right settings, you look through it, press the button and turn the base until the spot vanishes. This is rather like using a good SLR focussing screen: when it happens it really happens, it’s quite dramatic.

You need to set the film speed by a little click-stop thing. There were two versions of the meter: the older one was in BSI, the newer in ASA. Mine is BSI. BSI is DIN + 10, so ISO 100 film is BSI 31. The sort of person who used the SEI, and certainly the sort of person who wrote its manual, would certainly have carried a slide rule at all times to do such conversions where they had not memorised them. I wrote some functions in Common Lisp, which perhaps is the latter-day equivalent.

And now you discover just how strange a device it is, and how little you understand about metering.

An approach to metering

So you set the right film speed, pick an aperture and read the exposure from the right scale. And it’s insane. I used two meters which I know agree well enough with each other and with a modern digital camera to meter a spot on a white wall lit by an LED light. The two good meters read 1/30s at f/8: the SEI read 1/320s at f/8: more than three stops out. Well, I thought: this thing is at least 60 years old and perhaps more than 70: probably it should just join my collection of beautiful cold-war relics, most of which are thankfully non-functional.

And then I looked at the film speed markings on the meter: there are two, one for negative film and one for reversal film, and they’re about 6.5 stops apart. And that makes no sense, either. So, in desperation I read the manual again, more carefully. And finally, I understood it, and it’s brilliant.

We all know the rule of thumb for negative film: expose for the shadows and the highlights will usually be OK. Fewer perhaps know that the rule for reversal film is the opposite, and so the same as for digital: expose for the highlights and the shadows will usually be OK. This is easy to understand in terms of density of the image on the film: it can keep on getting denser for quite a long time, but once you’re just looking at film base that’s it.

With an averaging meter, or a spot meter with a fairly large field of view you can’t really exploit this: your best bet for negative film is to meter an average grey and then decide, based on what the scene looks like, whether you need to overexpose to keep shadow detail.

With a really narrow-angle spot meter you can exploit it, and this is what SEI worked out. So, for negative film you are meant to use the negative film mark and then meter the point where you want shadow detail to just be appearing. For reversal film you use the reversal mark and meter the point where you want highlight detail to start getting lost. And you can do this because the acceptance angle is so tiny. Once you’ve done this you just stop worrying and let the other end of the scale manage itself. This is quite similar to how you’re meant to work if you use the zone system I think.

Once I’d understood this approach, it was clear that the meter is in fact pretty accurate. But at this point I have done no more than play with it. It really is a device that wants to be used for fairly slow photography: it’s not going to replace the averaging meters that most small film cameras have, possibly shoe-mounted, not least because it’s heavier than such meters, and much slower to use.

I do intend to try it in anger when I have recovered the ability to process and print large-format film. In the meantime it is resting in a drawer, from where I occasionally remove it just to hold it: it really is that nice as an object.

Should you get one?

They are beautiful objects. They’re all old, and there are things which can go wrong with them which are fairly final, so the one you get may be a beautiful but nonfunctional object: I fully expected that of mine. They are heavy and not fast to use: you’re not going to be using one of these for street photography or anything like it. If you find one with any of the attachments I alluded to above it is probably very rare. You should, perhaps be rather discrete about what you own in that case: I have heard stories from reliable sources of 3AM knocks on the door. In any case you should check for contamination, both radiological and chemical.

In next month’s exciting episode: either the zone system or the Minolta Autometer II, depending on which I can face.