Silverbased

Projects and ponderings for film photographers

Archive for March, 2008


Shallower Depth of Field?

Few issues could be more vital to image-making than knowing what is or isn’t in focus. Yet figuring depth of field correctly seems to endlessly bedevil photographers. The more subtle aspects of DOF (to use its geeky nickname) have enough twists and turns that just the Wikipedia entry on the topic becomes quite an exhausting read.

But as digital cameras have come to dominance, there’s increasing awareness that small-format sensors exact a penalty—limiting the options to use shallow depth of field creatively. I’ve written about this myself in an earlier post.

Photos with minimal depth of field can have a classic, even romantic quality, which can be quite nice for certain subjects. The iconic example is a close-up portrait where someone’s eyes are sharp, but even the rest of their head dissolves away into softness.

Portrait with Shallow Depth of Field

Pleasingly shallow DOF from a Zuiko 85mm f/2.0 portrait lens. Any lens gives its thinnest depth of field when used at close focus.

Now these days, there any number of jazzy lenses on the market: Primes of f/1.4 or even faster, super telephotos, and all manner of zooms. So what should you chose if you want the absolute minimum, whisker-thin depth of field?

Which Focal Length?

Your first instinct might be that a telephoto would give shallower depth of field than a normal lens—and that’s halfway right.

But to make a fair comparison, you need to move the camera position until each lens includes the same subject height within the frame. And when you satisfy this condition, all lenses of the same f/number give the same depth of field—or darned close—providing we don’t change the size of the image format.

Play with one of those online DOF calculators and check for yourself, if you don’t believe me.

Yet subjectively, telephotos do have a different look. If you switch to a lens of twice the focal length, and then scootch back twice as far from your subject to restore the same framing, chances are your distance to the background didn’t increase by nearly the same proportion. So longer lenses still enlarge any background objects; and their blur seems more pronounced.

Conversely, a wider-coverage lens captures a broader sweep of background details. Even if you move closer to your subject to blur those distracting eyesores, more of them still remain in the frame; and so you have less chance to exclude them by shifting the camera from side to side.

This is why, practically speaking, a longer focal length is still preferable if you want to deliberately “disappear” the background.

Bigger and Longer

As focal lengths increase, maintaining the same widest f/stop means the diameter of the glass must increase in proportion. But remember that doubling the lens diameter quadruples its surface area (and hence the amount of grinding & polishing needed); and the volume of the glass elements (and thus their weight) increases by a factor of eight. This is why fast lenses become increasingly exotic and expensive in longer focal lengths.

For any given format, the fastest lenses tend to be the “normal” focal lengths. Yet these aren’t the best choice for a frame-filling head & shoulders portrait. There’s a well-known problem where getting too close to the subject leads to bulging, exaggerated noses and tiny, disappearing ears. However shooting from too far away (and thus seeing too much of the sides of the head) can give a “balloon face” look that’s unnatural too.

So to work at a distance that gives a flattering perspective for portraits, we’d prefer a focal length ranging somewhere from 1.5 to 2.5 times the image diagonal.

Which Format?
My personal standard for nice depth of field is my Zuiko 85mm f/2.0. This is still a much-sought-after OM System lens, but nothing wildly exotic—it generally sells for a couple hundred dollars on eBay. Shooting wide open from 4 feet away, the DOF is about 3/4 of an inch.

But as I discussed in my earlier post, depth of field deepens with smaller image formats and shrinks with larger ones. So I spent an amusing morning calculating what f/stop would be required on other film formats or digital chip sizes to yield equivalently-shallow DOF.

To answer that question, I experimented with plugging various numbers into this online calculator, to see what aperture would be needed. (That particular page is especially useful, since you can input arbitrary custom values for aperture, focal length, circle of confusion, etc.) *

To get DOF equivalent to my Zuiko, the other formats would require a comparable lens with the following maximum apertures—

  • Four-Thirds System DSLR: f/1.0 [sorry, none available!]
  • 1.6x crop factor DSLR (e.g. Canon): f/1.2
  • Leica digital M8 (1.33x crop factor): f/1.4

and what about larger film formats?

  • 6×6 and 645 format film: f/3.5
  • 6×7 and 6×9 format film: f/5
  • 4×5 inch film: f/9

Now, ultrafast small-format lenses get into some costly territory. But wow—finding medium format lenses matching those speeds is not that hard! Actually, these days they’re shockingly affordable on the used market.

Mamiya 645, 80/1.9

Mamiya 645 with 80mm f/1.9—sadly not a portrait lens, but still crazy-shallow DOF at fire-sale prices

Portrait lenses of f/3.5 or faster were made for Bronica, Mamiya, Pentax, and Kowa medium-format systems. Mamiya & Pentax made many better-than-f/5 lenses for their 6×7 cameras. Bargain hunters might seek out a Kiev-60 or Pentacon Six system, to try the “eastern bloc” f/2.8 portrait lenses which fit them.

Onward to Exotica

Naturally, a 50mm Leica Noctilux at f/1.0 (!), mounted on a digital M8 body would yield even shallower DOF. It works out to a mere 1/2″deep, if used at the distance that matches the coverage of my Zuiko at 4 feet. (The ‘Noct is a little shy of true portrait focal length on the M8 sensor.) But the total $11,500 price tag might be a little daunting!

Otherwise, 85mm f/1.4 lenses for a full 35mm frame do exist—which yield similarly paper-thin DOF. But they clock in at about $1000 each. Canon even makes an f/1.2 version, costing somewhere north of $1,700.

These ultrafast thoroughbreds may be a godsend in extreme low light; but with typical daylight levels and ISO speeds, it becomes problematic to use them wide open. The moderate widest f/stops of medium-format lenses may more practical to handle.

Larger Film to the Rescue—Again

Yet you can even get the same same razor-thin DOF as the Noctilux or those exotic 85’s—but with nowhere near the same cash outlay. An example would be a nice old Pentax 6×7, fitted with one of their 150 or 165mm f/2.8 portrait lenses. At today’s eBay prices, this combo seems to sell for four hundred dollars or so.

Or, if you move up to a 4×5 camera, any large-format lens of f/5.6 aperture can do the same. Plus—think of the smooth tones, the amazing detail! (Admittedly, the size and unwieldiness of a 4×5 setup might weigh against using it for casual portraiture… )

For my money, the sweet spot for shallow depth of field happens with medium format cameras. The 110mm f/2.8 I got for my Mamiya 645 even ekes out slightly shallower DOF than my Zuiko comparison—yet as “bargain” grade from KEH it was less than $70, shipped.

*I wanted to be fair to all formats, regardless of their differing aspect ratios. So I noted the largest crop from each negative size which would yield a print of 3:4 aspect ratio. Then I scaled the equivalent circle of confusion and focal lengths using these image sizes. The 3:4 ratio is simply a compromise between the various format proportions—not an artistic preference, nor any special law of optics.

Siphon Film Washing: A Better Way

Okay, you sun-belters out there quit snickering right now. Here in Michigan, it’s definitely still winter. (March 27th dumped yet another mini-blizzard on us.)

After many months of low temperatures, the earth our city water pipes pass through gets rather chilly. The other day, I measured the water from my cold tap at below 43°F (6°C). Ouch! This presents a problem when you want to mix photo chemicals; or when you need to rinse developed film and prints.

Frigid Waters

Whoah, bummer.

Water of such low temperature is not effective at removing the fixer that remains in a gelatin emulsion. And years later, that fixer residue might cause brown stains on the image. Also, the shock of going from developer temperature (68°F) to something so cold could make a tender negative emulsion shrink and crinkle.

A negative is irreplaceable. So I generally exceed most manufacturers’ wash recommendations—erring on the side of caution and going for 15 or 20 minutes. And Lo, how I have often fiddled with balky hot and cold water taps, trying to get the rinse temperature to stay at a constant 70°F or so during that whole time.

There are really several irritations with this method:

  • Dissolved gas bubbles from the hot water always cling to the film surface. Does this interfere with washing efficiency? I don’t know, but I worry it might.
  • No matter how much I fiddle, the water temperature always drifts from the temperature I set by tweaking the tap handles (yes, I know thermostatic taps exist, but one at the kitchen sink where I develop film seems like overkill).
  • I often walk away to do something else while the film is washing—then forget I’ve left the taps running until an hour later, wasting water.
  • In fact, I only have a vague sense of how much water the running-taps method uses; and little confidence that my washing is consistent from film to film.

What are we trying to do here?

Now, my hunch is that any liquid fixer clinging to the surface of the film is rinsed away within a few seconds. The purpose of film washing is to get the fixer molecules that have diffused into the gelatin to migrate outwards again.

For this purpose you shouldn’t need to blast the film with extreme water velocities. You only need enough water flow to give a good diffusion gradient, from the fixer embedded in the emulsion to the fresh water outside it. (I suspect that the time the film remains submerged in the rinse water is more important than the water velocity.)

So this month I had a brainwave for a new low-velocity method for rinsing film, one that solves several of my earlier problems.

The New Way

First, fill a 5 gallon bucket with 70°F/21°C water, and allow it to stand until all the dissolved gas bubbles float to the surface. Obviously you need to use a well-scrubbed bucket, without any residues from its prior uses. I might set my stopbath and fixer bottles into this water for a half an hour before starting to develop, to equalize them to the proper processing temperature.

Take a length of tubing 4 feet long (120cm), and with an inside diameter of about 3/16″ (5mm). Attach this to the lip of the bucket in whatever convenient way will suspend one end at the bottom. Here I’ve used a binder clip:

Bucket, Binder-Clip

Five-gallon bucket, tubing, & binder clip

Place your film tank in a sink where the rinse water can overflow from its top; support the bucket at a height where its bottom is at least 8 inches above the top of the tank.

When you are ready to begin rinsing, suck on the free end of the tubing until water begins siphoning out of the bucket—keep the tubing’s end below the water level.

Fill the tank once with water, then dump it all out. Then stick the tubing into the center spindle of your film reel; and once you’ve confirmed that water is overflowing from the top of the tank… walk away!

Siphon Washing

Siphon washing at work

After about 20 minutes the bucket will empty. A soft gurgle will announce the end of your film rinse. An unexpected bonus is that otherwise, this method is totally silent—unlike the distracting whistle of keeping taps running.

The water flow will be highest at the beginning, then slow down as the water level drops—which is appropriate for rinsing away fixer. There’s no clinging bubbles; and no surprise temperature lurches if someone in the house flushes a toilet. And you always know exactly how much rinse water you used.

If you’re feeling especially thorough, you may wish to check how the rinse is going at the halfway point. Dump out all the water and invert the reel, just to insure totally uniform washing.

This inspiration for this new method only came to me after, oh, 40 years of developing my own film—struggling with the minor irritations of the running-taps method. Oh well. That’s another thing I love about photography: You never run out of new things to learn.

[In case you’ve never developed your own B&W negatives, it’s easy—see this earlier article to get started.]

Quirks of the Zinc-Air Cell, Part 2

In an earlier post I discussed the problem of old photo gear designed for 1.35 volt mercury cells—a battery type banned in recent years because of the metal’s toxicity.

I concluded that the easiest solution for most people is to substitute an inexpensive #675 hearing-aid battery (with a spacer ring if needed to fit your battery compartment). Their zinc-air chemistry is a near match to the voltage of the original mercury cells.

Mercury, Hearing-aid, & Wein cells

PX13 mercury, #675 hearing-aid, and Wein cell batteries

I also noted that there is a more expensive commercial product, the “Wein cell,” which is also a zinc-air type. But these are specifically marketed as mercury-replacement batteries for photo equipment.

While Wein cells are indeed effective for this, photographers have often speculated whether there is anything special about them. Are they simply standard hearing-aid batteries repackaged at a higher price?

The Wein replacement for a PX13 or PX625 battery does have some physical differences from a 675. The Wein’s metal case only has two air holes, not four. This helps keep them from drying out as quickly.

The Wein battery has the same dimensions as the 675 hearing-aid size, but adds a metal ring crimped around it. This can be important in making electrical contact with some battery compartments, for example the metered Yashica Mat models.

But is the Wein cell chemistry any different?

I have no insider knowledge of how Wein produces these batteries; and my only test equipment consists of a simple Radio Shack multi-meter.

Yet I made some curious discoveries about the voltage of zinc-air batteries (of both types), which photographers using these cells ought to know about.

While researching my earlier post, I became curious exactly how much current a typical camera meter circuit was drawing. By making temporary connections to the battery compartment of my Olympus OM-1, I found that with its meter needle centered, it drew roughly 0.2 milliamps from its PX-13 mercury cell.

It was more convenient and accurate to use a simple resistor rather than a camera as my dummy load when comparing batteries. So I used a 6.8 kΩ value I had handy, which drew a current comparable to the OM’s meter.

Battery Voltage Test with Dummy Load

Testing zinc-air battery voltage with a resistor dummy load—hmm!

At this point I discovered an interesting anomaly. The voltage of a mercury or silver-oxide battery scarcely changed when put under this load—a drop of maybe 0.005 volts. But with either type of zinc-air cell, the story is very different.

Under no load, the zinc-air readings begin at about 1.42 volts. But, when connected to the load, their voltage begins to drop significantly. After five minutes or so, the 675 hearing aid battery dropped to 1.37 volts. The Wein cell went down to 1.33 volts. All these numbers showed slight variations when I repeated the tests.

I wondered if the Wein cell’s fewer air holes were “starving” the cell for oxygen, causing the lower voltage. But covering two air holes on a 675 cell didn’t seem to change its voltage readings.

Now, my original PX13 mercury battery tested at about 1.34 volts under load. So there does seem to be some support for the claim that Wein has somehow “tweaked” the zinc-air chemistry, to give a voltage closer to that of mercury cells.

But is this difference important?

Hardly! Far more significant for accurate exposures is that you turn on the meter several minutes before taking a reading. With either zinc-air battery type, the voltage will be too high immediately after turning on the camera—possibly leading to underexposure. Once you become aware of this effect, you will definitely notice your needle drifting for the first few minutes, even when metering a scene with unchanging light.

Now, the “gold standard” for mercury battery replacements is to use a silver oxide battery (approximately 1.58 volts) with an adapter that adds a miniature diode to the circuit. By introducing this small voltage drop, a very close approximation to the mercury battery voltage can be obtained. The voltage doesn’t drift after being switched on.

But in fairness, we should note that even this solution can give different voltages depending on the specific diode chosen, and the current draw of the meter circuit. The voltage also changes somewhat with temperature. Frans de Gruijter discusses both these factors at some length in his definitive battery-adapter article downloadable here.

Given all these uncertainties, I continue to feel hearing-aid batteries are the most sensible real-world mercury replacement for many photographers. Not perfect perhaps—but close enough.

But remember, don’t try to “save the batteries” by constantly switching the meter on and off. When you go out shooting, switch on the camera, uncover the lens cap, and leave it that way. Zinc-air batteries have a high capacity; they’re less likely to die from electrical exhaustion than from the inevitable drying out of the chemical paste inside them.

After that you’ll need to splurge on another 75-cent hearing aid battery. Then grab that vintage camera, get out there, and keep shooting.

See this earlier post for more about the mercury battery issue.