Why meter off sky, then change aperture and shutter speed?

Question

In section "Night and Low-Light Phtography" that starts on page 128 of Bryan Peterson's "Understanding Exposure" book, he has several examples where he (spot) meters off the sky with a large aperture, then recomposes, reducing the aperture and increasing the shutter speed in order to maintain the same exposure. My question is: why does he not simply take the spot meter reading using the desired aperture or shutter speed? For example, on p128:

... I began with my aperture at f/2.8 and pointed the camera toward the sky... I then adjusted my shutter speed to 1/4 sec... I set the lens to f/22 and simply did the math to determine the new and correct exposure ... I needed to increase the exposure time to 30 seconds.

Would he have obtained the same result by taking a meter reading off the sky with the aperture set to f/22? Perhaps the 30s shutter speed in this example would make this a challenge for the camera's metering system; however, in other examples he does the same, but increasing the shutter speed to, say 4s.

Answer 1

With an old fully manual aperture lens you want to meter wide open so you give the metering sensor as much light to work with as possible. However with a fully electronic camera the aperture stays open 'till the point of exposure so all meter readings are taken wide open. With this in mind I can't think of any reason why you'd want to meter with the camera set to f/2.8 and then do the calculation yourself - instead of letting the camera do it! The only exception is if you're using bulb mode and need to calculate exposures of longer than 30 seconds as the camera wont do this.

My approach to metering for night shots is to simply shoot a bunch of exposures and look at the histogram - I find the camera's main sensor to be the most accurate meter :) If I want to save time I'll use a higher ISO to shorten the exposures and then multiply up the shutter time.

Answer 2

You're right, he mostly likely could have set his camera to f/22 first and then metered. But that doesn't appear to be how he works.

One of the major principles he talks about in Understanding Exposure is that there are often at least 6 combinations of aperture and shutter speed that will give you the same technically correct exposure, but after you've found one it's also up to you as an artist to select the proper one that gives you the desired artistic depth of field and motion blur.

So in describing the technical details of his photographs, he often describes that he first determines the correct technical exposure, and then changes his settings to create the desired artistic exposure. Like this:

1. In manual mode, meter the scene and zero the meter.
2. After you've zeroed the meter, adjust aperture or shutter speed to achieve the desired artistic affect.
3. Adjust the other setting to re-zero the meter. For example, if you opened up the aperture, you'll need a faster shutter speed.
4. Take the exposure.

The example you've provided from his book sounds like it matches that pattern.

Answer 3

As another possibility, I wonder if he's trying to more accurately meter, considering reciprocity failure.

I found two references for it from Wikipedia and Geoff Lawerence.

The Geoff Lawerence link is more succinct.

Reciprocity failure, as I understand it, is that as your exposure time is increased, metering accuracy decreases. The link on Wikiedia only refers specifically to film, though it also uses the phrase, "For most photographic materials...". So this could also include a digital sensor.

So he may be thinking that by metering for a shorter exposure, then extending his actual shot, he'd get a more accurate exposure. This assumes that the meter is more accurate for shorter exposures.

However, it may as simple as he though he'd wouldn't be able to use the in-built shutter timer...