# Thread: Exposure measurement at dawn/dusk

1. ## Exposure measurement at dawn/dusk

I am having some difficulties calculating exposure when photographing at dawn and dusk, as exposure times are long and light changes rapidly. What I would like to is to come up with a good approximation that can yield acceptable result. I am using color transparency film, so exposure has to be fairly accurate.

I usually shoot at around EV 1 to 5 (8 min at f32 to 2 min at f64?). By the end of exposure time, the light level has increased or decreased by 1-2 stops.

My question is: Is there a good shorthand method of measuring during exposure and compensating exposure time as the light changes?

Kind regards, Lars (at the moment in outback Australia)

2. ## Exposure measurement at dawn/dusk

Lars, I think you need to experiment. Measure the EV over a period of, say, 10 minutes at an appropriate time of day. Plot the values on the graph - the required exposure will be the integral.

3. ## Exposure measurement at dawn/dusk

I spent the better part of the 1960's in Hollywood assisting the late Sid Avery shoot Japanese automobiles on location around Southern California. These photographs were made on 4x5 Ektachrome from first light until sunrise and again from sunset until dark. At these times the sunlight bounced off of all sorts of clouds and pollution at the horizon and the light levels and (kelvin) colors were all over the place.

Our method consisted of having one assistant with a Spectra Combi 500 incident meter pointed straight up at the sky constantly calling out f-stops as the light changed. Another assistant stood by the front of the camera adjusting the aperture and cocking the shutter as Mr. Avery handled the film holders and cable release.

We would expose fifty sheets of film in (the golden) ten minutes. While we were able to keep the film density fairly constant, the color of a red car would photograph as every hue from orange through magenta as the light changed. No time to play around with CC filters.

4. ## Exposure measurement at dawn/dusk

You may want to experiment with different film tyoes as well. Provia, for instance has a much wider exposure latitude than other 4x5 transparency film.

5. ## Exposure measurement at dawn/dusk

Interesting problem, Lars. Thinking through the problem, it seems to me there ought to be a way of pre-calculating the various factors that come into play, and arrive at the shorthand compensation method you're looking for. But, it also strikes me that the short hand may have longish, but variable, fingers, so to speak. I'm not sure how to do the math, but here are the factors I think might come into play.

I haven't tested this, but my guess would be that the morning light level with a cloudless sky will be a swooping curve, where the light level and rate of change increase as you get closer to sunrise. Where your initial exposure reading falls on that curve would depend on how close you are to local sunrise time. Thus, placement on the curve, and your proximity to local sunrise time, would also affect the amount of compensation necessary. The reverse (or, inverse ?) would be true for sunset, I'd think.

For this purpose, it may be helpful to think of the exposure in terms of something like lumen-seconds, if there is such a thing. What you're looking for, I think, is the cumulative exposure, or essentially the quantity of photons needed to arrive at your desired exposure. That "desired exposure" is also affected by where your image objective fall in the range from "photographing the object" to "photographing the light" - that is, whether you want an object within the light to be exposed "normally", or if you want the image to give the sense of the time of day and quality of the light.

Once the math is done, you should be able to arrive at your desired compensation factor, or "multiplier" of the initial exposure reading. In that respect, your proximity to local sunrise time strikes me as being the key factor. While I think the multiplier is always going to be less than 1.0 for morning shots (greater than 1.0 for sunset shots), I'm guessing that the rate that it gets smaller (or larger) increases as you get closer to sunrise (or sunset).

If clouds are in the sky, the pre-calculations would be affected considerably, but in a strongly variable way. Size, altitude, and movement of the clouds (acting as pre-dawn reflectors) complicate the calculation in a way that could not be predicted, I'd think.

I'm not sure my ramblings are really helpful, but they might spark some additional ideas.

6. ## Exposure measurement at dawn/dusk

I think that sometimes it's possible to get too technical. I recently was shooting a landscape at dusk with a 4x5 using 160 Porto VC negative film. Certainly the lattitude is a couple of stops wider than chrome, but I think the principle is close. There was no published reciprocity data for the film - all I knew that the film should be good from 1/10,000 to 1/10 of a second. So I did what I sometimes do - decided to take a test shot or two, examine the results, and see how far off the film seemed to my eye. According to a reflective light meter, not a spot meter, my exposure would be something like three minutes at f64. I took the photo, had it developed, and was quite happy when it actually came out pretty well - maybe a half stop over exposed.

I've done the same thing even with roll film - taking shots, trying out different exposures (while recording them, or at least the EV difference between what the meter said and what I did), and seeing which of the results pleased my eye. It's not as scientific as what many people do, but I like the results, and it generally only takes one test. With luck, one of the test shots even ends up as what I wanted in the first place.

7. ## Exposure measurement at dawn/dusk

A few notes on the conditions here: I am in the outback shooting 8x10. There is no 8x10 Provia in sight, possibly not available in Australia except as special order. I talked to E6Plus, one of the pro labs in Melbourne - the owner had not developed any 8x10 chromes for over a year. I will not see any developed results until I am finished with four months of shooting, so testing is not really an option.

Erik, with all due respect I think that negative film is enough forgiving that the comparison is not quite valid. I am a little confused about your description - I take it that you are not referring to bracketing, but rather to testing a certain film at a certain light level? Another (implied?) good point of yours is that perhaps negative film is better suited for these light levels. (Is the new Kodak UC 100 available in 8x10?)

Ralph, I love ramblings, and I think you are pushing in the right direction... And I have tried to develop your line of reasoning below.

Now, my situation is that I am in the middle of nowhere - literally. I just drove down a gravel road 630 miles long, with no towns and only one gas station which is closed Tuesday to Thursday - plus I am on the road with no fixed address so shipping off film to be developed in Sydney or Melbourne is a bit impractical.

I am also traveling, which means that I usually only get one shot at each location - and only one shot per dawn setup and one shot per dusk setup. (Unless I wait 24 hours to reshoot the same composition, but I do want to avoid that kind of bracketing if possible.)

My own thoughts on this problem, with some input from Ralph above, are as follows:

- Avoid exposure times that lead to more than +/- 1 stop change in light levels (should be around 3-10 minutes, depending on a zillion conditions);

- At dawn choose the largest acceptable aperture, and at dusk choose one stop smaller than largest acceptable aperture.

- Keep exposure time fixed (makes reciprocity failure compensation calculations simpler);

- Measure light before and during exposure, and adjust aperture along the way, at for example, 0, 25, 50, and 75 percent of exposure time.

The alternative, which is what I have been doing up till now, is to shorten or extend exposure time, but I find the involved calculations to be quite difficult to do on the fly (there is absolutely no time to take notes!).

8. ## Exposure measurement at dawn/dusk

Get the Olympus OM-2n which has off the film metering. It will close the shutter when the film has enough light (up to two minutes)! and the light meter goes down to -5,5 EV ;-)

http://www.mir.com.my/rb/photography/hardwares/classics/olympusom1n2/om2/index.htm

Øyvind

9. ## Exposure measurement at dawn/dusk

Lars, yes, I was talking about testing a film, not bracketing. That implies an amount of time and access to processing that you clearly don't have, though the approach does work with chrome - like the lazy man's way of finding personal exposure ratings for particular films. All I was saying was that I was actually testing the Porto when I had no reciprocity information at all, and as luck would have it, it actually worked decently at a two minute exposure. Negative film would give you more lattitude - and if you ended up scanning the results, you could always increase the color saturation a bit to try and get closer to what a transparency might provide.

Even if you are moving about, would you ever know roughly where you might be a week or so in advance? Hikers doing the Appalachian Trail in the US, which runs something like 1500 miles, will often ship themselves supplies under "general delivery" to post offices along the way. The post office keeps the package for pick-up by the person. Could you do the same, maybe sending out a couple of your transparencies to a lab and having them returned to a spot where you could "meet" them to be sure things were working?

As far as changing light, I think the idea of sticking to exposure times that don't result in major changes in the amount of light is wise. I wouldn't try adjusting the aperature during an exposure, as you'd run the bigger risk of jostling the lens and reducing the sharpness of the image. If you do have to use the longer exposure and experience tells you that the light will be two stops off by the end, then my guess is that it would be safe to model the fall off as a linear function, and to take the average of the readings that your experience is showing. So, if it drops by two stops, increase your exposure by one by opening the aperature, as increasing the time would just complicate things even more.

One other possible solution, though maybe not one that you'd like aesthetically, would be to have your film push processed. Yes, you'd increase the grain some, but given the size of your film, it might not be that big a problem. You'd half your exposure times - possibly more with the reciprocity factor - which could take care of the issue of fall-off.

By the way, I checked Kodak's site and the UC films are not available in sheets. The Porta 160VC (more saturated version) is available in sheets, as is the 400NC (I think it means natural color).

10. ## Exposure measurement at dawn/dusk

Here's another thought for you, Lars. "Invest" one morning and one evening in measurement taking. Let's say, take incident readings (dome straight up) every minute for 30 minutes before sunrise, and 30 minutes after sunrise. Do the same for before and after sunset. Then, plot each set of EV readings on a corresponding sheet of graph paper. Connect the dots to see if the rate of change is linear or non-linear.

As both sunrise and sunset are precisely predictable based on location, and can be downloaded in advance from the NOAA site (http://www.srrb.noaa.gov/highlights/sunrise/sunrise.html), you can estimate the amount of correction (the multiplier) based on how long before or after sunrise/sunset you start your exposure. If you have a GPS unit, it might already have the sunrise/sunset calculation function built-in.

Take the duration of your typical exposure time as the increment, and then for each increment of time before and after sunrise/sunset, take an average of the EV readings off your graph. Divide the reading at the beginning of that increment by the average to arrive at a multiplier for that increment, and make a little card with the increment times and multipliers - essentially, pre-calculating the amount of correction. Based on the corrected time, you can also pre-calculate reciprocity for the film(s) you are using, and add that to the chart.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•