Canon 10D and Sekonic L358

[dan_tate]

This is a photography student question that has stumped the photo professors

I've asked. My current assignment is a studio portrait which has finally given

me hard evidence to the oddity I've been suspecting for a while now: my Canon

10D and Sekonic L358 differ in their opinions of exposure by about 3 stops.

 

The setup in the studio today was a single Calumet strobe for my key light,

another for my background light. These details are largely insignificant but I

include them so I don't have to answer the question of setup later.

 

In an effort to save a few Polaroids I occasionally check my setup with my Canon

10D before shooting with the 4x5 I use for the assignment. After I composed my

model, I took my meter reading in strobe cable trigger mode: 1/60th and f8 for

ISO 100 film. I then fire off a shot digitally in RAW mode and the histogram and

image preview show 80% of the scene overexposed and totally blown out. I

increase my f-stops until the highlights have calmed down leaving me with the

trial-and-error reading: 1/60th and f22 for ISO 100. Why the discrepancy?

 

I'd also like to take a moment and say I've had similar experiences with this

10D and this Sekonic meter with my Sunpak hotshoe-mounted, full-manual flash I

regularly use so this is not a unique case with the studio strobes. I have not

ruled out my light meter or 10D camera body as oddly calibrated. My results are

consistent though, between the Sekonic being dead-on for film readings and about

3 stops over for my 10D for strobe exposures.

 

Any ideas or suggestions will be appreciated.

 

- Dan

[garry edwards]

Dan,

 

I take it you're using the Sekonic to take incident readings?

Is there any chance that you're using it in REFLECTIVE mode, i.e. without the diffuser in place?

 

This would account for an error of about 2 1/2 stops.

[mike_lepp1]

My sekonic and my 10d agree to within a 1/3 stop.

[brooks short]

What's your setup ? Just kidding though as Garry mentioned, it would be helpful to know if you're doing an incident or reflected reading with the Sekonic.

 

Also it would be good to know what focal length lens are you using on the 4x5, how close to your subject is the camera, and what is your calculated bellows factor for that bellows length?

 

I use a Seconic 358 for both film and Kodak 14SLRN and medium format digital back and it's accurate for both.

[leicaglow]

It might be worth asking in Mamiya's Sekonic forum: http://www.macuserforums.com/webx?14@252.SFWLacxPu5n.0@.ee70439

 

I remember a year or so ago the question arising of meters being 2-3 stops off. Mine has always been spot on with my Nikon and Hasselblad systems, and in comparison to other meters I own. You should also use a grey card with your Canon to ensure you're measuring apples to apples.

[jim_doty]

Dan,

 

Since I'm not there watching what's going on, it is hard to figure out what is wrong.

 

If I understand what you are doing in the studio, you aren't metering with the 10D at all, just taking a picture based on the Sekonic meter reading. The meters can't disagree on readings since the Canon 10D can't do an "incident" flash meter reading.

 

Like others, I'm curious about the Sekoninc meter being in the wrong mode, or set to a reflected reading instead of an incident reading. Is the incident sphere/disc in place? Are you pointing the incident sphere/disc at the light source or the sugject? Is the meter the same distance from the light source as the subject?

 

I'm assuming you are setting both the aperture and shutter on the 10D in manual mode. In Av mode the 10D will drag the shutter to create a proper ambient light exposure which, in addition to the flash exposure, can lead to serious overexposure. Tv mode is worse and you shouldn't use program mode in the studio.

 

The simple test for the Sekonic meter is to get two or three flash meters, all set to incident flash mode and the same ISO. Point all the meters at the light source from the same subject position, fire the flash and compare the readings. They should be close.

 

You can test the 10D meter by metering an 18% gray card in full sun at ISO 100 in Av mode with the aperture set to f/16. The shutter speed should be somewhere around 1/100 second.

 

With the Sekonic meter in incident ambient mode, point the incident sphere/disc at the full sun, ISO 100, aperture at f/16. The shutter should be close to 1/100 second.

 

If you do flash exposure with a dedicated, EX series Canon flash unit, does the 10D meter provide good flash exposures with the subject within the acceptable Guide Number distance for the aperture in use?

 

Jim

[dan_tate]

Wow, thanks for all the responses everyone!

 

Gene Kane - I'll try googling overrated ISO, thanks.

 

Everyone else:

I'm using the incident dome aimed at the strobes from the same location within my setup for each reading. I am not taking incident readings from the 10D, just basing my exposure off the histogram. I'm making the exposure on the 10D in full manual mode setting both shutter speed and aperture. The 4x5 was using a 210mm lens with the bellows extended to about 8 inches, no compensation applied to the readings taken with the 358. I also have a 135mm and a 90mm lens for the 4x5. I've been using the 358 with all my exposures so far with no problems of overexposing film.

 

As for further testing, unfortunately I do not have an EX flash at this time. I will look into borrowing one for some tests.

 

Today is overcast with a forecast of cold, rain and/or snow, and flurries of shooting for homework. I think next week there will be some sun and I'll try the gray card metering. If I remember correctly the meter and the 10D are closer to agreeing in outdoor available light. It's been a while and I will try to test this theory today, cloudy or not. I'll also compare to some reflective readings outside since the 10D is metering through the lens.