Highlight Hell

[hugh_hill]

I have been using digital now for over 6 years cameras used are DCS-

520, D30, D60, 10D, & the 1Ds, I have also have the Nikon D100 & the

D70.

I love Canon and choose it almost always over the Nikons, but I am

not saying that the Nikons are any less a camera its just my

personal choice and no more.

 

I have a question and it is regarding the 1Ds,

when I am shooting outside I almost always experience a great deal

of highlights when I am using the 15-30 aspherical yet with the same

settings and a similar lens on the Nikon it very rarely happens, I

cannot understand how this happens, its so frustrating. Is there

something I am missing?

 

I know if I use a tripod and take various shots with different

apperture/speed settings I can get around this problem but for

quality reportage/street shots this is not possible,, This only

happens with this lens (Sigma 15-30mm f:3.5 - 4.5 DG).

 

 

Thanks

 

hughhill@photo.net

[antony_bichon]

Could it be that with the 1Ds, and its full frame sensor, your 15-30 is fooling the meter by getting so much stuff in the picture?

I don't think the Nikon you mentioned are full frame hence include lens stuff hence might get better exposure.

For a picture like the one you attached (eventhough the exposure isn't bad at all) you might benefit from changing your metering system (isn't it backlit anyway?)

 

:)

Antony

[bill_b2]

This can be resolved, Hugh. It involves a little testing, but that is so easy to do on digital.

For an apples-to-apples comparison, be sure of the following:

 

1. Same viewfinder scene. Remember that if Canon is full frame and Nikon is not, using lenses that will cover a full 35mm frame will use only their best part on the Nikon. You would expect to see less dropoff, etc. on the D70, if using full frame Nikkors.

 

2. Same metering scope. Comparing spot to matrix will often give differing results. With spot metering, you can more easily expose for the highlights, avoiding blowout. The spots should be the same size, of course. If that is not available on your Canon, use a substitute target, such as a white card held in your hand, measured from close up. Before internal spot, matrix, etc. metering, a sub was common to do. Sometimes, we just need to be reminded of what we already know and may have practiced in the past.

 

3. Aperture or Shutter priority on both and same ISO.

 

4. If you do the above, you should get the same result on both cameras, regardless of scene appearance. I assume you would be exposing for highlights.

 

5. If results vary:

a. Is Canon complete scene always lighter or darker? Adjust ISO or exposure compensation.

b. Are Canon highlights still blown? Increase ISO or adjust exposure compensation to reduce exposure.

 

If shooting 15mm, full frame, you will need to get very close for a spot to be small enough to cover just the highlight. A substitute subject might be better. This is a longstanding problem with ultra-wide angle lenses. Big street scenes can have a huge dynamic range for a 15mm lens. The telephoto shooter is more apt to choose a scene of more limited range(pure guesswork on my part, by the way.), which will be easier to meter. You do not have to meter every shot unless the light is changing a lot. Unless you are shooting a 10mm Nikkor, which is the equivalent of 15mm full frame in what you see through the Canon, you are comparing and metering differently on each camera.

 

If the differences show up on matrix metering, it just means that Nikon's programmers did a better job regarding highlight exposures, at least as your eyes judge it.

 

Thanks for hanging in there with me on this. I personally guarantee that some part, at least, of this may be correct.

 

Bill

[andrew robertson]

Use less exposure? Your scene includes a large swath of sky, which will make the meter add

exposure to the scene. Maybe the solution is to dial in -2/3 stop exposure compensation

with this lens, or look at the histogram and adjust manually.

[wiley]

Based on comments that I've seen on the Web, a number of photographers (myself included)

have found that the 1Ds consistently overexposes photos by about 1/2 to 2/3 stop. I've just

taken to dialing in -2/3 stop and going from there.

[hugh_hill]

Thank you - Anthony, Bill, Andrew and Wiley.

 

Your knowledge and help is very much appreciated I will be putting some of these methods to the test in around 4 hours when I get up for 'The Liberty Festival' a festival taking place in Trafalgar Square, London from 1-6pm, I have been requested to act as the official photographer for the Mayor who is organising the festival through The GLA (Greater London Authority) Its great knowing there are great guy's like you around always willing to share your knowledge, I will be thinking of you when I am out there....

 

Thanks

 

HH

[poulbh]

No answer to your problems, but as digi cameras have problems with exposure lattitude, giving a lot of blown highlights, its a good practice to always underexpose,as it is much easier , in photoshop, to retreive any lost information in the dark areas as in the light.

[digitmstr]

>>This only happens with this lens (Sigma 15-30mm f:3.5 - 4.5 DG).<<

 

Right. In the sample you posted you can't expect the sky to be on the same EI as the subjects! Either use fill flash or shoot RAW and develop two pictures later, one for each level you desire.