1D Mark II + 550EX Flash Test

[scottbergerphoto]

I have read all the posts about the underexposure produced by the combination of the 1D Mark II and 550EX combination. I had a very disappointing experience yesterday with the combination and the 24-70 f/2.8 at the Museum of Natural History in NYC. All the shots were underexposed by 1/2 to 1 stop. I thought it might be a lens issue as I had gotten consistently good exposures with the 85 f/1.8. Tonight I did a test of a number of Canon lenses. The test conditions were as follows:
1D Mark II: Manual Mode, 1/60, f/5.6, Evaluative Metering, Cfn. 14 tested at 0 and 1 for each lens.
550EX: ETTL
Ambient light level: 1.3s at f/5.6
Here are the results:
http://www.pbase.com/scottbergerphoto/1d_mark_ii_flash_test
The pictures were shot JPEG 10, minimal cropping, USM 100, .6, 0
I was shocked to find that the results were all acceptable exposures and that it didin't matter much which lens or Cfn 14 setting was used. Now I'm confused. I don't know why or when the underexposure will reoccur.
Regards,
Scott

[Tom W]

Well, if I were testing, I'd do phase II - I'd take just one of the lenses, and vary the aim of the camera so that each of the various colored objects was center frame. I'd also shoot center focus point, even though focus point isn't supposed to influence the new E-TTL-2 (I don't think).

Did the museum have any weird lighting?

[scottbergerphoto]

Well, if I were testing, I'd do phase II - I'd take just one of the lenses, and vary the aim of the camera so that each of the various colored objects was center frame. I'd also shoot center focus point, even though focus point isn't supposed to influence the new E-TTL-2 (I don't think).

Did the museum have any weird lighting?
Hi. Tom.
All shots were Center AF point. Active AF points have no effect in ETTLII. The next phase of the test will be with lower light. I suspect that something weird happens when the ambient light gets very dark.
Scott

[Mike H]

Scott, thanks for sharing the test results, they are quite helpful.

Only the shots taken with the 16-35 look slightly under (on my monitor), and they are easily fixed. If you really enjoy testing, you might want to follow Tom's lead and try with white, and then black, outfits and backgrounds to see what happens. I'm especially curious about how things might change when you use bounce (rather than direct) flash.

I don't have any test dummies, but I might start looking through my closet for good test shot mixes.

I strongly suspect that you've done this already--but just to make sure--you might want to double check your EXIF data from the shoot that produced the underexposures. If you accidentally dialed in exposure compensation (easy to do with the quick control dial switched on), that could save you some testing time! (I'm glad they've offered the option to take exposure compensation off of the QCD altogether with a personal function.)

Mike H

[Jon Borcik]

I just shot over 132 diffused flash shots of my niece on four different days and time of day using my mkII and 550ex and was extremely impressed with the results. Times were between early morning, late afternoon and sunset (at the beach this week). I have found that underexposure seems to occur with very dark background and little light.

[scottbergerphoto]

I have found that underexposure seems to occur with very dark background and little light.
I suspect that to be true. The lighting conditions in the museum were very dark where the shots were underexposed. You would think that the results would be just the opposite. I tried to duplicate that tonight after the first test by turning the lights in my living room down to very low. The exposures were still dead on. I remain perplexed.
Scott