stephen_lutz

Hello,

 

I bought a 70-200 2.8L IS, used. It had been dropped and broken, and

then repaired. The repair was to replace the outer and inner lens

barrel, and the AF and IS. I got the lens with a 6 month warranty

from a local shop I buy from regularly. My question is AF

performance. I went to a baseball game, and was shooting players on

the field from the stands and testing the lens. IS worked fine, and

the AF worked, but didn't seem to be quite right. When I focused at

the minimum focus distance, and then tried to focus on something on

the field, the lens wouldn't auto focus. It would hunt a bit, and

then stop. However, when I turned the MF ring by hand to bring the

scene into approximate focus, the AF would then pick up on the focus

point and work properly. I know that AF works by judging contrast in

the scene and that this could be a camera issue (I was using a

digital Rebel). Is this normal behavior for this lens when used

under the circumstances described above?

 

I shot about 400 shots at the ball game, most with a 1.4 TC, and the

photo quality is excellent and the IS works properly. Even without

the TC, however, the lens exhibits this behavior. Any thoughts on

the matter? Thanks. sl

To check on whether the flash fired, select the file in Zoombrowser, right click it, and select "properties." The details of the shot will then be displayed, including whether a flash was used and if any exposure compensation was used. From the posted photo, it looked like the flash didn't go off.

I have used the omnibounce with mixed results, but have gotten more adept with it. You have to use flash compensation with the omnbounce most of the time. Generally, I'll take a shot, check the histogram and add from +2/3 to +1 1/3 of a stop more flash. One thing I have also done is modify my omnibounce to add a catch light in the eyes of the subject. What I did was use some double sided stick stuff and put aluminum foil inside the rear part of the omnibounce. Thus when I tilt it up (usually 45-90 degrees) the aluminum foil will bounce enough light into the eyes of the subject for a catch light. Here's an example, from a party yesterday.<div></div>

16-35 2.8L is wide, fast and a zoom. It also cost $1300+. I use mine 80% of the time on my 10D and I have no complaints.

I have been practicing with my 550EX, a stofen diffuser, and Canon 10D. I shot some prom pictures last weekend that came out too dark on bounce shots. Many suggested bumping the exposure +1 stop or so. I tried this over the weekend, at a party. I don't know if this is helpful, but here is one taken in P mode, 1/60, f/4. ISO 100, 10D, 16-35 2.8L, 550EX with stofen omnibounce, +2/3 flash exposure compensation. No shadows on the face, and very smooth skin tones.

 

In re: your shots, did you have the Canon "auto fill flash reduction" custom function turned on? If you did, the camera will automatically reduce flash output using an internal algorithm. This is both a good and bad thing. Good, in that you don't have to worry about it in sunlight, since the camera does it automatically. Bad, in the sense that you never know how much flash reduction the camera has selected. I would prefer for Canon to show you somewhere how much flash reduction is being used. Anyway, on my 10D, I turn this custom function off so I can select flash exposure compensation myself.

 

One of the real advantages of digital is "instant review" of exposure and the histogram. When I look at the display, I check out the histogram, not the picture itself. Anyway, good luck, and take careful notes on each practice shot. Good luck!<div></div>

I sometimes shoot at 1/30 of a second in manual mode with flash, but the results are somewhat inconsistent in re: subject sharpness. Should I use 2nd curtain sync so the flash will fire at the end of the exposure? Or perhaps this would make no difference? I am shooting some band photos tonight and will give some of these ideas a try. Thanks.

Thank you, that's very helpful. I knew when I chose 1/60 I was selecting a shutter speed well below the speed that would be needed if ambient light alone was used. 1/60 is sort of the default setting for flash to ensure subject sharpness by freezing motion blur. F/2.8 would have been what the camera would have selected if I had left it in program mode, but this would have left the photo softish on anything but the focus point. I wanted to ensure some degree of sharpness on all the subjects. I shot a couple of "safeties" with direct flash and they turned out ok. I considered using ISO 400, but decided to test the camera/flash to its limits. In any event, what was somewhat surprising to me was that the flash didn't pop enough light. I checked my flash confirmation light, and it glowed OK.

 

I have a stofen omnibounce attachment for my flash, but forgot to bring it. Next time I'll try this: I will add a stop, and use the stofen at a 60 degree angle. I will check the histogram for exposure confirmation, rather than rely on the flash confirmation light. Thanks again to all who responded. Any other suggestions would be welcome and appreciated, as I am trying to learn this E-TTL system and improve my flash photography.

I had the camera set to partial metering, and used direct flash. The way I understand Canon's flash system, the camera ALWAYS uses Evaluative metering with flash, regardless of what metering mode is selected on the camera. For the most part, I have found that when I set the exposure compensation up with direct flash I get blown out highlights. However, the next time I bounce flash, I will add a stop of exposure. Obviously, the bounced photos are too dark, so more light is needed. My curiousity is why the flash didn't provide enough light that close to the subject. Even at ISO 100 the 550EX should have had enough power for bounce in this situation. Perhaps the ETTL preflash metered on something incorrectly and it messed up the exposure? Anybody have an idea how ETTL meters when you bounce the flash? What would the camera have metered on in this scene? I locked exposure on one of their faces. With direct flash, this worked fine, with bounced flash, I got dark pictures. Grrrrrr. Next time, with bounce, I'll add a stop just to see what happens. Here is my stepdaughter, with direct flash, exposure locked on her face. Thanks!<div></div>

Thanks for your suggestions so far. I had two lenses, the 16-35 2.8L and the 28-70 2.8L. The ones I shot with the 28-70, with bounce flash, were very dark. The ones with the 16-35 were considerably better. Direct flash, with both lenses was fine, but I was careful to lock the flash exposure on one of their faces. Marc's adjustment looks very similar to the results I got with an autolevels adjustment in PS 6. I did a manual adjustment because the autolevels looked a little too contrasty to me, so I tweaked it to my taste. In any event, flash photography with my digital camera at least gives me results to look at right away. I hated it with film because I never knew what I was going to get until I got the film developed.

 

You're right about better to underexpose than over expose. Overexposure is unrecoverable. Of course what I want is correct exposure! :)

 

Here's one I shot with the 16-35 2.8L, at 16mm, with the 10D at ISO 100 and direct flash. Turned out pretty well. The exposure is pretty good, and there aren't any obnoxious shadows. Shot in manual mode. 1/60, f/4 to ensure DOF with the whole group.<div></div>

Here's a crop of the the fixed photo, showing detail.<div></div>

Here they are after a levels adjustment in Photoshop.

Hello,

 

I shoot a lot of social events, rather than weddings. Most of these

are for friends and family, but I have shot events for pay. Just

the other day, I shot some for my stepdaughter, before she went to

her prom. I have a question about lighting small groups with bounce

flash. I shot the attached photo with a 10D (ISO 100), a 16-35 2.8L

and 550EX bounced straight up on an eight foot ceiling. I shot in

manual mode, 1/60, f/4 because I didn't want the camera to default

to f/2.8 and leave some of them blurry. Overall, the shot was

dark. I improved it considerably using levels in PS 6, but wonder

what sort of technique or flash setting would provide enough flash

power to light this scene without having to salvage it in photo

shop. Obviously, shooting at a higher ISO is one way to do it, but

with a 550EX, from close in, I would have thought I would have

plenty of flash power. The photo isn't really underexposed so much

as devoid of any highlights. The histogram was normal looking and

evenly distributed, but went flat 2/3 of the way across the graph. I

will post the original and then the salvaged picture. Thanks in

advance for any help!<div></div>

At ISO 100, in "Large/fine" jpg mode, I have gotten super sharp 13x19 prints from my Canon s9100 inkjet of photos taken with a 10D. The prints I have gotten are as sharp as any I have gotten from a lab. Absolutely no pixelization was evident to the naked eye. I didn't look at them under a loupe, but I suppose if you magnify them enough they will degrade. Better yet, the prints looked very natural, very smooth, not "digital" at all. Of course, the 16-35 2.8L I was using may have a lot to do with that.

The 1D is analogous to the 10D as the 10D is to the Digital Rebel. The only "advantage" of the 10D is pixel count. I have seen 13x19 prints made on an Epson 1280, taken in JPG mode with the 1D that are flat out terrific, so the extra 2 MP of the 10D doesn't seem to be that big of a deal. In any event, it is really a matter of whether you have the money, and will put up with a big heavy camera. The 1D is far superior to the 10D in every way. (Not that the 10D is a bad camera. Far from it. I own one, and it is excellent, but it simply is no match for the 1D.)

With flash, you are essentially making a double exposure, one of the foreground (with flash) the other of the background (without). The camera meters the whole scene and then adds flash to the subject. You can control both exposures independently. Underexpose the scene and leave the flash alone, and you darken the background, for example.

 

I suggest you:

 

1) Turn off the auto fill flash function (via a custom function), so you can control the flash the way you want.

2) Leave the flash with no compensation, so the bride and groom stand out. Underexpose the background 1/2 a stop, to balance the exposure.

3) Use Av mode, and FP mode on the flash, but try to keep the shutter speed below 1/200 of a second. 1/200 @ f/8 should be fine if it is sunny outside.

 

This should work, but it's easier to test the effects with a digital camera. This works pretty well with my 10D, at ISO 100, 200 or 400. I adjust ISO sometimes if it is a little dark or light out.

I don't shoot weddings, but do shoot parties, events, etc. I use my 16-35 2.8L 90% of the time on my 10D for indoor shots. I just can't wide enough when I am in tight quarters with any other lens. Plus, I can hold the camera up over my head, zoom out to 16mm and get a crowded dance floor shot. Those are hit or miss, of course, but still can turn out pretty well. Here's one shot over my head, at 16mm, a 10D at ISO 800 and a 550EX flash.<div></div>

I haven't done any "tests" per se, just shot about 100 pictures with the Digital Rebel and kit lens. I own the 16-35 2.8L, which is my "standard" lense on my 10D. I also have 5 other L lenses, so I know first-hand how fantastic they are.

 

I walked around today, and shot a variety of subjects with the kit lens, and then printed a couple out on my Canon CP-200 printer. The 4x6 prints were sharp, colorful and pleasing to the eye. All in all, I thought, "Well, why not? The prints look good. The shots look good on my computer screen."

 

Is it an L? No, of course not. Does it take good photos in "general photography" situations? Seems to. It certainly isn't the dog that some have said it is, at least in my experience.

 

Oh, I shot some New Year's Eve party photos with the Digital Rebel and 16-35 2.8L and they were terrific. I don't know if the difference would be all that noticable from the 18-55 kit lens, at a 4x6 print size.

I will probably buy a digital Rebel to replace the D-30. The main reason is that the D-Rebel has the same sensor size as the 10D, and the image quality should be the same. The default sharpening parameter on the D-Rebel is the same custom setting I use on my 10D. The D-Rebel doesn't have the "favorite" focus point button, nor does it have the control wheel on the back. No custom functions on the D-Rebel, either. The absence of flash compensation is a drag, but I can do that on my 550EX anyway. All in all, the D-Rebel is a very attractive camera for its intended market: first time D-SLR owners. The D-Rebel may be a "loss leader" in that it may not be too profitable, but the gravy for Canon is the lenses, accessories, etc. that people will buy with it. I would imagine this will put a serious crimp on the new 4/3 sensor cameras, like the new D-SLR from Olympus and force Nikon's hand with a roll out for a D-SLR N-75.