Using a flash in a dark room

I was doing some experimentation this weekend at a friend's wedding at applying some flash usage techniques I had

read about on the Planet Neil's blog, as well as the Strobists blog.

I was using the a Sigma EF500 DG Super on a Digital Rebel XT + Tamron 17-50 f2.8, with the flash in full ETTL

mode, and well, to sum it up, I was quite disappointed in the results. I doubt it was my gear that was failing

me, so I'm sure it was 100% due to my in-experience, but it did bring up a lot of questions to which I thought

I'd ask here:

 

- One thing I noticed as I bounced the flash off different walls or ceilings is that I never knew how much flash

power I was using, or rather if my flash was being maxed out or not.

A lot of the pictures looked dark on the histogram, but then again, shooting at shutter speeds of 1/100 or so,

will make most of the background pretty dark. Only indication of overexposure was the highlight warning.

So since the histogram showed an image biased towards the left, I would add some flash compensation. Sometimes it

would work in brightening the image, other times it would not. But I never really knew if that was because I had

too low an ISO or too high an f-stop, and the flash was being maxed out or not.

The room was not huge (maybe 20x50 ft or 30x60ft), and for the most part, the lights were dimmed (during the

dinner, and rest of the reception)

 

- The other issue I had was that since I was ALWAYS trying to bounce the flash of something (was doing my best to

avoid the straight down the lens flash) the exposures were not always consistent. I guess this ties in with the

point aboive, but it was frustrating, since I didn't know if the flash was maxed out or not.

 

How do you know when you're maxing out your flash (other then using it manual mode, and dialing in your own

fraction of max power)?

 

Any words of advice?

A good point to start is to understand how the EOS system works with flashes. ...

 

http://photonotes.org/articles/eos-flash/

Some basic calculations on flash power: the GN at the 50mm setting is 40, so at f/8 and 100 ISO you have 40/8 = 5 metres range. When bouncing the distance is from the flash to the bounce surface and then to the subject, and flash power is also lost to absorption and scattering by the bounce surface - this can cost 1-2 stops over the lens setting. Every stop faster ISO adds 40% to the guide number (2 stops doubles it). In practice, many wedding reception rooms have ceilings that make bounce flash almost impossible without substantially boosting the ISO. Ways of dealing with this include mounting the flash on a bracket so that it is centred over the lens, but at a high enough angle such that shadows fall mostly behind the subject, and providing your own bounce surface with e.g. a Lumiquest Pocketbounce.

How do you know? Basically, by experience. You should know your guide numbers and all, but if you test and have a feel for what your flash can do, at what ISOs and apertures, in medium to large places of varying colors, you will know what you can get away with. I can tell you that bouncing off the ceiling is generally more consistent since the distance from the flash to the ceiling stays more consistent. Wall bounce is great if you have good surfaces fairly close, but as you found out, it isn't always possible while still using a reasonable ISO. Color and value of the walls and ceilings, even the floor, also affect what is possible with bounce.

 

Actually inconsistency of exposure from frame to frame is not entirely due to bouncing only. I don't find ETTL, or any other automatic flash metering, extremely consistent without riding compensation constantly (with digital, which will show even 1/3 stop difference).

Also--you should still be dragging the shutter to 'bring in' more ambient light.

Nadine,

I was dragging the shutter.

I kept doing test throughout the night (without the flash) to see what shutter speed would bring in a good amount of light, to illuminate the background.

But thinking back, I think my problem was mostly that I was too far away from the walls that I was trying to bounce from.

You also mentioned that you don't find the ETTL all that consistent. So what do you do?

Set the flash to manual and shoot that way?

I used the same combo a lot (mostly bouncing off ceilings mostly) and found that 1 to 1 1/2 stops of FEC gives pretty good exposures. Walls are trickier than ceilings because the distance from wall to subject varies a lot and because many rooms have white ceilings and colored walls.

 

Make sure your batteries are fully charged when you begin testing. When the batteries get weaker the flash takes longer to recharge and you'll get underexposure if it fires with insufficient charge.

Adrian--you mentioned using 1/100th, and in dark conditions, that isn't dragging the shutter too much, although it does depend on ISO used and the EV in play. I mention dragging the shutter because it helps a lot in these kind of situations. If you let the flash do all the illuminating, you are certainly going to have reach problems.

 

I actually use auto thyristor flash metering when indoors in darker conditions. I find it more logical to predict than ETTL. There are fewer compensation 'rules' to remember. However, it, too, isn't always ideal, particularly in very dark situations, like outdoors at night with no reflectivity from walls and ceiling. Some people have found manual flash workable for bounce situations. You might try it. I find it works OK if you accept a bit of variability in exposure which you optimize later in post.