Canon 300D FEL

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

Andy - 31 Jul 2005 20:34 GMT

Hi,

I've been playing around with ways to diffuse the built-in flash on my
300D and wondered about the following.

AIUI the pre-flash is supposed to calculate the flash exposure and fires
again for the actual shot but with I'm not seeing this happening. With a
simple diffuser in front of the flash the shots are a good stop under
exposed.

Yet, if I use the Flash Exposure Lock prior to the shot I get a good
exposure every time. Why is this different from the pre-flash, or have I
mis-understood the pre-flash?

Signature

Andy

Reply to this Message

Colin D - 01 Aug 2005 01:48 GMT

> Hi,
>
[quoted text clipped - 12 lines]
> --
> Andy

I have the same camera. If you just frame and shoot, the exposure is
centre-weighted around the AF-point - one of the seven available - and
it may not be on an item you want correctly exposed. I have my AF point
selected as the centre point only, and try to place it on a key tone in
the subject, or at least I try to get the framing to exclude bright
highlights that might depress the exposure.

With FEL, the metering switches to partial, a 9% area in the centre of
the VF, and when you push the FEL button the pre-flash fires and sets
the subsequent exposure for whatever part of the subject the centre AF
point was looking at. You then have 16 seconds to frame and shoot the
subject before the exposure information is cancelled.

Here are a couple of relevant websites you may find useful:

http://photonotes.org/articles/eos-flash/index.html
http://revolution.cx/rcx/fecset.htm

Colin D.

Reply to this Message

Andy - 01 Aug 2005 06:48 GMT

> I have the same camera. If you just frame and shoot, the exposure is
> centre-weighted around the AF-point - one of the seven available - and
> it may not be on an item you want correctly exposed. I have my AF point
> selected as the centre point only, and try to place it on a key tone in
> the subject, or at least I try to get the framing to exclude bright
> highlights that might depress the exposure.

I'm using the centre point also however a simple test shows the
difference I'm talking about.

Find something to act as a diffuser ( a piece of thin white paper, paper
towel, etc) and rig it over the flash.

Take 2 program mode shots of a plain, evenly lit wall - one with and the
other without FEL.

I'm seeing a clear difference in exposure between the two. Whatever
metering mode is used I would have expected the results to be the same??

Signature

Andy

Reply to this Message

Colin D - 01 Aug 2005 12:03 GMT

> > I have the same camera. If you just frame and shoot, the exposure is
> > centre-weighted around the AF-point - one of the seven available - and
[quoted text clipped - 17 lines]
> --
> Andy

Agreed there. I'll have a shot at that and post what happens. One
thought strikes; did you do this test with the camera at your eye, or on
a tripod? If on a tripod, did you blank off the viewfinder? Without
your face in the way, stray light will enter the eyepiece and cause
underexposure. Perhaps your hand was in the way when pushing the FEL
button and the preflash fired, but out of the way when pushing the
shutter, thus altering the amount of light entering the VF? IIRC the
300d came with a rubber plug to cover the eyepiece for just this
purpose.

Colin D.

Reply to this Message

Andy - 01 Aug 2005 12:33 GMT

> Agreed there. I'll have a shot at that and post what happens. One
> thought strikes; did you do this test with the camera at your eye, or on

Test were done with the camera to my eye along the lines of:

Half press, AE/AF Lock, shot
Half press, AE/AF Lock, FEL, shot

I didn't take the camera away from my eye between shots.

I'll be interested to hear what results you get, thanks.

Signature

Andy

Reply to this Message

Colin D - 02 Aug 2005 00:24 GMT

> > Agreed there. I'll have a shot at that and post what happens. One
> > thought strikes; did you do this test with the camera at your eye, or on
[quoted text clipped - 10 lines]
> --
> Andy

Ok, here's my results. Standing about 8 feet from an off-white wall,
420EX flash on camera;

shot 1: direct flash, evaluative metering, no paper;
shot 2: direct flash, FEL metering, no paper.

shot 3: bounce flash from ceiling and wall behind, evaluative metering,
no paper;
shot 4: bounce flash, FEL metering, no paper

shot 5: direct flash, evaluative metering, with paper towel;
shot 6: direct flash, FEL metering, with paper towel.

Results: shots 1, 2, 3, 4, and 6: practically no exposure variation.
Histograms were very close. Shot 5, evaluative with paper towel, was
underexposed by a half-stop or more.

Conclusion: there appeared to be no difference between the evaluative
and the FEL shots, (which use partial metering, the central 9% of the
image) - with the exception of shot 5, which was underexposed. This was
the underexposed shot with evaluative metering and paper over the flash,
which leads me to think that the paper towel was upsetting the light
distribution, causing the evaluative metering to shorten the exposure.

Colin D.

Reply to this Message

Andy - 02 Aug 2005 06:49 GMT

> the underexposed shot with evaluative metering and paper over the flash,
> which leads me to think that the paper towel was upsetting the light
> distribution, causing the evaluative metering to shorten the exposure.

Thanks for that Colin, the results are what I was seeing. I wonder if
there is something in the E-TTL formula that assumes a certain
power:light ratio for direct flash?

Your other post re the test pictures make perfect sense, I think maybe
my brain is too used to thinking about light in the simplified world of
ray tracing ;(

At least I know that if necessary I can diffuse the built-in flash and
get good results from FEL if I take a reading from a sensible part of
the frame.

Signature

Thanks again,

Andy

Reply to this Message

Vince - 01 Aug 2005 13:33 GMT

Can i see some test images. let me know i will give my email address.
thanks.

Vince....

Colin D wrote:

Signature

Andy

Reply to this Message

Andy - 01 Aug 2005 15:09 GMT

> Can i see some test images. let me know i will give my email address.
> thanks.

Examples are here:

http://www.pxl8.co.uk/without_FEL.jpg
http://www.pxl8.co.uk/with_FEL.jpg

in this test FEL produced a darker exposure than without, but surely,
given the content, both exposures should be almost identical?

Signature

Andy

Reply to this Message

Vince - 02 Aug 2005 03:28 GMT

Thank you Andy

Vince ...

Vince wrote:

> Can i see some test images. let me know i will give my email address.
> thanks.

Signature

Andy

Reply to this Message

Colin D - 02 Aug 2005 04:20 GMT

> Thank you Andy
>
[quoted text clipped - 14 lines]
> --
> Andy

Looking at those shots, there seems to be some sort of sheen on the
wallpaper in the centre of the shot, which is returning more light than
the periphery. Since 'without FEL' uses evaluative metering, i.e. from
the whole image area, and FEL metering uses 9% of the area in the
centre, the FEL is going to see a brighter area than the evaluative
average, therefore I would expect the FEL exposure to be darker, and it
is. I don't think there is a problem there.

Colin D.

Reply to this Message