Exposure and ETTL

[beaconlightboy]

Ok, i have what i believe is a simple question, but i'm sure there is no simple answer, or maybe there is.

This question is based on a few assumptions that i believe to be true thru my experience (the little i have) so please correct me if any are wrong. each assumption is based on the previous one.

assumptions:

When using a flash, the only thing that matters for exposure is the F-Stop. Shutter speed determines background exposure only.
The F-Stop is based upon the intensity of the light at a specific distance irregardless of the subjects skin tones, or outfit colors.
Reflective light is never brighter than incident light, hence the fact reflectivity can't affect a properly metered shot. (white dress for instance)Now if all those assumptions are in fact true. Why isnt flash exposure in the Canon, or any other camera for that matter based solely on distance of the subject. there technically would be no need for reading reflective light. you just need the distance. When i setup my stobe for wedding shots i take a meter reading and my shots are always properly exposed. While i have found a few cheats for getting good exposures with E-TTL.

i find it very unpredictable when mixing blacks and whites at weddings. you never know where your camera is going to be aimed when composing a shot, primarily when couples are moving so you never know whether to over or under the FE. i use FEL whenever possible. Anyways, can anyone simply tell me why the heck the camera doesnt just get the distance and use that for its exposure setting. Seems really simple to me.

PLEASE HELP ME UNDERSTAND!!1

[tim]

I don't understand your point #3, but it is 2.45am here, but I don't know if I agree with it. When ETTL meters it can only rely on reflected light, and different materials and colors reflect differently. If you metered your Speedlite on manual and kept your distance correct then the exposures should be great. Also remember flash heads zoom.

For weddings 90% of the time I use a hand held light meter to determine my exposure and use flash only as fill, camera on full manual of course. I've gotten much more accurate exposures and less post processing needed that way.

ETTL is a little unpredictable, aye.

[beaconlightboy]

I don't understand your point #3, but it is 2.45am here, but I don't know if I agree with it. When ETTL meters it can only rely on reflected light, and different materials and colors reflect differently.

Tim - i may have confused you with #3. when i said
"
Reflective light is never brighter than incident light, hence the fact reflectivity can't affect a properly metered shot. (white dress for instance)
Now if all those assumptions are in "

i meant that if the shot was metered incident. not reflective.

The point is that the camera doesnt need to meter the flash at all to properly expose a subject it just needs to know the f-stop, ISO, and the distance of the subject. I'm not an expert, but i thought that was the foundation of the Guide system. so many feet, certain ISO etc.

the only thing the light meter meters is light. light falls off at a certain distance, hence changing its ability to properly expose a subject. so knowing this formula one should be able to figure out how far to stand away from a subject and at what F-stop and ISO the camera should be set at for that light setting. if you can do this manually, then why can't the flash systems. it seems so simply complicated.

does this help confuse you more..lol

[beaconlightboy]

i'm either crazy, or canon's got a lot of engineers that can't change light bulbs.

[tim]

I think you're correct, with distance you can work out the power that needs to be put out. I wonder why they don't do that. What if you bounce the flash?

And yes, you've just confused me more!

[gmen]

Do a google search on 'Flashmatic system' (or Guide Number system)...

A number of cameras used this system: The flash is set to manual, the guide number of the flash, the ISO and the shutter speed are set on the camera... and the camera calculates the aperture required based on the subject distance.

It does make for excellent exposures that are not influenced by light/dark clothing etc.

The Flashmatic approach was often used on rangefinder models... the Olympus 35-SP springs to mind by way of example. The 'Flashmatic' feature was 'coupled' to the lens focusing mechanism.

I owned a 35mm P&S Contax T2 during the days of film (wazzat?) - it also used the Flashmatic approach for flash exposures and was consistently peachy.

I'm not 100% sure that this is what the OP is getting at... but it might be on the right lines.

EDIT: Here's a link to the Olympus 35-SP, scroll down for various details on the Flashmatic system: http://www.ph.utexas.edu/~yue/misc/35SP.html

---- Gavin

[beaconlightboy]

I will read up on this. but is any of this technology canon friendly?

[Wilt]

The point is that the camera doesnt need to meter the flash at all to properly expose a subject it just needs to know the f-stop, ISO, and the distance of the subject. I'm not an expert, but i thought that was the foundation of the Guide system. so many feet, certain ISO etc.

You are abolutely correct on this point!!! TTL was invented before it became more commonplace for lenses to tell the focus distance, however. So brightness of the scene was used. When digital sensors came out, however, TTL was rendered ineffective since the surface of the sensor was so much more reflective than film emulsion, so an alternative had to be engineered, like the much (and often deservedly!) maligned ETTL. If all lenses for Canon cameras had the distance information, your theory would be a nice easy way to deal with it (no different than flash tables on flash units in the days of Manual flash!). But then you also have no way to deal reasonably with the many other lenses you might mount (via adapters, etc) on that Canon body.

[TheSteveMadden]

You'd also have problems with multiple flash/off camera flash scenerios. You'd almost need a distance measuring device on the flash itself, and not use the lens data as it's the flash to subject distance which is relevant.

[beaconlightboy]

Thanks for the explanation wilt. I'm not sure i understand what the reflectiveness of the sensor has to do with it. could you elaborate.

[beaconlightboy]

You'd also have problems with multiple flash/off camera flash scenerios. You'd almost need a distance measuring device on the flash itself, and not use the lens data as it's the flash to subject distance which is relevant.

yes, but the flash can fire an infared beam. that could be used to determine distance. or some similiar technology. when using a portable light triggered by infrared, radio, or sync.. you're most likely to use a light meter anyway

[Wilt]

Thanks for the explanation wilt. I'm not sure i understand what the reflectiveness of the sensor has to do with it. could you elaborate.

Many TTL flash control were OTF (off the film). The photosensor read the light striking the surface of the film, then instructed the flash to stop output when enough light stuck the film surface. Apparently the shiny surface of the digital sensor makes it difficult to use the same technique. (You'd think they simply reduced the sensitivity of the photosensor, so it probably is more complex that that!)

[beaconlightboy]

Many TTL flash control were OTF (off the film). The photosensor read the light striking the surface of the film, then instructed the flash to stop output when enough light stuck the film surface.


with that being said. wouldn't that still rely soley on reflective metering. if we Metered distance alone this would be irrelavent,... yes?

[Wilt]

Well, yes the white bridal gown from the film would throw things off like white bridal gown seen by your camera's sensor the reads the focusing screen. So yes the distance thing is the best, and is no different that using GN and distance to determine f/stop, in concept. If you had a filter on the lens, the simple calculation would fail, though, if the filter absorbed any light, and it would fail if you used a teleconvertor. Also, GN is based upon room of average reflectivity, so in a large hall with walls far away and a high ceiling, your distance-derived f/stop would result in underexposure as well! So you see, it's unfortunately not a simple as you make it.

[beaconlightboy]

probably not, but it would be more reliable than E-TTL is, especially when subjects are of different colors and moving.

[Wilt]

probably not, but it would be more reliable than E-TTL is, especially when subjects are of different colors and moving.

I won't dispute the LACK of reliability of ETTL, but if you look at all the exceptions to the rule that I posted wouly afflict the purely distance-derived algorithm, to someone who is not familiar with all the exceptions it would seem just as unreliable!
I love the TTL systems from 10-20 years ago, and found them absolutely predictable...the only exception was if the SUBJECT brightness was not greycard level. And I do find ETTL systems too unpredictable. But I fault the lack of documentation (provided with the camera) from the engineering/marketing departments of Canon for why thousands of users have to resort to online photo forums to figure out some of what makes our ETTL/ETTL2 flash system tick! Perhaps it is due to the 'dumbing down' of the dSLR that causes them to deliberately NOT put details into the user manual so as not to scare them off, since the average Joe or Jane seems to be unwilling or unable to read user manuals with so many other things they buy?!