How Canon 430 II and 580 II work.

So when you go through the Canon Master Flash series the videos explain it. ETTL uses an ambient reading, a flash reading (pre fire) and a combined flash and ambient reading to determine flash output. If available depending on lens it also uses distance information to confirm distance information.

Much like using a guide in manual it will determine the output power based on distance to subject. This would explain why if you put on lets say a Stofen (not sure why you would want to - just an example) the image will be underexposed. ETTL does not know there is a diffuser on so it will still output the same power based on distance to subject. You need to add FEC.

Based on another thread in another world I read - if you tilt the flash 45 degrees up ETTL stops using lens distance information and you will get a more accurate exposure. Also if your lens does not provide distance information and flash is in the forward orientation same thing. This does not make sense to me because the videos state that ETTL uses an ambient reading, a flash reading (pre fire) and a combined flash and ambient reading to determine flash output. Lens distance information just provides extra distance confirmation. I think?

I also read that this would apply with no diffuser and bouncing. Again this does make sense to me. If I bounce my flash and it has to travel 16 feet and my subject is 8 feet away how can it determine the correct distance?

The thing is the system knows the flash has been tilted so how does it compensate? Or does it? Is there something else in the algorithms going on? Is it looking for whites or basing it on 18% grey and actually determine correct exposure as well, not just distance?

I know I have to FEC so it is no big deal. I just like know how things work. Often you can use that to your advantage.

It reads the preflash that bounces off the subject and back through the lens to a sensor inside the camera. So as long as the distance is not so great to max out on power the reading taken at the sensor controls the output. Keep in mind that ETTL is evaluative Through The Lens. This technology replace the old thyristor systems where the flash was reflected off the subject to a sensor on the front of the flash and the auto circuit shut off the flash when enough light returned to meet a certain fixed exposure amount. The problem with that system was that a white table cloth nearby but not in the frame would reflect much more light than a black tuxedo in the subject. By moving the sensor inside the camera you bring the camera's metering system based on what it see's through the lens into play making it more accurate. Still subject to the vagaries of making the world 18% gray, but better than before.

So it is just not working on distance alone but also determining correct exposure also. You figure it could not be just distance alone. Direct flash has been pretty accurate. They only time it really struggles a little I found is a fashion show. When the models are on one move.

Interesting about the thyristor. Seems like a lot of people like Metz but it did not go so well for Canon in the II series flashes.

Yeah its not purely based on distance, and because it uses the camera's metering system I am a bit skeptical of any claim that a metz flash would be perform any better on the same body than a canon flash in terms of exposure calculation.

Edit for clarity, I meant a metz flash using ettl.

digital paradise wrote in post #12522309
Much like using a guide in manual it will determine the output power based on distance to subject. This would explain why if you put on lets say a Stofen (not sure why you would want to - just an example) the image will be underexposed. ETTL does not know there is a diffuser on so it will still output the same power based on distance to subject. You need to add FEC.

To me, this statement stands out as being false or misunderstood.

While distance information is sometimes used(not all lenses have distance information capability), the main variable is the ETTL preflash. The Stofen diffuser should not result in a change to the exposure because the ETTL preflash will compensate for any light modifier.

gonzogolf wrote in post #12522356
So as long as the distance is not so great to max out on power the reading taken at the sensor controls the output.

That's the way I understand it too. The ETTL preflash isn't full-power. The camera interpolates the preflash reading and adjusts the flash power accordingly.

If the stofen is causing under-exposure for the OP with ETTL, I suspect that it's blocking the flash from being able to provide full light output.

msowsun wrote in post #12523186
the ETTL preflash will compensate for any light modifier.

unless the modifier prevents the flash from delivering full-power

gonzogolf wrote in post #12522517
Yeah its not purely based on distance, and because it uses the camera's metering system I am a bit skeptical of any claim that a metz flash would be perform any better on the same body than a canon flash in terms of exposure calculation.

Metz flash Auto mode (photosensor in flash) is very accurate, unlike the absurdly underexposing Canon implementation of External mode in the 580EXII.

Because the body is what tells the flash to output (after emitting the preflash which is read by the body's sensor), ETTL is equally flawed (or on target) in accuracy regardless of whose ETTL-compatible flash you are using.


As for use of a device like Stoffen, in Manual flash mode it certainly is reducing light, causing ultimate underexposure.
In ETTL mode the Stoffen-like devices (with the flash firing forward, not upward bounce) might UNDERexpose or they might OVERexposure even in ETTL. PaceAce and I conducted tests with two Wescott 5x7" softboxes on Canon and on Metz flashes, and we experience UNDERexposure with some bodies and lenses, but OVERexposure with other bodies and lenses. We tried to determine the variable which affected UNDER vs. OVER, but we have not arrived at any definitive conclusion about this, merely puzzlement.
Aiming the flash upward without any modifier on the flash does result in good exposure. Why the same thing does not occur with the flash firing foward with a modifier is a puzzlement caused by the ETTL programming 'wizards' at Canon and the lack of good information from Canon about that automation and its design characteristics.

msowsun wrote in post #12523186
To me, this statement stands out as being false or misunderstood.

While distance information is sometimes used(not all lenses have distance information capability), the main variable is the ETTL preflash. The Stofen diffuser should not result in a change to the exposure because the ETTL preflash will compensate for any light modifier.

This is just based what I read based on the flash firing in the forward position. The person did the tests and had the same repeated results. When the stofen was installed it underexposed. I'm just curious if you tested this? I haven't but I'd be curious to know. I'm interested about how it works. Canon has little literature on this.

digital paradise wrote in post #12525011
This is just based what I read based on the flash firing in the forward position. The person did the tests and had the same repeated results. When the stofen was installed it underexposed. I'm just curious if you tested this? I haven't but I'd be curious to know. I'm interested about how it works. Canon has little literature on this.

Maybe the flash ran out of power rather than intentionally underexposing the scene? The stofen should make no difference otherwise.

Wilt wrote in post #12523384
Aiming the flash upward without any modifier on the flash does result in good exposure. Why the same thing does not occur with the flash firing foward with a modifier is a puzzlement caused by the ETTL programming 'wizards' at Canon and the lack of good information from Canon about that automation and its design characteristics.

rudy_216 wrote in post #12525072
Maybe the flash ran out of power rather than intentionally underexposing the scene? The stofen should make no difference otherwise.

I don't know. You are saying it should and I Wilt is saying it doesn't. I guess I'll have to try it to know for sure. Just curious if anyone anyone tried this before? I'm not saying I'm right, I'm just trying to figure this out. If it does not make a difference all the better.

I don't use my cheapo ebay "Stofen" type diffuser very much and I always bounce. I just tried it out with direct flash and noticed different results depending on whether or not I used Evaluative or Average ETTL metering. I usually use Average and found that in all my photos the diffuser gave a slightly brighter photo than without. I suppose it was because the light was now being bounced all around the room.

On a hunch I switched to Evaluative I got very different results. Almost every photo with the diffuser was noticeably darker. (as described above)

It doesn't make sense to me but there is something weird going on.

Wilt I just had a thought on this. Perhaps the preflash and distance information are contradicting each other.

Scenario 1. No diffuser on the flash and pointed forward. Flash prefires and determines that the correct flash output is 1/2 power. The lens distance information confirms it.

Scenario 2. Diffuser is on the flash and pointed forward. Flash prefires and due to the diffuser it determines the correct flash output power is 3/4. The lens distance information is telling the system it needs 1/2 power and chooses the lower output power to prevent overexposure. I doubt Canon had diffusers in mind when they designed it.

You tilt the flash up, the lens distance information is disabled and a good exposure. I am going to try this out just to see for myself.

Here is a bit from Bob Atkins.

http://www.bobatkins.c​om/photography/eosfaq/​ettl2.html

msowsun wrote in post #12525365
I don't use my cheapo ebay "Stofen" type diffuser very much and I always bounce. I just tried it out with direct flash and noticed different results depending on whether or not I used Evaluative or Average ETTL metering. I usually use Average and found that in all my photos the diffuser gave a slightly brighter photo than without. I suppose it was because the light was now being bounced all around the room.

On a hunch I switched to Evaluative I got very different results. Almost every photo with the diffuser was noticeably darker. (as described above)

It doesn't make sense to me but there is something weird going on.

That is interesting. Different metering modes were not discussed.

Here are some examples: All shot with a 7D and 15-85mm. ISO400 Manual mode 1/160 f/5.6 and direct flash 580EX II ETTL mode "0" FEC

Average metering NO diffuser

I must have bumped the wheel because the rest are at f/8


average without