E-TTL II examples

[bill_goldman]

Some highlights on the features of the new EOS-1D Mark II can be

viewed at a Japanese Canon website (in English). One page purports

to show how the new E-TTL II improves flash pictures when the

averaging option is set for the camera (C.Fn 4-1).

 

There are three examples. The first shows a girl in a white blouse

holding a black dog with a light colored wall. The next shows a

girl with a lamp and night scene through a window in the

background. Finally, the third shows a back lighted subject with a

bright sky in the background.

 

It seems to me that the original E-TTL with the active focusing point

(s) on the girl�s face would have yielded the same exposures as the

E-TTL II averaged ones shown. Furthermore, if the girl in the first

picture was not holding the black dog, the E-TTL II example would

have been underexposed.

 

Anyhow the URL is:

 

http://web.canon.jp/Imaging/eos1dm2/html/09.html

 

 

Any comments?

[the_macman]

Is that the only body to have E-TTL II ?

[PuppyDigs]

"Is that the only body to have E-TTL II ?"

 

Nope. The Elan 7N/7NE with cool blue illuminated LCD also sports E-TTL II.

[mark u]

I agree that ordinary E-TTL should have got the third exposure right. I'm not sure what the first example would have done without the dog. With the picture as it is, the average of black and white is gray. If the dog was out of the picture then the girl's shirt would have had high reflectivity for the distance which the algorithm might be able to correct for properly - but it's something you'd have to test to know for certain how it performs. The same is true with a predominantly black subject.

 

I can think of many other situations which would require a learning process - e.g. how different is performance when flash is indirect, or when slaves are added. It's a bit like learning to live with any new evaluative metering algorithm that supposedly corrects for certain conditions that otherwise require compensation or spot/partial/incident metering.

[the_macman]

That's one of the crappiest comparisons I've ever seen... seems like canon.jp

is aiming that camera to clueless consumers (possibly weekend shooters

midwives), otherwise they wouldn't feed us with such a junk.

 

I don't question E-TTL and as I love their products I'd gladly believe them

when they say the E-TTL II is fantastic... but would it hurt them to tell us with

what the f* they're comparing it? Is the "bad" image E-TTL, no flash or a

randomnly set flash?

[bill_goldman]

One correction. The averaging option on the 1D Mark II is set with C.Fn 14-1, not 4-1.

[jon_austin]

I think if I wasn't able to get better flash photos in similar shooting conditions with my (non-E-TTL II) 10D and 420EX than those offered as "without E-TTL II" examples on the cited web page, I'd have returned my camera and flash to the vendor and looked for another solution. (And I'm no professional ... far from it.)

[steve_dunn2]

<p>I don't see anywhere on the page where it actually says that the "without E-TTL II" pictures are presumed to be the original E-TTL; "without E-TTL II" could mean E-TTL, A-TTL, TTL, or any other flash metering scheme. I think they're comparing to non-E-TTL, or perhaps to E-TTL as wielded by someone who doesn't understand how it works.</p>

 

<p>Certainly, picture #3 should be exposed properly by E-TTL as long as flash is metered off her face (either by putting the active focus point there, or using FEL). My limited experience suggests that #2 should also be metered correctly by E-TTL, and I think #1 would be, too - again, as long as the photographer has used an active focus point on the face for flash metering.</p>

 

<p>Keep in mind, too, that they're trying to make a point. They're not going to show two pictures with subtle differences in flash metering (which is what I found in most cases when I compared my old Elan II's pop-up TTL flash to the 380EX E-TTL flash); they're going to make an effort to give you glaring exposure errors with the non-E-TTL II pictures. Even if that means that they have to fake them. The page mentioned is a marketing page and should be treated as such.</p>

[beauh44]

Yeah, those examples were pretty lame. It's not as though you couldn't have gotten the same results without ETTL-II. Maybe I'm wrong, but if it's just an algorithm, then it's just some new code that Canon tweaked a little differently than the old code and made a decision to market it with the new "II" moniker. Oh well . . .

[mark_davidson]

The sad fact is that ETTL does not work very well and Canon knows that. It is also a sad

fact that Nikon's implementation of exposure calculation for digital imaging was equally

poor. I have several friends that shoot professionally that switched to Canon because of

the promise of ETTL. Though things got better, it was a far cry from the reliability they

had with an auto flash and the latitude of film.

 

I do believe that ETTL-2 will be a significant advance since DISTANCE to the subject is

actually taken into account. For crying out loud doesn't anyone in this world remember

how flash exposure used to be calculated? Doesn't anyone remember the 45GN Nikkor?

I suppose I'll have to shuffle off to the home for retired and nearly dead photographers to

find anyone who does. ;)

[maureen_m]

<i>" "without E-TTL II" could mean E-TTL, A-TTL, TTL, or any other flash metering scheme."</i><p>I figure that the "without E-TTL II" shots must have been made with regular E-TTL. They were taken with a 1D Mark II (it's that model's page after all), which doesn't support TTL or A-TTL (it's a DSLR), and if the shots were made with a manual flash, the blame would lie with whoever metered the shot.<p>I guess the "without E-TTL" shots could have been made with an "Auto" flash (Vivitar 283, etc), but not likely, as I assume they at least used a Canon flash.

[maureen_m]

Of course, I meant "without E-TTL <b>II</b>" in that last line.

[primoz]

This one looks for me as comparions of E-TTL II and flash set to whatever it feels and without E-TTL.<br>

If E-TTL would be so bad as it's on these photos then I have to be genious :) And since I don't really believe in this I'm pretty sure E-TTL doesn't have anything to do with sample photos marked as "without E-TTL II".<br>

I agree E-TTL II is probably better then E-TTL but even E-TTL is not bad in most of cases.

[jim_larson1]

This web page reminds me of the dRebel TV commercial: the one with the guy taking fast action telephoto's of photoball game with a 18-55 lens (you know, the type of situation that a 400/4.0L would struggle with).

 

My take on ETTL-II is that Canon listened to the -quite legitimate- complaints about ETTL not having an averaging function. FEL is fine, but a tool that is not suited to the fast moving photography that the 45pt AF systems are designed for. I believe averaging is a CHOICE on this flash, and as such I welcome ETTL-II as an advance for the system.

 

With E-TTL on my 10D, I would not have had any trouble with shots #2 and #3 => unless I failed to use FEL or failed to have the focus point on the girls face. I would argue that correctly used ETTL should yield SUPERIOR results in these cases, unless the ETTL-II was set to a "non-averaging" mode.

 

Shot #1, with a white subject against a white wall (with a black dog!) is real trouble. My experience would have had me dial in a -1 FEC for the first shot (and double check at the histogram afterwards). In this case, the averaging of the ETTL-II would save the day. I agree that if the dog was not there, or was white, averaging ETTL-II flash would probably also have had trouble with this shot.