E-TTL II

[canon_user808]

Does any one know how E-TTL II work and will it work with my EX550? I currently have the DRebel and will almost certainly be getting the 20D once it hit the market (My local dealer had me down as #1 on his waiting list). Or, will I have to get the new EX580 that is being released the same time as the 20D to make the most of the added performance built into the camera?

[timmyquest]

Go look at some 1d MKII reviews, that should give you a pretty good indication.

The 550EX works, and works well.

[canon_user808]

Thanks!

[neil_r]

550 and MkII works fine so I guess the answer is yes

N

[Cadwell]

If I remember when the news about E-TTL II first broke, it is only fully compatible with certain lenses


Fixed Focal Length

EF14mm f/2.8L USM
EF20mm f/2.8 USM
EF24mm f/1.4L USM
EF28mm f/1.8 USM
EF35mm f/1.4L USM
MP-E65mm f/2.8 1-5x Macrophoto
EF85mm f/1.8 USM
EF100mm f/2 USM
EF100mm f/2.8 Macro USM
EF100mm f/2.8 Macro*
EF135mm f/2L USM
EF180mm f/3.5L Macro USM
EF200mm f/2.8L II USM
EF200mm f/2.8L USM*
EF300mm f/2.8L IS USM
EF300mm f/4L IS USM
EF300mm f/4L USM*
EF400mm f/2.8L IS USM
EF400mm f/4 DO IS USM
EF400mm f/5.6L USM
EF500mm f/4L IS USM
EF600mm f/4L IS USM
EF1200mm f/5.6L USM

Zoom Lenses

EF16-35mm f/2.8L USM
EF17-35mm f/2.8L USM*
EF17-40mm f/4L USM
EF20-35mm f/3.5-4.5 USM
EF24-70mm f/2.8L USM
EF24-85mm f/3.5-4.5 USM
EF28-70mm f/2.8L USM*
EF28-80mm f/3.5-5.6 USM*
EF28-105mm f/3.5-4.5 II USM
EF28-105mm f/3.5-4.5 USM*
EF28-105mm f/4-5.6 USM
EF28-105mm f/4-5.6
EF28-200mm f/3.5-5.6 USM
EF28-200mm f/3.5-5.6*
EF28-300mm f/3.5-5.6L IS USM
EF35-135mm f/4-5.6 USM*
EF70-200mm f/2.8L IS USM
EF70-200mm f/2.8L USM
EF70-200mm f/4L USM
EF70-210mm f/3.5-4.5 USM*
EF70-300mm f/4.5-5.6 DO IS USM
EF90-300mm f/4.5-5.6 USM
EF90-300mm f/4.5-5.6
EF100-300mm f/4.5-5.6 USM
EF100-400mm f/4.5-5.6L IS USM

EF-S18-55mm f/3.5-5.6 USM
EF-S18-55mm f/3.5-5.6

*Indicates discontinued product


There are some notable popular lenses missing from that list, including the EF 28-135mm IS, which I know a lot of people here have.

[scottbergerphoto]

ETTLII is a function of the camera body not the flash. It will work on all EX series flashes. The camera body uses distance information from the lens and reflected light from the preflash off the subject to determine flash output. You can choose via CFn.s to have the reflected light evaluated in two ways by the camera body 1. Evaluative or 2. Averageing. ETTL2 differs from ETTL not only in terms of using distance information, but in that it is totally independent of active focus points. Switching from AF to MF has no effect on the flash output as it does in ETTL. In Evaluative ETTL2,(CFn. 14-0), the camera looks at the relative amounts of light reflected back off the subject at 17 metering segments within the AF elipse. It selects the metering segments to use for calculating flash output based on relative intensity differences. It ignores data from those metering segments reporting wide variations from the rest. It was designed to solve the problem of the white gown and black tux. In Averaging ETTL2,(CFn. 14-1), the camera uses all 17 metering segment's data and averages it.
So far only two camera bodies have ETTL2: the 1D Mark II and the new 20D.
If you use a lens that doesn't support distance information, the camera will just work without the distance information. It never reverts to ETTL.
Scott

[JABACo]

If you use a lens that doesn't support distance information, the camera will just work without the distance information. It never reverts to ETTL Scott

I wonder about 3rd party lenses. Such as the Sigma's??????

BA

[scottbergerphoto]

If you use a lens that doesn't support distance information, the camera will just work without the distance information. It never reverts to ETTL Scott

I wonder about 3rd party lenses. Such as the Sigma's??????

BA
If the flash supports ETTL it should support ETTL2. The main thing the flash does in both systems is to fire a preflash. The camera body does the rest. It is how the camera body uses the pre flash information that sets ETTL and ETTL2 apart. I use a Q Flash T2D which was designed for ETTl. It works fine in ETTL2.
Scott

[J Rabin]

Scott and other users:
I have trouble nailing fill-flash exposure (indoor and outdoor) with 10D E-TTL and 420 EX using Lumiquest Pocket Bouncer. I imagine E-TTL II will not improve this, it's likely due to my lack of camera handling skill.
Especially with 17-40Lmm on the 10D I underexpose, or poorly expose (specially around white clothing or white tablecothes in viewfinder).
Does a person need to use a 550 EX (or=) and manually boost flash output power setting for the loss? What's the proper approach with Pocket Bouncers?
I have considered using the partial meter FEL Lock on skin, but for candid shots people are irritated getting hit by the pre-flash, or they blink and the exposure results in closed eyes, and I'm not convinced this will aid, since the flash output is still reduced due to spreading the light.

Any ideas on improving my success?
I use the Lumiquest to avoid hot sports, shadows, get a nice even spread, uniform skin illumination color/tone. It's wonderful outdoors. But, when the flash head is tilted up in bounce position, the flesh head zoom power output feature is disabled.
J

[scottbergerphoto]

You have to make adjustments to flash metering just as you do for ambient light metering. The problem you are having with whites is because the white reflects alot of the preflash back to the camera, telling it to emit less light from the flash. Your camera meter is programed to make white look grey. You need to add exposure to white for ambient and flash metering. One of the things ETTL2 was supposed to improve on is exactly this. ETTL2 in Evaluative mode, discards information from the metering segments giving very high or very low readings. It is a big improvement over ETTL.
One of the nice things about digital is that you can make adjustments and immediately see the results. Practice with different subjects and settings and see what works.
Regards,
Scott

[J Rabin]

Thank you Scott.
So, it is not so much the reduction in light power from the bouncer, or the distance (so E-TTL II won't help there)... it is a metering problem. ie, my lack of camera handling skill.
I suppose this is the place for EC/FEC? I should set FEC to +2/3 or even +1 or even > when walking around with the Lumiquest bouncer for candid shots with light colored subject matter. There is no time to meter and re-meter. The moment is gone. I'll experiment with it.
Thanks again.
J

[Jesper]

I don't want to push you into buying the 580 EX, but....: :P

The new 580 EX will have some interesting advantages over the 550 EX, especially on the 20D and 1D Mark II. According to Canon (http://web.canon.jp/Imaging/eos20d/580EX/index.html):

When the SPEEDLITE 580EX receives image size data from a supported EOS digital camera*, the flash head automatically zooms to match the effective angle of view. The resulting flash coverage is more accurate. When color temperature data is automatically transmitted from the SPEEDLITE 580EX to a camera*, the camera sets the color balance for the optimal flash shot.
* EOS 20D, EOS-1D Mark II (firmware update necessary) or later camera models

[scottbergerphoto]

Thank you Scott.
So, it is not so much the reduction in light power from the bouncer, or the distance (so E-TTL II won't help there)... it is a metering problem. ie, my lack of camera handling skill.
I suppose this is the place for EC/FEC? I should set FEC to +2/3 or even +1 or even > when walking around with the Lumiquest bouncer for candid shots with light colored subject matter. There is no time to meter and re-meter. The moment is gone. I'll experiment with it.
Thanks again.
J
Correct. Light colored objects need more ambient and flash exposure. ETTL2 in Evaluative Mode tries to do it for you.

Keep in mind however that each flash has it's limits. The Guide number of the flash tells you how far a direct on flash will carry at ISO 100 at F 1.0. To calculate the maximum distance your flash will travel at ISO 100 for a given f stop:
Distance = Guide number/f stop
for ISO 200 multiply by 1.4
for ISO 400 multiply by 2
The diffuser will cost you some distance. Check the papers that came with your diffuser or the manufacturers web site to see how many f stops you lose. So if your diffuser costs you 2 stops and you are using f/5.6, then you would use f/11 in the distance formula. You need to stay within the maximum distance for adequate flash power.
Regards,
Scott

[J Rabin]

Scott et al.
More on my E-TTL learning curve with 10D. I found following re: E-TTL II on P.K. Frary's web site at: http://emedia.leeward.hawaii.edu/frary/canon_elan7ne.htm
I will try using M metering suggested.

“Does E-TTL II insure direct flash exposures are more accurate than before? Maybe, but E-TTL on the Elan 7E performed extremely well, so I haven't noticed a significant flash improvement with the Elan 7NE, not even with slide film. E-TTL II results with EOS 1D MKII are said to be more striking, probably because digital sensors are more sensitive to small exposure fluctuations than film.

Although Canon improved E-TTL, they left one serious flaw to keep photographers thinking and spinning dials: Flash Negative Evaluative Exposure Compensation or NEVEC. If the flash is on in M, Av or Tv modes, the camera applies negative 0.5 to 1.0 exposure compensation to the background under dim lighting, e.g., Ev 8 at ISO 100. Thus, flash and ambient light are not balanced because the background is underexposed. Some like this look, but I hate it! If you're in a hurry, use Av or Tv mode and counter NEVEC with plus 0.5 to 1.0 exposure compensation (guesstimate). However, the most accurate method is to meter ambient and flash separately:

1. With flash off, meter ambient light in M mode.
2. Turn flash on and take picture. Be sure the active sensor is on the subject and don't adjust exposure no matter what the meter says.

If you're considering a flash for the Elan 7NE, the Canon 420EX Speedlite is a perfectly matched companion. It's the only flash that covers all seven of the Elan 7NE's AF sensors with AF assist. Incidentally, E-TTL II circuits are onboard camera, so all EX series Speedlites are fully compatible with E-TTL II's new features.”
NEVEC is discussed at:
http://eosdoc.com/manuals.asp?q=NEVEC

[Oblomov]

You can choose via CFn.s to have the reflected light evaluated in two ways by the camera body 1. Evaluative or 2. Averageing.
Scott
What is the best choice?

[scottbergerphoto]

You can choose via CFn.s to have the reflected light evaluated in two ways by the camera body 1. Evaluative or 2. Averageing.
Scott
What is the best choice?
That seems to be a matter of personal choice. Some people are happier with one or the other. Some feel that Averageing is more consistent. I haven't taken enough flash pictures to prefer one or the other yet. The test I did and posted didn't reveal any major differences:
http://www.pbase.com/scottbergerphoto/1d_mark_ii_flash_test
Major differences might occur when taking pictures of white and black objects together. Cfn 14-0 should be better for that.
Regards,
Scott

[Oblomov]

Thanks Scott, also for the test.

[DaveG]

I don't want to push you into buying the 580 EX, but....: :P

The new 580 EX will have some interesting advantages over the 550 EX, especially on the 20D and 1D Mark II. According to Canon (http://web.canon.jp/Imaging/eos20d/580EX/index.html):

When the SPEEDLITE 580EX receives image size data from a supported EOS digital camera*, the flash head automatically zooms to match the effective angle of view. The resulting flash coverage is more accurate. When color temperature data is automatically transmitted from the SPEEDLITE 580EX to a camera*, the camera sets the color balance for the optimal flash shot.
* EOS 20D, EOS-1D Mark II (firmware update necessary) or later camera models

I read that and realized that it's neither of these features is all that important, at least to me.

The size of the cone of light that the flash throws is important only if it's too narrow, and with a 550 on a 1.6 camera the cone will be too big. It will weaken the flash output but the 550 has lots of horsepower and will be fine. It will have to pump out a little more light per shot, and this may impact on battery drain but only a little I would think.

I also wonder about the flash coverage when the flash head is in a vertical position in a bracket while I'm shooting. I often use this configuration with my 10D and a 16-35. When the flash is in this position there is more up and down coverage than left and right and with a really wide lens I think that I'm counting on the sloppiness of the 550's light cone to cover. I doubt if the flash knows that the flash head is now up and down rather than left to right, so would the 580 cover?

Finally the colour balance is only important if you are shooting jpegs. For us RAW shooters it's moot.

[Jon]

The size of the cone of light that the flash throws is important only if it's too narrow, and with a 550 on a 1.6 camera the cone will be too big. It will weaken the flash output but the 550 has lots of horsepower and will be fine. It will have to pump out a little more light per shot, and this may impact on battery drain but only a little I would think.

I also wonder about the flash coverage when the flash head is in a vertical position in a bracket while I'm shooting. I often use this configuration with my 10D and a 16-35. When the flash is in this position there is more up and down coverage than left and right and with a really wide lens I think that I'm counting on the sloppiness of the 550's light cone to cover. I doubt if the flash knows that the flash head is now up and down rather than left to right, so would the 580 cover?

The 550 in bounce mode, automatically zooms to 50 mm (p. 90). I'd expect the 580 to behave similarly. You can override if you want to, but it doesn't pay any attention to the lens you're using.

[DaveG]

The size of the cone of light that the flash throws is important only if it's too narrow, and with a 550 on a 1.6 camera the cone will be too big. It will weaken the flash output but the 550 has lots of horsepower and will be fine. It will have to pump out a little more light per shot, and this may impact on battery drain but only a little I would think.

I also wonder about the flash coverage when the flash head is in a vertical position in a bracket while I'm shooting. I often use this configuration with my 10D and a 16-35. When the flash is in this position there is more up and down coverage than left and right and with a really wide lens I think that I'm counting on the sloppiness of the 550's light cone to cover. I doubt if the flash knows that the flash head is now up and down rather than left to right, so would the 580 cover?



The 550 in bounce mode, automatically zooms to 50 mm (p. 90). I'd expect the 580 to behave similarly. You can override if you want to, but it doesn't pay any attention to the lens you're using.

True, but what if you don't want to (or can't) bounce?

I often find myself in crowded rooms doing "social shots" of cocktail parties for corporate clients. There's a 10 meter ceiling so only a fool would try to bounce. I want to use my 16-35 so I don't have to knock people down to get the shot. My flash is in a flash flip bracket so when I shoot vertically the flash head needs to be rotated to point forward.

At this point the head is vertical and the flash will produce a lot of top/bottom coverage, but reduced left/right coverage. This is the situation where I wouldn't want the 580 to say, "Hey he's using a 16mm focal length but it's on a 10D so I'll narrow the beam.", and then I'll get fall off on the edges. This is where with the "extra" coverage from the 550 make things OK.

[Andy_T]

This is the situation where I wouldn't want the 580 to say, "Hey he's using a 16mm focal length but it's on a 10D so I'll narrow the beam.", and then I'll get fall off on the edges. This is where with the "extra" coverage from the 550 make things OK.

But wouldn't the flash recognize that it is on the 1DII which has a different sensor size and needs a wider flash?

IMHO that would not be clever engineering, if the flash didn't take this into consideration.
Or is it a flash that is exclusively targeted on the 1.6 crop factor cameras (an EF-S flash :lol: ), and the 1D and 1Ds people are supposed to stick to the 550?

Best regards,
Andy

[DaveG]

This is the situation where I wouldn't want the 580 to say, "Hey he's using a 16mm focal length but it's on a 10D so I'll narrow the beam.", and then I'll get fall off on the edges. This is where with the "extra" coverage from the 550 make things OK.

But wouldn't the flash recognize that it is on the 1DII which has a different sensor size and needs a wider flash?

IMHO that would not be clever engineering, if the flash didn't take this into consideration.
Or is it a flash that is exclusively targeted on the 1.6 crop factor cameras (an EF-S flash :lol: ), and the 1D and 1Ds people are supposed to stick to the 550?

Best regards,
Andy
No, even the 10D needs a wider flash cone when the flash head is turned sideways and the 580 may not account for that. Imagine the camera sitting on a desk with a flash in the hot shoe. The flash's head is horizontal in this position and will throw more light left to right than up and down, and that's the way it's supposed to work and I'll call this position A.

Turn the camera into a vertical position. Hold the flash above it but side ways. It makes sense now that the you'll get more flash spread top and bottom, rather than left and right. This is the position my flash is in when it's in the flash bracket. If I use this set-up with a very wide angle lens I could have the flash decide to narrow the beam, since its point of reference is always "Position A". That's where the extra "slop" from the 550 comes in handy.

[Andy_T]

No, even the 10D needs a wider flash cone when the flash head is turned sideways and the 580 may not account for that.

That's quite logical.

My mistake :roll:

Had I read your post as thorough as I should have, I would have gotten that you put the flash sideways. Duh.

Most likely the flash realizes if you tilt it, but not when it's in a bracket. So they should include a 'manual mode' for that, as most pros will very likely use flash brackets.

Best regards,
Andy

[Jon]

I, too, read "when the flash head is in a vertical position" to mean the head was tilted 90 deg. Thanks for clarifying. Yes in that situation, you're right - Canon needs to develop a 590EX which has an orientation sensor coupled to the camera's orientation sensor to handle these situations. Or we need to set things manually to allow for that.

[Andy_T]

Canon needs to develop a 590EX which has an orientation sensor coupled to the camera's orientation sensor to handle these situations.

Yes. Everything that's worth doing, is also worth overdoing :lol:

Best regards,
Andy