How reliable do you find Canon ETTL

Messages
4,234
Edit My Images
No
Just wondering how reliable you find ettl. I find that to get the best results I have to use FEL (flash exposure lock - or * button). If I don't it's hit miss.

Love to hear your thoughts - be put straight!

Cheers

Dav
 
I love using Ettl, especially wireless Ettl. I don't tend to use FEL though. Indoors it works flawlessly, I only have issues outdoors, firing the slaves can be a pita sometimes.

I've finally got hold of a decent 10 mtr Ettl cable (all the way from California :eek:) which I'm positively bursting to try out properly, which should sort these issues.

You don't say how you're using Ettl though, it has its uses and then again sometimes its much better to use manual flash. I find the best time to use Ettl is when subject to flash distance can change frequently, it really comes into its own when shooting kids, but if I'm posing an adult subject its sometimes better to use manual.

Check out Syl Arena - Canons answer (sort of) to David Hobby. He looks odd but knows his stuff when it comes to Canon's flash systems.
 
Cheers fella. I'm referring to on camera. All my off camera stuff is done manually.

Cheers.

Dav
 
I find it a bit inconsistent too, and I often need to add +1/3 EV or so when ceiling bouncing. It's a little annoying when shooting something fast paced (like my little girl running around) and don't have the time to do things more manually or make an adjustment and re-shoot. But generally I cut Canon a bit of slack here - it's a complicated set of calculations that the cameras performing very quickly, so I don't expect perfect results every time, I just wish it was a tiny bit better than it is. Metering is one of my top 3 improvements for a 5D mk3. And I think (no direct experience) that this is one area Nikon beats Canon.

Edit: I should add that the majority of the time the ETTL nails the exposure (or near enough not to matter), so I just talking about the minority of shots. I'm being a bit demanding expecting this minority to be smaller than it actually is, but I paid a lot for the kit, and I'm not expecting perfection, so I don't think it's totally unrealistic.
 
Last edited:
Cheers fella. I'm referring to on camera. All my off camera stuff is done manually.

Cheers.

Dav

Try it off camera - that's where it really works well. Read the manual, scratch your head a bit :thinking:. Read Syl's website and understand it...

The thing about manual OCF is you have to set your power (well doh) and get your stand in the right place. Fire a test shot. Adjust. Test shot, hmmm not quite there...adjust...Perfect! (Unless your subject moved a bit...)

With Wireless Ettl, you fire a shot, Hmmm, needs a tad more. Adjust using the camera menu, fire again..perfect, even though subject has moved. Works a treat. This is why I love it.

On camera Ettl is ok, but wireless is where its at.
 
I'm sure wireless ocf is great - but I'm specifically talking about on camera.

I shot two weddings this weekend. The other tog (main) shot Nikon. His hit rate was without doubt much better than mine

Now it could be that I'm doing something silly, but I've never found canon ettl consistent.

Cheers.

Dav
 
I'm sure wireless ocf is great - but I'm specifically talking about on camera.

I shot two weddings this weekend. The other tog (main) shot Nikon. His hit rate was without doubt much better than mine

Now it could be that I'm doing something silly, but I've never found canon ettl consistent.

Cheers.

Dav

I know that I was just saying...:D

To be honest, I've never had a problem with it on camera, bouncing. I set it and forget it. I also use it with my ring flash adapter, needs a +2EV usually but again its consistent.
 
Just wondering how reliable you find ettl. I find that to get the best results I have to use FEL (flash exposure lock - or * button). If I don't it's hit miss.

Love to hear your thoughts - be put straight!

Cheers

Dav

E-TTL has a mind of its own, and unfortunately, Canon doesn't publish details of it's undoubtedly complex exposure algorithms.

One of the main influencers, when fired direct, is it uses focusing distance data, which appears to be overriding at close-ish range, and also AF point selection if that's on auto select. Pressing the FEL button appears to disable distance data, as does tilting the head very slightly (or just pressing the tilt release button).

Try this, first with flash fired direct at typical portrait distance, then drape a tissue over the flash - it will be underexposed because the flash believes the distance information (which would normally be very accurate) over the pre-flash data. Press the tilt release button and correct exposure is restored, or press the FEL botton and it will probably change slightly again.

I use E-TTL a lot and think it's brilliant. Sometimes it's not perfect straight off the bat, but it's very easy to control and adjust with compensation. I don't worry about 'accuracy' too much because with something like fill-in flash there is usually no such thing as a consistently accurate flash level. It varies a lot from a tiny dash to quite a major wallop, depending on the subject, the circumstances, and the kind of result you're after. I usually find the auto result pretty acceptable, but if not, tweak it a bit and then it stays how I want it.

Metering pattern selection can change things, and the ambient light level and dynamic range when using fill-in also influences the outcome. Conclusion - it's complcated :thinking: :D
 
Thinking about this a bit more, I suppose it depends on what mode the camera is in? I normally use M to get maximum creative control, so aperture, shutter, and ISO are locked down my me, and the instensity of the flash will ramp up and down to exposure the scene correctly. Essentially, I'm at the mercy of ETTL here, but when it works I get the shot I want.

What if I used AP mode instead? The camera now has to determine the shutter AND the flash levels itself. I'm not sure if this would lead to more or less consistency???

Maybe someone who has good success can chime in with what camera mode they usually use?

Damned ETTL :bang:
 
What if I used AP mode instead? The camera now has to determine the shutter AND the flash levels itself. I'm not sure if this would lead to more or less consistency???
I believe Canons in Av/Tv mode meter flash differently than for M and P modes. Essentially, in Av/Tv mode, they are doing fill-in flash...
 
I believe Canons in Av/Tv mode meter flash differently than for M and P modes. Essentially, in Av/Tv mode, they are doing fill-in flash...

Grrrr. I always thought it does fill when the flash head is pointed straight forward, but as soon as you tilt it upwards it changes.

Right, that's it, I going to go back a read properly, from cover to cover, this amazing fountain of ETTL knowledge that I never got round to fully reading before. I always promised myself I'd re-visit this, and maybe its time now, even if it is a very long read.
 
Grrrr. I always thought it does fill when the flash head is pointed straight forward, but as soon as you tilt it upwards it changes.

Right, that's it, I going to go back a read properly, from cover to cover, this amazing fountain of ETTL knowledge that I never got round to fully reading before. I always promised myself I'd re-visit this, and maybe its time now, even if it is a very long read.
Yup. That's a great resource. Take a look at: http://photonotes.org/articles/eos-flash/#faq10 to see how Canons meter in the different modes
 
Of all the many variables that Canon has thrown into E-TTL, I don't think the exposure mode is one of them - not as such. For example, if you shoot on Av the camera will naturally adjust the shutter speed to match the ambient, say 1/30sec at f/5.6, but if you make those same settings in manual, you should get the same result.

However, if you decide you want the shot to be 100% flash and set the shutter speed at max x-sync in manual, that will change how the camera interprets the scene and will adjust the flash output, as it should do because the ambient element has been removed - maybe not by much, but it's certainly a potential factor.

I use both Av and manual. Av is brilliant in a fluid social situation because it adjusts for distance, adjusts for ceiling height if you're bouncing, and the shutter speed will float to balance the changing ambient.

With a more static situation, when you know what you want and need it to stay there, switch to manual on both the camera and flash and everything is locked down. You know it won't change, whereas if you're in any auto mode slight movements of the subject can alter the way the system reads the scene and variations are possible.

In a situation like that, there is no advanatge to shooting on auto. Use auto when things are changing and you'll get usable exposure straight out of the can, a quick tweak and they'll be perfect. Basically use E-TTL and auto modes with flash when you just haven't got time to mess about.

NB, I sometimes use manual on camera and E-TTL on the flash - keeps the ambient background constant but adjusts the flash according to distance etc. Lots of permuations ;)
 
Grrrr. I always thought it does fill when the flash head is pointed straight forward, but as soon as you tilt it upwards it changes.

Right, that's it, I going to go back a read properly, from cover to cover, this amazing fountain of ETTL knowledge that I never got round to fully reading before. I always promised myself I'd re-visit this, and maybe its time now, even if it is a very long read.

Yes, that's a good resource, but he also bemoans the fact that Canon doesn't publish anything about how E-TTL works in detail. I also have a hunch that it varies slightly between different cameras, in addition to stuff like whether or not the lens reports focusing distance. Also see his very good book Mastering Canon EOS Flash Photography.

An even better book IMHO, not quite so technical and better on the practical technique stuff, is Syl Arena's Speedliter's Handbook. They're both about £20 on Amazon.
 
Yes, that's a good resource, but he also bemoans the fact that Canon doesn't publish anything about how E-TTL works in detail. I also have a hunch that it varies slightly between different cameras, in addition to stuff like whether or not the lens reports focusing distance. Also see his very good book Mastering Canon EOS Flash Photography.

An even better book IMHO, not quite so technical and better on the practical technique stuff, is Syl Arena's Speedliter's Handbook. They're both about £20 on Amazon.

Awesome, thanks for the tips.

One last question (and I admit to having not looked for myself), but how has the invention, and evolution of auto ISO influenced ETTL? I imagine that with different implementations between 7D, 5Dii, 5Di etc then the ETTL algorithms will be different?
 
Awesome, thanks for the tips.

One last question (and I admit to having not looked for myself), but how has the invention, and evolution of auto ISO influenced ETTL? I imagine that with different implementations between 7D, 5Dii, 5Di etc then the ETTL algorithms will be different?

I'm not sure it has, not directly.

The starting point for flash exposure is always what is set on the camera, either by the user or by the camera itself. The gun then fires a pre-flash, after that has been decided, and sets the power output according to those settings.

For example, if you are boucning in a very large hall and the gun has run out of power, it's up to you do raise the ISO (or drop the f/number) to sort things out.
 
Guys - thanks for all the input. Really appreciated.

I think over the weekend, I may run some experiments to get this sorted in my head.

I have to say when shooting the recption at a wedding on Monday, I found the best bet was to stick the flash in manual (as Richard suggests). It worked a treat.

Cheers all.

Dav
 
I don't think it's been mentioned, but also consider that ETTL exposure is affected by very pale and very dark subjects in just the same way that ambient exposure metering is. Fill the frame with the bride and you are likely to see the flash underexpose. Fill the frame with the groom and the reverse is likely.

Canon flash does seem to favour underexposure as a default, which can catch the unwary out until they understand the quirks. My first experience with a 580EX was at a friend's wedding and I was appalled at the results produced. Shooting in Av mode at the time, and suffering NEVEC did nothing to help either. At the time I'd only had the camera for three months and the flash for a week and I was expecting the camera to deliver great results, but I had no time to experiment before the wedding and on the day I was shocked by what happened. In short I did not know my equipment. I was expecting "normal" performance but everything with flash was underexposed. Everything!

My preferred approach these days for on camera flash in run and gun mode is to shoot with manual exposure for ambient and the flash on ETTL, but nudging it up to +1/3 FEC as my default starting point. You may prefer the results from the flash if you switch to average flash metering instead of evaluative flash. Evaluative supposedly works better for daylight/outdoor fill. Average may work better for indoor/dim conditions where the flash is your primary light source. TBH I haven't really tried out the options. Certainly I like to take manual control of the ambient exposure and that is one less thing for the camera to screw up for me.
 
Last edited:
Never really worked well unless I use FEL. On ETTL it's just not consistent enough to rely on.
I ditched speedlites in favour of Qflash, but on the odd occasion I still use them I use them on manual, or FEL with +2/3 dialled in on the flash.

ETTL2 is an improvement on the previous version (that was unusable), but it's still way off Nikons .
 
Back
Top