A Few TTL Flash Questions/Statements

[jasonlprice]

I’ve read through the Flash 101 and EOS flash bible and would like to see if I have the theory straight……

If I have not so bright available light (say F4 1/30 iso 800) then in P mode the 580 will choose either fill or main light depending on the calculated EV. In the case above EV is kinda low (maybe 10 or so?) so I’m thinking the flash would be the main light?

In Tv mode I would set the shutter speed to say 1/60, the aperture would be chosen by the camera to properly expose the background (f2.8) and flash exposure would always be fill. (Unless the maximum aperture of the lens is reached in which case the 580 would have to become the main light and expose foreground). With a fast lens “dragging” the shutter would be difficult.

In Av mode I would choose an aperture say f8 and the camera would choose a shutter speed of 1/8 @ iso 800 (get out the tripod) to expose for ambient light and the 580 would be fill only. This mode would be very useful outside with lots of available light or indoors with tripod. This mode would be useful to “drag the shutter” in low light situations to flash-freeze the subject.

In M mode if I set the camera to F4 1/30 iso 800 to expose for the ambient light correctly the flash will expose the subject and would be very reduced. The flash will always try to be the main light source for the subject? How is the subject determined? By distance if the lens can send that information? If the lens doesn't send that information then by manually zooming the 580 based on distance I read off the lens distance scale? If EV is low and I choose a fast shutter speed and/or closed aperture (say F8 1/60) under exposing the camera then the ambient light will be destroyed by the 580 trying to expose the foreground.

I have cf 4 set to 1 so I can’t lock flash exposure; or does half pressing the shutter also switch to FEL with a dedicated flash attached?. Do I really need to switch back to cf.4 0 when using the 580 (xt is not easy for me to change focus points so I f-recompose a lot).

Am I understanding Curt’s Flash Distance Chart correctly that (580ex set with the 580’s gns @ iso 100 in the chart) If I’m at iso 800 and shooting at 105 mm (and the flash zoomed) any aperture greater than 22 (with a subject distance of 25 feet) the flash will be capable of providing “enough” light for what it deems a proper exposure? If subject distance is 200 feet and I’m at iso 800 shooting at 105 mm I better have at least f 2.8 or bigger?

Thanks.

[Curtis N]

Wow. That's a lot to digest. ;)
I'll try to generalize in the hopes of covering some of it.

In Av and Tv modes, the camera will attempt to expose properly with ambient light. This may not be possible in Tv mode if the max. aperture of the lens is not enough, given the light level and your chosen shutter speed and ISO. And yes, Av mode will tend to generate slow shutter speeds in dark environments.

E-TTL is designed to properly expose the subject. It will account for the ambient exposure when calculating how much power is required to do that. Don't get caught in the "main" vs. "fill" terminology. It will either be the primary or secondary source of light, depending on the ambient exposure and how you set the FEC. Regardless of camera mode, if the ambient exposure is insignificant then flash will be the primary light source. If the ambient exposure is significant then flash will be only a secondary light source.

How does E-TTL determine what the subject is? Good question. It somehow guesses, and it often guesses wrong.

With custom function 4-1, you can lock the ambient exposure by half-pressing the shutter button. You do not have flash exposure lock (FEL). You can use flash with C.Fn. 4-1, you just can't use FEL.

Flash distance-
At ISO 800, multiply the guide number by 2.8
So your 580EX will have a GN of 162 meters or 536 feet.
536/25=21.4
So yes, your flash will have enough power to properly expose a subject 25 feet away at any aperture wider than f/22.

536/200=2.7
So you will need an aperture wider than f/2.8 for that distance.

[scottbergerphoto]

This explains how ETTLII determines where the subject is:
http://photography-on-the.net/forum/showthread.php?t=46311&highlight=ettlii

[jasonlprice]

Thanks Curtis! I feel much better about my understanding of ttl now (and your flash distance chart). It is a lot to digest and I've been reading, practicing, re-reading and re-practicing the flash bible and your 101. I just wanted some confirmation that after months of practice and studying my understanding was clear on some of the main points of ttl fash. After your response I believe I actually have a pretty good understanding.

I know I can still use c.fn 4 - 1 and flash, that is what I've been doing. I should have worded the question more like this....

Do the benefits of c-fn 4-1 (for me it is wonderful to have focus seperate) outweigh the benefits of c-fn 4 - 0 and FEL considering that I do recompose alot? What is your c-fn 4 set to when useing e-ttl flash versus non-flash? It is very much a habit for me now to focus with * so I'm afraid I'd miss shots if I went back to 4-0. I can chimp and FEC so for now I'm sticking with 4-1, but should I start practicing with flash and FEL?

[scottbergerphoto]

I always use CFn. 4-1. ETTLII determines where the subject is regardless of the active AF point.

[Curtis N]

Personally, I don't use FEL much. On those rare occasions when I feel the need I switch C.Fn. 4 back to 0, but mostly it stays at 1.

It's just a personal preference thing.

[jasonlprice]

Thanks Scott and Curtis. After reading the link you provided Scott (a couple of times!), I see that ettl 2 seems (in my small mind anyway) to render FEL useless (speaking purely about ttl flash photography). I'm not going to worry about going back to 4-0; as Curtis says it is my personal preference to use 4-1...

I'm still struggling with the term "subject" as it is used by Chuck Westfall when explaining the new e-ttl algorithm. I'll paste his explanation below just so it is right at our fingertips. I don’t really see why with the new algorithm there even needs to be a “subject” involved. The algorithm uses all metering zones and weights the ambient and preflash exposures of each zone and then averages those “scores”. Lets say with my example above that I’ve set the camera to F4 1/30 iso 800 which will properly expose the picture with available light. The pre-flash is fired and the EVs are very close so flash will be secondary light. Bright objects (highly reflective) will have a higher flash exposure value (the difference will be greater between ambient and flash) and will be weighted accordingly by the algorithm (or completely dropped out of the calculation if it is a highly reflective object). If I stop down to say f11 and keep the 1/30 and the algorithm calculates a greater difference (lets say below 10 ev) between the two exposures the flash becomes the main light. So what does the photographer’s subject have to do with that?
__________________________________________________ __________________
Copied from
http://photography-on-the.net/forum/showthread.php?t=46311&highlight=ettlii

By Chuck Westfall
Here's what we published in the EOS-1D Mark II White Paper document:

ADVANCED E-TTL II
New algorithm gives greater flash exposure control

In the new algorithm, ambient light is measured when the shutter button is pressed. Next, a pre-flash is fired and the metering sensor takes readings at the central 17 metering zones. The ambient and pre-flash readings are compared. The metering areas having a small difference are selected as the flash exposure metering areas. (Areas with very big differences between ambient and pre-flash readings are excluded or down weighted because they are assumed to contain a highly reflective object or that the subject is not in that part of the frame. The algorithm avoids chronic underexposure problems in such situations.) These readings are weighted, averaged and compared with the ambient light reading, and the main flash output is then set and stored in memory. E-TTL II weights and averages the flash metering for the subject and all other objects at the same distance as the subject. Even if the subject’s position, reflectance or size changes, the flash output will not change radically. The flash exposure will be highly accurate and stable.

[Curtis N]

Jason,

I'm not sure I understand your question, but I still think you're getting hung up on the "main vs. fill" concept. If you're shooting someone under a shade tree on a sunny day, the camera will expose the scene which is mostly bright background and the subject would be underexposed if only ambient light is used. The flash will add light to the subject, as much as necessary to get proper exposure. It makes no difference whether you call it main or fill. What's important is that the sum total of light on the subject from all sources is correct.

But my personal take on E-TTL has to do with the bullsnot in this excerpt:E-TTL II weights and averages the flash metering for the subject and all other objects at the same distance as the subject. Even if the subject’s position, reflectance or size changes, the flash output will not change radically. The flash exposure will be highly accurate and stable.Both common sense and personal experience tell me that this is a load of crap. A gray object 10 feet away might reflect the same light from the flash as a white object 15 feet away. There is no way for E-TTL to figure out the difference, which I believe is why this system is anything but "accurate and stable".

In my opinion, E-TTL is a great system for outdoor fill, because the background is typically so far in the distance that it won't reflect any measurable light from the flash, and hence the system can correctly identify the subject. But indoors, the background is typically close enough to reflect some light from the flash and confuse E-TTL into thinking that it's part of the subject.

[jasonlprice]

Curtis, the only question (and also my original question) I still have is how does e-ttl know what your "subject" is? Chuck's post talks about "subject" but doesn't answer how "subject" is determined, although it does say how non-subject items can sometimes be eliminated. You have already answered that question above, that it doesn't :D, so I'm not expecting any other answer or response if you want this to die;) .


How does E-TTL determine what the subject is?
Curtis N
It somehow guesses, and it often guesses wrong.

[Wilt]

But my personal take on E-TTL has to do with the bullsnot in this excerpt:
Quote:
E-TTL II weights and averages the flash metering for the subject and all other objects at the same distance as the subject. Even if the subject’s position, reflectance or size changes, the flash output will not change radically. The flash exposure will be highly accurate and stable.

Both common sense and personal experience tell me that this is a load of crap. A gray object 10 feet away might reflect the same light from the flash as a white object 15 feet away. There is no way for E-TTL to figure out the difference, which I believe is why this system is anything but "accurate and stable".


Curtis, there are two key points of Chuck Westfall's info on ETTL-II...
1. The camera looks at brightess values before and after the preflash, and ignore the zones that change radically in brightness...that is, the zones that change a lot are reflective surfaces like mirrors, so they are downweighted.
2. The camera looks at zones which are at the same focus distance as the point of focus, and considers them with increased importance in the calculation of flash output.

[Curtis N]

Well my original answer had a bit of sarcasm in it. But if you want to know how E-TTL attempts to determine the subject, I suspect it starts with this part:Next, a pre-flash is fired and the metering sensor takes readings at the central 17 metering zones. The ambient and pre-flash readings are compared.So it first makes the assumption that your subject is not out at the edge of the frame somewhere (if the shot is composed with your subject way at the edge, this might be a good time to use FEL). When it compares the ambient and pre-flash readings at those 17 zones, if there is no difference (or very little difference) in a particular zone, it might assume that that part of the frame is background, with nothing to reflect the flash back through the lens. If there is a significant difference between the ambient and pre-flash readings at a particular zone, it would assume that something is there reflecting the flash, and this must be part of the subject.

Make sense?

[Curtis N]

The camera looks at brightess values before and after the preflash, and ignore the zones that change radically in brightness...that is, the zones that change a lot are reflective surfaces like mirrors, so they are downweighted.I understand that part, but it contradicts personal experience. Make the mistake of shooting in front of a window and the shot will be underexposed every time.The camera looks at zones which are at the same focus distance as the point of focus, and considers them with increased importance in the calculation of flash output.A clever stragegy, frought with potential for error. But I'll get off my soapbox now. ;)

Thanks for pointing out those details, Wilt.

[Wilt]

I understand that part, but it contradicts personal experience. Make the mistake of shooting in front of a window and the shot will be underexposed every time.A clever stragegy, frought with potential for error. But I'll get off my soapbox now. ;)

Well the open window in the scene...was that at night, or was it during the day? and were you where you could see the reflection of the flash in the window? The theory of the situations...

a. If you could see the reflection of the flash at night, the meter should have ignored that zone because of the big change in brightness, and so a proper exposure results.

b. If you could not see the reflection of the flash in the day, the daylight probably fooled the ETTL system into thinking "...not much change in brightness, so I won't ignore that zone", but it also thought, "ahh, I have enought brightness, time to tell the flash to shut off" (which was too soon!)

c. If you could not see the reflection of the flash at night, the meter was thinking "...not much change in brightness, so I won't ignore that zone", but it also thought, "ahh, I have need to tell the flash to output more light until I see there is enough light"

Are you seeing cases which don't follow that logic?

I'm a critic of ETTL, too, having come to depend upon TTL in film cameras. But I'm trying to learn to live with the oddball behaviors that I can get my arms around. But, like you, the oddball things frustrate me to no end.

[jasonlprice]

Curtis N
If there is a significant difference between the ambient and pre-flash readings at a particular zone, it would assume that something is there reflecting the flash, and this must be part of the subject.

Ahhhh now I see and at some point the subject gets completely ignored when the difference reaches some critical threshold. Thanks Curtis and Wilt, I now at least understand the theory.

I'll do some window/mirror etc testing of this and see what I get :lol: . Please keep up the debate, it is very interesting to me.

[Curtis N]

I was referring to indoor shots where the reflection of the flash in the window can be seen in the shot. But that's an ugly distraction so I normally delete those anyway.

If I figure out a systematic way to demonstrate the frustrating quirks of E-TTL, I'll take some sample shots and post them.

[scottbergerphoto]

I believe one of the reasons for the change to ETTLII was to help get correct exposures in situations where there are wide differences in reflectivity, such a bride and/or groom. The Bride with a white dress and the groom in a black tuxedo. ETTLII is supposed to help reduce the flash systems overreaction to the black and the white.