E-TTL and softbox

It also depends on the metering mode. Evaluative has multiple metering zones. When you press the shutter half way the camera takes the ambient reading. When you press the shutter all the way the pre flash compares the ambient reading to the closest object (which is normally the subject/s) that reflected light back and isolates it. A bride in a white dress, a groom in a black tux and the bride and groom all reflect light back differently.

So much like the cameras ambient light meter which requires compensation for shooting black tar or snow you need to adjust FEC for what was reflected back during the pre flash.

Some photogs prefer to set the flash to average when shooting indoor venues which does not isolate the subject.

digital paradise wrote in post #16318807
It also depends on the metering mode. Evaluative has multiple metering zones. When you press the shutter half way the camera takes the ambient reading. When you press the shutter all the way the pre flash compares the ambient reading to the closest object (which is normally the subject/s) that reflected light back and isolates it. A bride in a white dress, a groom in a black tux and the bride and groom all reflect light back differently.

So much like the cameras ambient light meter which requires compensation for shooting black tar or snow you need to adjust FEC for what was reflected back during the pre flash.

Some photogs prefer to set the flash to average when shooting indoor venues which does not isolate the subject.

Certainly. So when you are shooting in those conditions where your subjects are changing (i.e. from a bride in a white dress, to a groom in a black tux)... how much EC/FEC is also having to be adjusted to compensate from shot to shot? If adjustment is necessary then I don't call it accuracy... I'd call it, getting you in the ballpark.

EDIT: I've heard this analogy a while back, but I'll summarize. ETTL is like the public transportation system which drops you off in close proximity to your house (i.e. ballpark). On occasion, the route and/or stops may change and it might even be right in front of your building... when you're really lucky (on occasion), it's in front of your own door. A light meter is like having a dedicated concierge service which gets you to the front door each and every time.

I have read so much on this and from personal experience I don't care if ETTL knows there is a modifier in front of it or not. I try to predict the scene as best I can but I know I'll be playing with FEC all evening. I've accepted it and it just makes life that much easier.

ETTL is subject to the same vagaries as the reflective metering in the camera. If your subject is more reflective than average (white) you will have underexposure. If your subject is less reflective (black) you will have overexposure. But I can assure you its a lot better than the old auto flash settings where the flash used a thyrister and a magic eye on the flash. A shiny bit of silverware on a table could play havoc with that.

ETTL 2 is supposed to compensate for highly reflective surfaces. I purchased a Metz fot it's legendary auto mode and it did not disappoint. Only thing was the tubes could not shut down and drain fast enough at high ISO so I kept getting over exposure warnings. That is one good thing you can say about pre flash and ETTL.

I believe I read here that ETTL has improved so much with the 600EX-RT and is a lot more accurate.

Is this accurate? I'm thinking about getting 3-4 600EXs for wedding and just keeping it in ETTL mode to make sure I focus more on capturing pictures than having to worry about constantly changing the power output of strobes (cheetah lights specifically that I am thinking about).

The last time I checked, it was the camera that does the metering of the ETTL system. Because of this, I'd be surprised if there was any difference whether using a 580EXII or a 600EX-RT.

That is correct.

The only thing they could have improved on is the Auto calibration because the flash decides on correct flash exposure. I have read it is not much better than the 580 II.

gonzogolf wrote in post #16318735
so you could have the same two subjects but two different exposures because of a minor change in position.

drvnbysound wrote in post #16318781
That is the lack of accuracy that I was referring to.

And that is precisely the greater accuracy I referred to. The light meter does not take that changed circumstance into account; E-TTL does. It may not be as convenient for those who have not learnt to work with it.

When I setup a studio shot, my light falls on the subject the same way for each and every shot - until I choose to change the light setup. It doesn't matter if the subject turns her head to one side or the other, or changes her pose all together; if she is in the same place, the light doesn't change, neither does the light falling on her. However, with ETTL if she was wearing a reflective earring, it could create a different exposure because of the way the light was reflected back into the camera. There is NO way that I want to do a test shot between each and every pose to see how ETTL decided to expose, and need to adjust FEC accordingly. That's not what I call accurate, but you can call it whatever you want

However, if I'm shooting an event, where I'm moving around and I'm shooting a bunch of different subjects in different places, etc. then ETTL is 'good enough'... and I'm not going to take the time to use a meter and get every exposure perfect.

drvnbysound wrote in post #16318831
Certainly. So when you are shooting in those conditions where your subjects are changing (i.e. from a bride in a white dress, to a groom in a black tux)... how much EC/FEC is also having to be adjusted to compensate from shot to shot? If adjustment is necessary then I don't call it accuracy... I'd call it, getting you in the ballpark.

EDIT: I've heard this analogy a while back, but I'll summarize. ETTL is like the public transportation system which drops you off in close proximity to your house (i.e. ballpark). On occasion, the route and/or stops may change and it might even be right in front of your building... when you're really lucky (on occasion), it's in front of your own door. A light meter is like having a dedicated concierge service which gets you to the front door each and every time.

As an analogy that's a poor one.
incident metering, which is what people generally are referring to when they are using a light meter measures the light that is emitted from the light source and is striking the subject. The single benefit is that if you know the amount of light striking the subject, and the subject's tonal range is within the dynamic range of the sensor a properly exposed image is the result. However, if the scene is beyond the dynamic range of the sensor the photographer has to make a judgement about what to sacrifice.

On the other hand, one can use a reflective meter which is conveniently built into the camera. It is calibrated, as is the incident meter, to 18% gray. Thus, the meter reading indicates the exposure considering that all the reflective illumination from the scene to the sensor is mid gray. If its not, then the photographer has to make a judgement about how to change the exposure, based on the meter reading, to place the key elements of the scene into the proper exposure zone.

That's the neat thing about meter's, once the photographer understands how they are measuring the light that will eventually excite the sensor, its relatively easy to use them to get a reliable reading. Then to use that information to set the camera to make a good exposure.

Perhaps a better analogy is that incident and reflective meters are like open end and box wrenches. Both will tighten or loosen a nut. Which to use depends on circumstance.

I'll have to do some homework and find the source of that analogy, but I'm pretty certain it was either Zack Arias or David Hobby - both above my pay grade in terms of photography.

Trying to find the source I did come across this from Zack, "Incident meters is reading the light falling on the subject regardless of whether the subject is white, black, green, grey, blue, dog, etc. You walk over to the subject, put the meter where the light is falling on it, take a reading. It says 5.6 at 60th of a second or whatever. You set your camera to 5.6 at 60th and boom! Balls on accurate. You don’t have to think about it."

Yes, so a reflective meter is trying to turn every picture I take into 18% gray. So when the model changes her top, the ETTL metering will probably change the amount of flash power. Sure, this may be accurate if you live in an 18% gray world; I don't. I live in the real world where things are of color.

When the light position doesn't change and the model's position doesn't change, the light output shouldn't change; inverse square law. Can a photographer use either meter to get a good exposure, yes - absolutely. I 100% agree that they are different tools, and that you can understand how to use one or the other to make proper exposures. But the original question pertains to accuracy... so can you reliably get your shot in a single shutter click with ETTL, or do you need to chip and make EC/FEC adjustments? If you require EC/FEC adjustments, then it's just getting you to the building... and you are compensating to get yourself to the door

CliveyBoy wrote in post #16319602
And that is precisely the greater accuracy I referred to. The light meter does not take that changed circumstance into account; E-TTL does. It may not be as convenient for those who have not learnt to work with it.

But I dont want ETTL to take that changed circumstance into account when I'm doing off camera studio work. I want the light power to be the same because the distances are the same, the exposure of the scene hasnt changed, only minor changes in the position come into play.

gonzogolf wrote in post #16321041
But I dont want ETTL to take that changed circumstance into account when I'm doing off camera studio work.

I agree. But isn't ETTL because it is too accurate, finnicky?