01-19-2013 03:21 PM
Canon stated that the ETTL II is designed to treat the exposure as a "plane" and not as "point" to correct the inconsistent exposures due to various reflections.
I have tried to vary the "evaluative" meter, the "average" meter, AV mode, Tv mode, Manual mode which doesn't seem to make a different in consistency. I noticed that some other cameras have spot meter for flash, and tend to have much brighter background than my 5DmkII at the same camera setting. I can try to use the flash memory lock which sometimes can not afford the added steps.
Thanks for your comments.
01-20-2013 10:51 AM
Both flashes are going to behave the same in ETTL mode. Both flashes are consistent with ETTL even though ETTL is best used with dynamic subjects. Both flashes are essentially the exact same except the 580 has a higher guide number (more power) and couple more features/settings.
To answer your questions, I do not believe you have any calibration issues with your lens or flash. Yes ETTL works the same on all modern Canon bodies. I truly believe your problem is improper flash exposure and/or flash coverage.
I do not know what Canon means by "plane" and "point." I'm guessing they're referring to evaluative-average, as opposed to a specific point (spot) metering to determine the correct amount of flash
We really need more information on your shooting conditions. Flash and camera model, lens and focal length, subject to camera distance, flash and camera exposure settings. Can you post some sample images?
Here are a few things that came to my mind:
1. Is is possible the front row subjects are shadowing the 2nd row subjects?
2. Is your flash mounted on camera or off? Does the flash need to be higher to illuminate the second row?
3. Does the flash zoom focal length match the focal length of the lens?
4. Do you have C.Fn-09 enabled so the auto zoom matches the sensor size?
5. Is there movement or changing brightness with the subjects
6. Are you using any FEC or FEB?
7. What shooting mode are you using (AV, TV, manual, green box)
Generally speaking you want to use ETTL with dynamic subjects where the brightness changes and the subject to camera distance changes. Then use FEC to increase or decrease flash output.
For static subjects (like what you appear to be shooting), use manual flash and increase or decrease the power to change the flash output.
Some things to try
1. Make sure your flash coverage matches the focal length of your lens.
2. Make sure your flash is positioned to fully cover your subjects. If this is a large group shot, you may need a second flash
3. Shoot some test shots with manual flash using different power settings
4. You can also try some ETTL test shots using FEC
You might even want to reset all your camera and flash settings first before you start testing.
Good luck, keep us posted
01-20-2013 09:05 PM - edited 01-20-2013 10:25 PM
I did eliminate these possible faults which you have mentioned above. Looking at the reviews of the 580exII from the CanonUsa.com, I saw at least 3 pros who have the same exact problem with the 580exII.
Photographers used OTF TTL (off-the-film through-the-lens) flash reliably more than 30 years ago, one couldn't imagine how ETTL could be that much more difficult.
I'm not totally sold on the 580exII as the culprit since it can handle direct on camera mount shots OK. It's the bounce shots (ceiling, bounce card, etc) that push the ETTL II to be all over the place.
Canon said that it only takes 1 battery to operate below optimal level to cause the flash to be faulty. Detecting the bad battery is not as easy as one thinks? maybe that's why we don't even have a battery check on these flash? Incidentally even Boeing 787 has problems with its lithium battery couple days ago. It's things like this that make photographers jump ship.
01-20-2013 10:28 PM
It makes sense that's the way physics is.
I know the 580exII has enough power because it can be manually over-compensate to blow out the side curtain or vase to get the center subject the correct exposure although this is what ETTL II and "Evaluative meter" suppose to do.
The other thing is the other brand camera next to me doesn't have to obey this Inverse Square Law, and its background is much brighter. If it's the CPU sensitivity, sure hope the 6D and 5DmkIII have caught on.
In the mean time, back to Lightroom and Photoshop to tweek each photos one by one.
01-21-2013 01:02 PM
Inverse Square Law does not answer everything or solve all problems. It is just another tool. I believe that it is just an important for a strobist to understand it as undestanding how your shutter, aperture and ISO work together (without flash).
All light obeys physics. Just like there is no mircale flash diffiser out there that can bend light. There is something else going on there.
01-21-2013 05:08 PM - edited 01-22-2013 11:53 PM
I read your first post again.
First bullit. Are the subjects are in the second row, seated of standing. You said inches apart. If the subjects in the front are correctly exposed then ETTL is doing it's job. If you want to expose the back subjects correctly then the ones in the front would be overexposed. It can't redistribute light equally so this is where fall off is important. It also relies heavily on the pre flash along with the distance information. Although flash has nothing to do with the cameras meter a bride in a white dress alone, a groom in a black tux alone and bride and groom together will all reflect light back diffferently during the pre flash and will be affected just like when your cameras meter is metering ambient light.
Second bullit. Well technically yes. If your main subjects are properly exposed then it is doing it's job. Evaluative metering on the flash tries to isloate the closest subject to you because that is typically the case. It is just doing what it was programmed to do. Average just tries to average the whole scene and I did not find had any advantage over Evaluative.
Third Bullet. No. A flash is flash and light follows the same rules of physics for all camera/flash combinations.
Don't bother with AV mode. Not used by many with flash. The difference between M and AV modes is as follows. I call AV mode meter dependant. I made that up. That means the camera meter has to be on center. Since you are bringing pretty much most of the ambient light you need the flash is now a fill. Canon has set it up that way and will only provide fill flash. You have to be careful in a dark room with low ISO as it will bring your shutter speeds as low as it needs to maintain that light meter based on the aperture you select.
M Mode is meter independant. Again I made that up. You meter will fall wherever based on the available light. In a dark room it will more than likely show underexposed. Now your flash becomes the main source of light, again exposing for the main subject and may leave the rest of the exposure dark.
A couple of things. To get natural looking flash your need to maintain a balance between your flash and ambient light. In dark venues you need to increase your ISO and slow the shuter down. Even though flash is controlled by aperture opening it also helps. If you can try to maintain about 2 stops underexposed that is oretty good. I often shoot 3200 and have no issue at 6400 if I have to. All depends on your camera. This will illuminate the subjects in the back a bit more and prevent your main subjects from loooking like they were nuked. Modern high ISO capable cameras have opened new doors in this area.
This is the part you probably won't like. ETTL is not an exact sciene. The scene and changing conditions will always make us chimp and adjust the FEC as required. I call this run and gun shooting and I am continually adjusting the FEC as the conditions change. That is what it is. If I know the condtions will not change (studio style, ect) then I will go to manual flash which I prefer.
So there so nothing you are doing that is techncally wrong. Remember in a dark room if you are not brining in much ambient light your flash will do it's job to correctly expose your main subject. It does not cares what is behind it. That is your cameras job.
01-21-2013 05:15 PM
Here is a great place to learn about flash. My guru. Start in the blue column on the right - flash photography techniques.
Book is good too.
01-23-2013 03:20 PM
If you bounce the head of the flash or use a Master / Slave off-camera set-up then the camera cannot use the distance information.
Sy the lens reports the subject of 20 feet away, the camera can use that information if the flash is on the hotshoe (or a dedicated cable like the Canon OC-E3 or OCF-33). If the flash head is bounced or twisted then the camera has no way to know how far the light travels from the flash to the subject. If you have 30 foot high ceilings then the flash travels 30 foot up, 30 foot down to light your subject 20 feet away. Distance info is not valid in this case.
Same if you use master / slave off-camera flash since there can be different distance between flash and subject and camera and subject.
Brian / p4pictures.
01-23-2013 11:56 PM
Exactly. And now the system is relying on the light from the pre flash only to determine distance to the subject. Couple that with bouncing. You know you have to add FEC before your first shot.