Confusion about TTL metering modes

I'm trying to understand my TTL work flow and I'm wondering how most people are metering their subjects. I have no problems setting ambient exposure.....lots of tools here ...... manual settings, use histogram, evaluative metering, chimping. Next, on to exposing my subject..... I have read that some people simply stay on spot metering mode, meter a face, as an example, and let TTL properly expose a face. As I continue to understand this, I see on my 5D3 there are only two different types of TTL metering: Evaluative, and Average, whatever that is. That's on Pg 191 of the 5D3 manual. If that's the case the subject should be exposed the same regardless of whether one is in spot, evaluative, or whatever metering mode. That's not the case as I see different results when I change metering modes. I know TTL has mysteries but I need to better understand the relationships between TTL metering modes, camera metering modes, and what tools I have available. I'd sure like to hear how folks use their exposure meter modes to determine ambient, then subject exposures when using TTL. Thank you

mpstan wrote in post #18224468
I'm trying to understand my TTL work flow and I'm wondering how most people are metering their subjects. I have no problems setting ambient exposure.....lots of tools here ...... manual settings, use histogram, evaluative metering, chimping. Next, on to exposing my subject..... I have read that some people simply stay on spot metering mode, meter a face, as an example, and let TTL properly expose a face. As I continue to understand this, I see on my 5D3 there are only two different types of TTL metering: Evaluative, and Average, whatever that is. That's on Pg 191 of the 5D3 manual. If that's the case the subject should be exposed the same regardless of whether one is in spot, evaluative, or whatever metering mode. That's not the case as I see different results when I change metering modes. I know TTL has mysteries but I need to better understand the relationships between TTL metering modes, camera metering modes, and what tools I have available. I'd sure like to hear how folks use their exposure meter modes to determine ambient, then subject exposures when using TTL. Thank you

Hi, something here may help you.
https://www.youtube.co​m …=ttl+portrait+p​hotography
Russ

mpstan wrote in post #18224468
As I continue to understand this, I see on my 5D3 there are only two different types of TTL metering: Evaluative, and Average, whatever that is. That's on Pg 191 of the 5D3 manual. If that's the case the subject should be exposed the same regardless of whether one is in spot, evaluative, or whatever metering mode. That's not the case as I see different results when I change metering modes. I know TTL has mysteries but I need to better understand the relationships between TTL metering modes, camera metering modes, and what tools I have available. I'd sure like to hear how folks use their exposure meter modes to determine ambient, then subject exposures when using TTL.

Although the 5DIII uses the same sensors for ambient and flash, as you point out, you can setup the camera to use those sensors independently for ambient and ETTL II.

See:
http://cpn.canon-europe.com/content/edu​cation/infobank.do

there are plenty of pages just on flash to work through. (The rubric on the Mandarin Duck photo is incorrect.) And:

http://www.learn.usa.c​anon.com …ite_tip_pt2_art​icle.shtml

I set my camera to M because the system has no idea how I wish the ambient lit parts of the frame to look. If using ETTL II, I then use Average for flash or FEL (forces a central spot pre-flash & lock). I expect to set FEC in both cases depending on the tone & reflectivity of the pre-flash metered area.

There are two types of metering modes. TTL and ETTL. TLL is for film cameras, ETTL is for digital.

First off to get flash you need separate the two. Even if the flash is on the camera pretend the flash and camera do not know the other exists. When you half press the shutter your camera light meter and metering modes go into action just like it would if you did not have a flash on. However metering off your subject is really not doing anything for ambient exposure. You want to meter the ambient light to get a good ambient exposure of the surrounding area, not your subject. It is the flashes job to illuminate your subject.

Unless you have a light meter there is no way to predict flash exposure because when you half press the shutter the cameras light meter is metering the ambient light only. This meter has nothing to do with the flash. The flash event has not taken place yet so it can't meter or predict flash exposure.

The flash has it's own exposure system or better said piggy backs of the exposure system. When you fully press the shutter ETTL fires a pre-flash. It is so fast you can't see it. This pre-flash isolates the closest object which is usually your subject by comparing the ambient light and pre-flash reflection back to the camera. ETTL determines the correct flash exposure of your subject based on that reflected light and determines correct flash output duration to correctly expose "your subject".

You can use spot metering when using the flash but it is not via the camera's metering system. The pre-flash just uses a smaller area to determine flash exposure of the subject. This is done using the FE lock.

The flash has two modes if your camera menus offer it - Evaluative and Average which works with flash exposure only, not ambient.

Evaluative. There are multiple metering zones. This is how the flash determines the correct exposure and I described it in paragraph 4. Forget the flash for a second. When shooting ambient only you know that you have to compensate exposure for images of pure white snow or pure black tar. An equal mix of tar and snow will provide a good exposure. The flashes exposure system during the pre flash will see reflected light from a bride in a white dress, a groom in a black tux and both together. This is why you need to FEC often.

Average. The flash metering system just averages out the whole scene. A bride in a white dress is usually surrounded by darker ambient at a reception thus it just balances it out. It does not isolate her and determine exposure off her dress only.

People say average is good for indoor flash photography and evaluative for outside. I use evaluative for both indoor and out but you should avoid average for outdoor flash. Does not work good.

I just came across this video looking for something else. There are lot's of good videos on how to expose for flash and ambient. I can post a few more if you like.

https://www.youtube.co​m/watch?v=shgNCky1c5Y

Forgot to add. I also use evaluative for the cameras ambient metering most of the time. Spot occasionally if I want to meter the sky when shooting outdoors to set up the ambient exposure. I then I shoot, chimp and FEC for correct flash exposure of my subject.

You can over time start to predict much to FEC and you get closer but you almost always have to fine tune a bit. That is why manual flash is better and if I do have the time I will use it. I have a flash meter for that.

http://www.learn.usa.c​anon.com …ite_tip_pt2_art​icle.shtml

I set my camera to M because the system has no idea how I wish the ambient lit parts of the frame to look. If using ETTL II, I then use Average for flash or FEL (forces a central spot pre-flash & lock). I expect to set FEC in both cases depending on the tone & reflectivity of the pre-flash metered area.

THIS. That explains how one would meter a face for proper exposure using flash metering. If you don't use FEL or average in settings, the ETTL II will use evaluative metering. Do I have that right?

Thank you so much. Never used FEL before. As I play with it I see that it should be the last thing one does before pressing the shutter; my back button focus or changing focus boxes appears to obliterate the setting. I really wonder if the people that say they are using spot metering with ETTL2 are really doing what they think they are doing.

mpstan wrote in post #18225015
Thank you so much. Never used FEL before. As I play with it I see that it should be the last thing one does before pressing the shutter; my back button focus or changing focus boxes appears to obliterate the setting. I really wonder if the people that say they are using spot metering with ETTL2 are really doing what they think they are doing.

FEL is not the last thing you do. You do it if you want the flash metering system to meter the light reflected back from the pre-flash within the spot metering circle only. If you want it to meter your whole subject then don't push it. Some people like to meter flash off skin and have their own secret sauce for FEC. If you were shooting weddings and metered off the white dress you could have a formula. I found it annoying.

Thanks for the post. Helps remind me just why I shoot manual and use a meter.

I chose not to allow a camera do my thinking and guessing for me. I'd rather have the exposure dead-on the first time. Under control so I can 'focus' on other aspects of the shoot. Variables are eliminated.

Incident light rocks over reflected light.

mpstan wrote in post #18224996
THIS. That explains how one would meter a face for proper exposure using flash metering. If you don't use FEL or average in settings, the ETTL II will use evaluative metering. Do I have that right?

Yes. And Evaluative for Flash will often do a pretty good job.

OceanRipple* wrote in post #18225101
Yes. And Evaluative for Flash will often do a pretty good job.

As long as we are separating things properly. Thought I'd bring this up just in case someone getting into this does not overlap.

There is evaluative, partial, centre weighted average and spot which the camera uses for the ambient meter reading and they have nothing to do with flash exposure.

Then you have the flash's evaluative and average modes that only pertain to flash exposure.

FEL that only applies to the flash exposure. The system decides on flash exposure from reflected pre-flash information it receives within the spot metering circle.

Most of the time I keep the camera on Evaluative, the flash on Evaluative and this does a pretty good job but there is almost always some FEC required.

mpstan wrote in post #18224996
THIS. That explains how one would meter a face for proper exposure using flash metering. If you don't use FEL or average in settings, the ETTL II will use evaluative metering. Do I have that right?

Dunno what applies in regards to the 5DIII, but most Canon dSLRs have the factory default flash metering set to Evaluative.
You can choose Evaluative vs. Average for flash metering mode in the CFn menu, and that is the default mode for flash metering unless you choose to use FEL (flash spotmetering).


BTW, all dSLRs today have 1) TTL (Thru The Lens) ambient metering AND also 2) TTL (Thru The Lens) Flash metering.
In history, TTL ambient metering was first available commercially in 1964 (Topcon), and TTL Flash metering first became available 16 years later (Nikon).
TTL is a generic term, simply to indicate metering 'Thru The Lens', what the lens sees is what is measured.

All dSLRs have TTL ambient metering, plus they also have TTL flash metering. But dSLR flash metering is quite different in implementation than film camera TTL flash metering.
In Canon's case the 'TTL' is also a flash metering name which applies specifically to flash metering for film cameras only
'aTTL' is an obsolete version of flash metering found in some Canon film cameras. 'eTTL' is the precursor to digital camera 'eTTL-II'. 'eTTL-II' is the flash metering in digital cameras starting from the Canon 20D to today.

Wilt wrote in post #18225277
Dunno what applies in regards to the 5DIII, but most Canon dSLRs have the factory default flash metering set to Evaluative.
You can choose Evaluative vs. Average for flash metering mode in the CFn menu, and that is the default mode for flash metering unless you choose to use FEL (flash spotmetering).

BTW, all dSLRs today have 1) TTL (Thru The Lens) ambient metering AND also 2) TTL (Thru The Lens) Flash metering.
In history, TTL ambient metering was first available commercially in 1964 (Topcon), and TTL Flash metering first became available 16 years later (Nikon).
TTL is a generic term, simply to indicate metering 'Thru The Lens', what the lens sees is what is measured.

Works this way in the 5D3 too Wilt. The other default, I think, is the multi function button controlling the metering: when this is invoked, a spot metering takes the place of evaluative and is supposed to stay active for 16 seconds.

mpstan wrote in post #18224996
THIS. That explains how one would meter a face for proper exposure using flash metering. If you don't use FEL or average in settings, the ETTL II will use evaluative metering. Do I have that right?

You choose the ETTL II Metering mode (either Evaluative or Average) in the cameras menu under "External Speedlite control". You can override the ETTL metering mode when you push the FEL button. When you push the FEL button the flash(s) will emit a preflash and lock in a flash exposure for 16 seconds. The FEL exposure, is based on an average of the metered tones, that fall within the partial metering circle viewed in the viewfinder.

When you choose an ambient metering mode (spot, partial, center weighted, or evaluative) you're basically only changing the size of the sampling area for the ambient meter reading. When you change ETTL II metering modes you are changing the algorithms of how the camera is arriving at a proper ETTL flash exposure. The ETTL II metering mode does not change the size of the sampling area as it meters the entire frame for both metering modes.

What does change between the two ETTL II metering modes is the way the camera arrives at a proper flash exposure. In Average metering it only meters a preflash reading and averages all the tones to arrive at a flash expose. In Evaluative ETTL metering the camera is trying to determine what needs to be properly exposed with flash (aka "the subject"). It uses focus distance information from the lens (if available) but it mostly tries to make this determination by comparing two meter readings (one ambient and one preflash). It compares these two readings and throws out what it deems to be erroneous information and what is left is what it determines needs to be properly exposed with flash. Over simplified but in a nutshell that is the difference between the two ETTL II metering modes.

The camera's meter tries to properly expose for mid gray whether metering ambient or flash. If you're metering tones that are not mid gray then you will have to use FEC (Flash Exposure Compensation) to correct for the meters misinterpretation of the scene. Just as when metering ambient light, the cameras meter when metering flash, will also be fooled by tones that are brighter or darker, and will trying to create a flash exposure based on those tones. If you have bright tones then the camera will try to under expose the flash exposure by reducing the flash power so you will need to add FEC to increase the flash power.

The rule of thumb for using either Evaluative or Average is to use Average in doors or in situations where your flash will provide some illumination for most of the scene and use Evaluative out doors or in situations where a greater part of the scene will not be illuminated by the flash or at best receive very minimal flash illumination. In doors or in a situation where most everything in the scene will be illuminated by the flash, the Evaluative metering mode can be fooled by all the tones filling the scene and may choose the wrong part of the scene to properly expose with flash. The problem with Evaluative metering in this type of a situation is that you do not know what part of the scene the camera is going to choose to properly expose with flash so, you may not always be able make a reasonable assumption of what FEC to apply. Consequently when trying to use Average metering out side or in situations where very little of the scene will receive illumination from the flash, the camera can once again be fooled by those areas not receiving any flash illumination during the preflash reading. Remember the Average metering mode only uses a preflesh reading and does not use an ambient meter reading for comparison. So, it assumes that all the dark open sky or those parts of the scene not receiving any flash illumination are just dark tones that need more flash power to be properly exposed. (see my images below of how these two metering modes differ when trying to meter a subject outside with a lot of open sky).

The best rule of thumb is to use Manual flash mode for the best flash exposure consistency but if you are using ETTL, try to use FEL when you can, so that you are only having to meter a small area that you want to be properly exposed with flash. By sampling only a small area you can more easily determine how much, if any, FEC is needed based on the tones within the partial metering circle. The rub with using FEL is that once you lock in a flash exposure, the distance from the flash to the subject cannot change. So in those situations where the distance from the subject to the flash is changing and where using ETTL may be a preference over using Manual flash, FEL may not be an option. You would then choose one of the ETTL II metering mode's, based on the scenarios I discussed above.

Below is a demonstration of how the two ETTL II metering modes handle a flash exposure for one of the scenarios I discussed above. The situation is an outdoor shoot, with the subject you want properly illuminated, occupying less than 1/2 (less than 1/4 actually) of the space, of the entire scene on a whole. An open sky, which will not receive any flash illumination during the preflash reading, dominates the scene. This scenario can be replicated during the day or night although during the day the ambient exposure will not be so under exposed, with the ambient light contributing more to the overall exposure, to where the camera will not need as much flash. After all the camera's meter is just a light meter, metering all light, whether flash or ambient, based on the camera's exposure settings. The Average metering mode tends to behave a bit more erratically, than does the Evaluative metering mode, as the camera's exposure settings increase or decrease the amount of ambient light contributing to the overall exposure, in this type of scenario.

I set my camera to Manual mode to lock in a camera setting that would allow my flash to properly expose my subject at 1/8 power in Manual flash mode. This is close to the middle of the flash power range giving me +3 stops and -5 stops of flash power leeway when I switched my flash to ETTL.

As you can see this is the type of scenario where it is best to use Evaluative metering and it did a pretty good job of properly exposing the test target. The last image provides the best exposure which was obtained by using FEL centered on the target.