SB-700 Under-exposing?

Joined
Dec 26, 2009
Messages
1,266
Location
Philadelphia
I'm noticing that when I angle my on camera SB-700 for anything other than direct to subject it's underexposing by a stop. Same results for spot or matrix metering on my D700. Weird, I don't have this issue with my SB-900's or my old SB-600, I thought in TTL, the camera would direct the flash to expose the scene correctly, why should I have to dial a EV +1 when slightly bouncing the flash?
Thanks,
Jon
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
If I understand you correctly you're saying that when using the same camera and either an SB-900 or SB-600 the resulting exposure, as determined by the camera, is correct in your opinion. But in the same or similar situations the SB-700 is approximately -1 EV? In other words, the SB-700 is capable of delivering adequate power (duration) but you have to consistently turn it up one stop by using +1 EV.

If so, is this behavior consistant regardless of the situation? That is, ceiling height and color, room size, ambient exposure, etc?
 
Joined
Dec 26, 2009
Messages
1,266
Location
Philadelphia
that is correct but I haven't done scientific testing, is this something others have noticed?

If I understand you correctly you're saying that when using the same camera and either an SB-900 or SB-600 the resulting exposure, as determined by the camera, is correct in your opinion. But in the same or similar situations the SB-700 is approximately -1 EV? In other words, the SB-700 is capable of delivering adequate power (duration) but you have to consistently turn it up one stop by using +1 EV.

If so, is this behavior consistant regardless of the situation? That is, ceiling height and color, room size, ambient exposure, etc?
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
I'm not aware of any reports of this, but of course that doesn't mean there haven't been any.

I suggest performing a simple experiment by ruling out as many variables as possible. Keep everything the same except for the flash. And use a tripod so distance and angle do not change. Keep detailed records and images for Nikon service in case no one else has the answer.

And lastly, be sure and rule out any errors in the user module (that being you). :wink:
 
Joined
Apr 3, 2006
Messages
5,614
Location
Texas
I'm noticing that when I angle my on camera SB-700 for anything other than direct to subject it's underexposing by a stop. Same results for spot or matrix metering on my D700. Weird, I don't have this issue with my SB-900's or my old SB-600, I thought in TTL, the camera would direct the flash to expose the scene correctly, why should I have to dial a EV +1 when slightly bouncing the flash?
Thanks,
Jon

One thought - TTL BL mode uses the focus distance reported by the D lens. This can override in some degree to modify what the TTL BL metering is otherwise saying... and can be more correct sometimes in the worst problem cases. However, this lens distance info is NOT used when the head is tilted from level, nor is it used in TTL mode (only when in TTL BL mode). This could change what you may see, but it really does not change the bigger overall situation, which is my first assumption, below.


This +1 EV is not routine, but it is not really unusual for TTL flash, sometimes, in some situations. It is what the Flash Compensation is for... you can simply watch and then fix it to be exactly like you want it.

It is simply not reasonable to imagine the light meter will always be "right". It will not always give the right result. It is a reflective meter, and it responds to the reflectivity of the subject. This is never what anyone wants to hear, and it may take longer to actually understand it, or to be willing to believe it, but ultimately, this will be the only workable answer. The first step towards understanding and believing is to try some different varied subjects, in different scenes.

If the scene is a black cat in a coal mine, the meter will try to make the overall average come out middle gray. But what it needs then is -2 or -3 EV Flash Compensation.

If the scene is a polar bear in the snow, the meter will try to make it come out middle gray. But what it needs then is +2 or +3 EV Flash Compensation.

This is simply how the light meter in the camera works. This is how they have worked since they were invented decades ago. We are the photographer, and this is our job, to watch and fix it.

It is NOT proper to set +1EV compensation and leave it there, imagining that we are "calibrating" the meter. Maybe in ONE same situation if needed, yes, but it is NEVER a one time adjustment to make the meter "right". The meter is already "right". What needs the compensation is the scene in front of the camera, and each scene can vary.

Indoors, we often have large expanses of white or light colored walls, or white shirts, or whatever, and the meter will try to make everything gray (underexposes white). THEN the answer may be to add say +1EV Flash Compensation. It is not terribly unusual, it happens. But each scene in front of the camera will be different.

I do hear you, but my first assumption is that the other flashes in fact all work the same way. Some scenes simply do meter a stop under, if their reflectivity is higher. So just to believe it, for starters, one time try all of the flashes with camera on a tripod, to exactly duplicate the situation, everything exactly the same except the flash model. The TTL exposure depends on the subject reflectivity. The metering is in the camera. You have already reported that the flash itself does OK when metered well.

See the light meter link below for more. This is simply how reflective meters work. It is about the first thing to be learned about photography.

Bottom line - Flash Compensation is how we control the TTL flash. It is a very good thing, when needed.
 
Joined
Dec 26, 2009
Messages
1,266
Location
Philadelphia
Quick thanks for the response Wayne, I'm going to have some coffee and dig in, appreciate your time on this.
Jon

One thought - TTL BL mode uses the focus distance reported by the D lens. This can override in some degree to modify what the TTL BL metering is otherwise saying... and can be more correct sometimes in the worst problem cases. However, this lens distance info is NOT used when the head is tilted from level, nor is it used in TTL mode (only when in TTL BL mode). This could change what you may see, but it really does not change the bigger overall situation, which is my first assumption, below.


This +1 EV is not routine, but it is not really unusual for TTL flash, sometimes, in some situations. It is what the Flash Compensation is for... you can simply watch and then fix it to be exactly like you want it.

It is simply not reasonable to imagine the light meter will always be "right". It will not always give the right result. It is a reflective meter, and it responds to the reflectivity of the subject. This is never what anyone wants to hear, and it may take longer to actually understand it, or to be willing to believe it, but ultimately, this will be the only workable answer. The first step towards understanding and believing is to try some different varied subjects, in different scenes.

If the scene is a black cat in a coal mine, the meter will try to make the overall average come out middle gray. But what it needs then is -2 or -3 EV Flash Compensation.

If the scene is a polar bear in the snow, the meter will try to make it come out middle gray. But what it needs then is +2 or +3 EV Flash Compensation.

This is simply how the light meter in the camera works. This is how they have worked since they were invented decades ago. We are the photographer, and this is our job, to watch and fix it.

It is NOT proper to set +1EV compensation and leave it there, imagining that we are "calibrating" the meter. Maybe in ONE same situation if needed, yes, but it is NEVER a one time adjustment to make the meter "right". The meter is already "right". What needs the compensation is the scene in front of the camera, and each scene can vary.

Indoors, we often have large expanses of white or light colored walls, or white shirts, or whatever, and the meter will try to make everything gray (underexposes white). THEN the answer may be to add say +1EV Flash Compensation. It is not terribly unusual, it happens. But each scene in front of the camera will be different.

I do hear you, but my first assumption is that the other flashes in fact all work the same way. Some scenes simply do meter a stop under, if their reflectivity is higher. So just to believe it, for starters, one time try all of the flashes with camera on a tripod, to exactly duplicate the situation, everything exactly the same except the flash model. The TTL exposure depends on the subject reflectivity. The metering is in the camera. You have already reported that the flash itself does OK when metered well.

See the light meter link below for more. This is simply how reflective meters work. It is about the first thing to be learned about photography.

Bottom line - Flash Compensation is how we control the TTL flash. It is a very good thing, when needed.
 
Joined
Sep 21, 2008
Messages
6,374
Location
Alabama
One thought - TTL BL mode uses the focus distance reported by the D lens. This can override in some degree to modify what the TTL BL metering is otherwise saying... and can be more correct sometimes in the worst problem cases. However, this lens distance info is NOT used when the head is tilted from level, nor is it used in TTL mode (only when in TTL BL mode). This could change what you may see, but it really does not change the bigger overall situation, which is my first assumption, below.


This +1 EV is not routine, but it is not really unusual for TTL flash, sometimes, in some situations. It is what the Flash Compensation is for... you can simply watch and then fix it to be exactly like you want it.

It is simply not reasonable to imagine the light meter will always be "right". It will not always give the right result. It is a reflective meter, and it responds to the reflectivity of the subject. This is never what anyone wants to hear, and it may take longer to actually understand it, or to be willing to believe it, but ultimately, this will be the only workable answer. The first step towards understanding and believing is to try some different varied subjects, in different scenes.

If the scene is a black cat in a coal mine, the meter will try to make the overall average come out middle gray. But what it needs then is -2 or -3 EV Flash Compensation.

If the scene is a polar bear in the snow, the meter will try to make it come out middle gray. But what it needs then is +2 or +3 EV Flash Compensation.

This is simply how the light meter in the camera works. This is how they have worked since they were invented decades ago. We are the photographer, and this is our job, to watch and fix it.

It is NOT proper to set +1EV compensation and leave it there, imagining that we are "calibrating" the meter. Maybe in ONE same situation if needed, yes, but it is NEVER a one time adjustment to make the meter "right". The meter is already "right". What needs the compensation is the scene in front of the camera, and each scene can vary.

Indoors, we often have large expanses of white or light colored walls, or white shirts, or whatever, and the meter will try to make everything gray (underexposes white). THEN the answer may be to add say +1EV Flash Compensation. It is not terribly unusual, it happens. But each scene in front of the camera will be different.

I do hear you, but my first assumption is that the other flashes in fact all work the same way. Some scenes simply do meter a stop under, if their reflectivity is higher. So just to believe it, for starters, one time try all of the flashes with camera on a tripod, to exactly duplicate the situation, everything exactly the same except the flash model. The TTL exposure depends on the subject reflectivity. The metering is in the camera. You have already reported that the flash itself does OK when metered well.

See the light meter link below for more. This is simply how reflective meters work. It is about the first thing to be learned about photography.

Bottom line - Flash Compensation is how we control the TTL flash. It is a very good thing, when needed.

I agree. I don't think its a flash problem. The camera and flash can only interpret so much. As a photographer, you have to adjust. Auto ISO is really nice too, but doesn't help too much when shooting a white dress on a sunny day with a dark background.
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
I think some are misinterpreting what the OP is saying. When asked if the issue was repeatable, "But in the same or similar situations the SB-700 is approximately -1 EV?" he said yes.

That is, given the same subject, subject color, reflectivity and distance, etc, the SB-700 is consistently resulting in exposures that are -1 EV when compared to the other two flashes on the same camera.

Granted, a black tux can result in an overexposed TTL flash image and a bright white wedding dress can result in an underexposed image for the reason Wayne said. But apparently that is not what is happening according to the OP.

Thus the suggestion of a controlled experiment where all variables are kept the same, including use of a tripod so the flash angle and distance are also identical.

More than likely, however, there is a variable involved that is not being taken into account and which results in the -1 EV exposure because, if things are as he stated, it is very odd behavior indeed. Hence the experiment should reveal what that variable is and will serve to educate the OP as to what must be considered. And it's fun to do as well.
 
Joined
Dec 26, 2009
Messages
1,266
Location
Philadelphia
Exactly, I suppose I should perform a more controlled test. I really think this flash is "behaving" differently in TTL than my other flash models. Also, anyone have any idea what the SB700 can only be set to TTL with spot meter? what I use matrix it's TTL BL only.....

I think some are misinterpreting what the OP is saying. When asked if the issue was repeatable, "But in the same or similar situations the SB-700 is approximately -1 EV?" he said yes.

That is, given the same subject, subject color, reflectivity and distance, etc, the SB-700 is consistently resulting in exposures that are -1 EV when compared to the other two flashes on the same camera.

Granted, a black tux can result in an overexposed TTL flash image and a bright white wedding dress can result in an underexposed image for the reason Wayne said. But apparently that is not what is happening according to the OP.

Thus the suggestion of a controlled experiment where all variables are kept the same, including use of a tripod so the flash angle and distance are also identical.

More than likely, however, there is a variable involved that is not being taken into account and which results in the -1 EV exposure because, if things are as he stated, it is very odd behavior indeed. Hence the experiment should reveal what that variable is and will serve to educate the OP as to what must be considered. And it's fun to do as well.
 
Joined
Apr 3, 2006
Messages
5,614
Location
Texas
Exactly, I suppose I should perform a more controlled test. I really think this flash is "behaving" differently in TTL than my other flash models. Also, anyone have any idea what the SB700 can only be set to TTL with spot meter? what I use matrix it's TTL BL only.....

Only TTL being available in Spot metering mode is a camera function, and is true with every flash model, including the built in flash and commander/remote modes.

TTL BL tries to balance the flash with the background ambient.
There is no concept of background or ambient in Spot metering mode. The spot you meter is what you get (as middle gray).
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
SuisseNik said:
...what I use matrix it's TTL BL only....

I believe you meant TTL BL is only available when in matrix metering mode. But make sure you are in TTL only (not TTL BL) when using CLS. Unless of course you have a backlit subject and want the flash to fill in the shadow, which is useful when the Sun is behind your subject, for example.

I only say this because if all you can get is in fact TTL BL when in matrix metering then your flash is clearly malfunctioning.
 
Joined
May 16, 2009
Messages
1,575
Location
Downey, CA
Funny, when I bounce my flash, SB-900, 700 or 600, I usually have to add at least +0.3 EV due to light loss, and +1.0EV isn't uncommon. It all depends on the color of the surface you're bouncing from (the lighter the better), and the distance the light must travel from camera to bounce surface to subject.
 
Joined
Dec 26, 2009
Messages
1,266
Location
Philadelphia
"I only say this because if all you can get is in fact TTL BL when in matrix metering then your flash is clearly malfunctioning. "
This is exactly what I'm saying, am I missing a way to change modes or something with the SB700?
 
Joined
May 16, 2009
Messages
1,575
Location
Downey, CA
I think some are misinterpreting what the OP is saying. When asked if the issue was repeatable, "But in the same or similar situations the SB-700 is approximately -1 EV?" he said yes.

Unless the OP mounts the camera on a tripod, keeps the composition the same, and tries the experiment with both the SB-700 and SB-900 with identical exposure settings (something he admits he hasn't done -- "unscientific"), then we really can't know what's going on.

To understand why TTL, even though measuring preflashes, needs positive flash compensation (in my experience, for any of my SB-nnn flashes), goes something like this: yes, in theory the preflashes should allow the TTL flash meter to measure the light just as it would if you used direct (on-axis) flash light. The problem is that not all the measured light is reflecting from your subject. Say you are bouncing from the wall to your left: the TTL meter will measure the light reflecting from your subject _plus_ the light reflecting from the side wall back to your flash unit. If you subtract the amount of light that is not reflecting from your subject, you can now visualize why your subject is under-exposed.
 
Joined
Apr 3, 2006
Messages
5,614
Location
Texas
Unless the OP mounts the camera on a tripod, keeps the composition the same, and tries the experiment with both the SB-700 and SB-900 with identical exposure settings (something he admits he hasn't done -- "unscientific"), then we really can't know what's going on.

That is of course true. Nothing else can be helpful, until the problem is known.

To understand why TTL, even though measuring preflashes, needs positive flash compensation (in my experience, for any of my SB-nnn flashes), goes something like this: yes, in theory the preflashes should allow the TTL flash meter to measure the light just as it would if you used direct (on-axis) flash light. The problem is that not all the measured light is reflecting from your subject. Say you are bouncing from the wall to your left: the TTL meter will measure the light reflecting from your subject _plus_ the light reflecting from the side wall back to your flash unit. If you subtract the amount of light that is not reflecting from your subject, you can now visualize why your subject is under-exposed.
Then how would you explain that bounce does not always need compensation? And even then, not always +EV (sometimes -EV too, but certainly more rare however).

It is true that indoors, I do always use TTL instead of TTL BL, because there seems no meaning of having TTL BL balance with insignificant ambient.
(however, and another subject, is that direct flash D lens distance info can keep wild TTL BL readings more limited. But that is direct flash only, not about bounce). But I'd say bounce TTL compensation is only needed maybe half the time (gut feel). However, certainly it is also true that +1EV is commonly needed too. Just not all time.

Metering accuracy depends on the subject, specifically, the reflectivity of the subject. The meter always tries to give a middle gray image, which works well with middle gray subjects. :smile:

Beginners always want to assume their light meter ought to always be correct. Of course, truth is, the way reflective meters works, it can only rarely be correct. It meters what you point it at. Point it at different things, and it will meter different values. At best, a reflective meter is a guide, to be used with the photographers brain.


I have recently enhanced the material at How Light Meters Work . I think the first couple of sections is clear enough now for everyone.

In particular, all of these photos below are with the same setup (direct flash, matrix metering, automatic TTL flash, f/8, 1/200 second).

The ONLY difference is that the black background paper was moved slightly left for each frame. ALL ELSE IS THE SAME. Yet, the exposure shifts from overexposure to underexposure. This is simply how it works.

What we meter depends on what we aim the meter at. It is good to know and expect that. :smile:

dse_4758.jpg
Subscribe to see EXIF info for this image (if available)


dse_4759.jpg
Subscribe to see EXIF info for this image (if available)


dse_4760.jpg
Subscribe to see EXIF info for this image (if available)


dse_4761.jpg
Subscribe to see EXIF info for this image (if available)


dse_4762.jpg
Subscribe to see EXIF info for this image (if available)
 
Joined
May 16, 2009
Messages
1,575
Location
Downey, CA
Then how would you explain that bounce does not always need compensation? And even then, not always +EV (sometimes -EV too, but certainly more rare however).

In my experience, this is rare. If the "light math" I proposed is true, then it is true regardless of the situation. It may be the case, however, that given your angle of bounce, or proximity to the bounce surface, there may be some situations where the non-subject reflected light is insignificant to an extent where the TTL meter produces nearly identical or identical results to those that would come about with 100% of the light reflecting back from the subject. Remember also that the TTL meter measures and adjusts in discrete steps, not infinitesimal gradations, so that round-off may yield equivalent or identical results to 100% reflected light.

I don't see how the examples you post here, which essentially are varying the scene metered or being lit, pertain to a discussion about keeping all things equal, except the particular flash unit that is being used as a bounced flash.
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
WayneF, your "How Light Meters Work" article is excellent. It demonstrates the reason I often suggest to people just starting out to use a handheld lightmeter. That way they can quickly see that what their cameras reflective meter suggests and what the handheld incident meter reads can often vary by two stops or more. Hence, they learn to 'see' and compensate for their cameras meter and soon find they become sort of their own lightmeter.

Then I no longer hear something like "I took a picture of a black cat and it came out sort of greyish." :wink:
 
Joined
Apr 3, 2006
Messages
5,614
Location
Texas
In my experience, this is rare. If the "light math" I proposed is true, then it is true regardless of the situation. It may be the case, however, that given your angle of bounce, or proximity to the bounce surface, there may be some situations where the non-subject reflected light is insignificant to an extent where the TTL meter produces nearly identical or identical results to those that would come about with 100% of the light reflecting back from the subject. Remember also that the TTL meter measures and adjusts in discrete steps, not infinitesimal gradations, so that round-off may yield equivalent or identical results to 100% reflected light.

That was my thinking too... if true, it should always be true. Yet, in my experience, the effect of +EV compensation is only needed maybe half the time for bounce flash. But... there are always surprises.

The problem is that not all the measured light is reflecting from your subject. Say you are bouncing from the wall to your left: the TTL meter will measure the light reflecting from your subject _plus_ the light reflecting from the side wall back to your flash unit.
The preflash and the final flash both follow the same route to the lens and metering. From the flash, off of the ceiling, off of the subject, and to the lens. Whatever is lost, and whatever makes it, is the same either way. Both are metering the reflection from the subject, which is where the lens is aimed. So that seems wrong by default.

I don't see how the examples you post here, which essentially are varying the scene metered or being lit, pertain to a discussion about keeping all things equal, except the particular flash unit that is being used as a bounced flash.

I thought it very clearly showed that the same camera and same flash and same TTL system taking a picture of the same rose in the same setting, gets rather different results by simply varying the reflectivity of the scene slightly. Same shutter button push, different results.

And yet, some of us expect the meter to always give accurate readings, no matter what we point them at. I'm trying to suggest to those people that they have a little more to learn.
 
Joined
Mar 31, 2010
Messages
684
Location
North of Seattle, South of Canada
SuisseNik said:
This is exactly what I'm saying, am I missing a way to change modes or something with the SB700?

No, I was mistaken. The SB-700, unlike the SB-900 for example, does not allow you to select just TTL even though the slide switch is only labeled TTL. Instead, when in matrix or center weighted modes it automatically displays TTL BL on the LCD.

Therefore, the SB-700 in combination with the camera, determines what mode (either TTL or TTL BL*) to use depending on the subject and whether or not it's backlit. But either way, the SB-700 LCD will always display TTL BL when in matrix or center weighted modes. So for example, if you focus on uncle Bob and he's strongly backlit the flash should theoretically use TTL BL and balance the flash with the backlight. However, like anything automated, it can potentially be fooled. And there may be situations where you don't want TTL BL but simply TTL. Of course, the opposite can also be necessary.


*In other words the SB-700 does this internally but the display won't change.
 
Last edited by a moderator:
Joined
Apr 3, 2006
Messages
5,614
Location
Texas
The SB-700, unlike the SB-900 for example, does not allow you to select just TTL even though the slide switch is only labeled TTL. Instead, when in matrix or center weighted modes it automatically displays TTL BL on the LCD.

Wow! I owe you an apology, because my first immediate thought was that this was not reasonable, and so there must be some other explanation. Yet page C-2 seems to say exactly that.. Camera Matrix or Center Weighted switches to TTL BL.

So then, that would be the same as the internal flash, and same as remote/commander mode, but apparently, now on hot shoe too.

So... looks like Spot Metering is now the only way to get into real TTL Mode...


Thanks much for the comments on the metering page. I knew it was good info, but sometimes the presentation is a problem.
 

Latest posts

Links on this page may be to our affiliates. Sales through affiliate links may benefit this site.
Nikon Cafe is a fan site and not associated with Nikon Corporation.
Forum post reactions by Twemoji: https://github.com/twitter/twemoji
Forum GIFs powered by GIPHY: https://giphy.com/
Copyright © Amin Forums, LLC
Top Bottom