Why do I get underexposure with flash for macro pictures?

I've been taking close up macro shots of flowers (near 1:1) and

I'm finding that my pictures tend to come out underexposed by 1/2

to 1 stop (unless I compensate).

 

Given that the nikon flash system should take into account distance,

pre-fires, and effective aperature of the lens, can anybody explain

why I get underexposed pictures? Flash pictures at normal distances

turn out properly.

 

My setup is an N80, 105 micro AF-D, sc-17, and an sb-50dx or sb-24.

Both flashes tend to underexpose the picture. I handhold the flash

off camera at about 45 degrees to the lens axis. I'm in aperature

priority, matrix metering, and f22 or smaller.

 

Thanks,

 

Stephen

SC-17 !!!

 

1. That means you are using flash off shoe, which makes distance information utterly useless.

 

2. Even if you are 'on-shoe' distance information has effect only in 3D Matrix Balanced Fill, no effect in plain TTL and you should be using plain TTL around 1:1 for numereous reasons.

 

3. Also plain TTL can get fooled extremely eaisily especially if your subjects are tiny compared to background but this is likely to cause overexposure, not under.

Also plain TTL will cause under exposure if your flowers are lightly coloured unless you apply exposure compensation.

<I>SC-17 !!! <P>

 

1. That means you are using flash off shoe, which makes

distance information utterly useless."</I> <P>Actually Umit .

where the flash is located (on camera or tethered with SC-17)

makes no difference to the "D" calculations. The "D" information

processing algorithms are evaluating the in focus distance to the

camera as compared to the out offocus areas distance to the

camera. Once again where the light is makes no differece to the

3D Matri TTL computations.<P>

But what might be making a difference is how close the flash is

to the subject. If the flash is too close it might be out the range of

the meters ability to accurately compute how much flash is

needed so it cuts off the exposure prematurely. I know this

sounds anti-thetical to my first statement but it isn't. <P>You

might also be having a problem with the reflectivity of the flowers.

If the flower is a pale color and dominates the frame than you

might needto use your wetware and make a flash or exposure

compensation adjustment to compensate for this overall

brightness that the meter is trying to force down to a more

normal brightness level. These camera systems are smart but

they still can't read your mind. Try bracketing with either the flash

compensation or the exposure compensation controls of your

flash and camera.<P>

A final possibility: If you are shooting color negative film and the

shots are being processed by automated machine than the

machine printer might likewise be forcing down the overall

brightness of the shots.

<i>Where the flash is located (on camera or tethered with SC-17) makes no difference to the "D" calculations. The "D" information processing algorithms are evaluating the in focus distance to the camera as compared to the out offocus areas distance to the camera. Once again where the light is makes no differece to the 3D Matri TTL computations.</i><br><br>

 

No Ellis. Preflash is a burst of flash with a fixed and known intensity, amount of bouncing light from the subject, combined with the distance information (light source to the subject distance) gives an idea about the subject reflectance using the inverse square law. If you arbitrarily change the light source to subject distance (unknown amount of light loss) reflectance will be wrongly calculated. Similar thing happens when you tilt the flash head, that's the reason preflash is cancelled automatically when you do so.

I don't know much (make that anything) about TTL flash, but if this were a manual set up, I would question the ultra-tight aperture range, unless you were using pretty fast film. Aperture priority with such a small aperture would send the shutter into pretty slow range. The shutter may be open longer than the flash fires. Again, I don't know much about the auto stuff.

Using the SB-24 will cancel out any 'D' info the camera might be expecting. You need a SB-25 (I think) or newer flash for 'D' data to incorporated.

 

You may want to test a roll (write down each exposure) for the sake of finding out what is happening. Bracket your shots, film is not that expensive. And if you want to eliminate the 'lab sees it one way,' use slide film.

Umit, Sorry you are still wrong. You'd have to move the flash a

great distance in a very short time (micro seconds) between the

preflash and real flash to make a huge difference. I've been

using SC-17 cables --up to three at a time-- with Nikon

Speedlights (SB-24, SB-25, SB-26, & SB-28dx) since they

became available, and I've been shooting mostly slides. I think

I'd have noticed this if it were a problem. If the metering sensor

were in the flash you'd be correct, but it is located in the camera

body. Once again the "D" function is an in camera bias algorithm

that is based on the distance the lens is focused, among other

factors which probably include the size of subject relative to

overall frame, focal length of lens, and ambient lighting levels.

<P>

The preflash is discontinued when you tilt the head because

Nikon assumes you are bouncing the flash when you tilt the

head and the very weak preflash would most likely be

unreadable by the TTL metering cell in most bounce situations.

<i>...I think I'd have noticed this if it were a problem....</i><br><br>

 

In your practice you may have observed that it doesn't make much of a difference probably because your camera to subject and flash to subject distances are practically equivalent, or the offset of flash is negligibly small compared to distance to the subject; but if your subject is 1m away from the camera and you place flash 1.5m away from the subject it is going to make a difference.<br><br>

 

<i>If the metering sensor were in the flash you'd be correct, but it is located in the camera body.</i><br><br>

 

Lets see:<br> <br>

 

1) Flash emits a preflash (with fixed and known intensity) and the camera analyses returning amount of light (from the lens) for 'something' irrelevant to us. OK? <br><br>

 

2) You make some (unknown) amount of this preflash light vanish via some means thats unknown to the camera (via changing distance, tilting, blocking flash with diffuser, blocking lens with a filter etc)<br><br>

 

3) How would you expect this data (corrupted in some way) yields correct answers?<br><br>

 

<i>The preflash is discontinued when you tilt the head because Nikon assumes you are bouncing the flash when you tilt the head and the very weak preflash would most likely be unreadable by the TTL metering cell in most bounce situations.</i><br><br>

 

Although your thinking is sound it is not the case, if returning preflash light were beyond metering cell's limitations (yes sometimes is) camera would know that and do processing accordingly. Inversely this means it would utilise it (preflash info) whenever it can (ie bounced light is significant for the camera's meter). Even if you tilt the flash head down (not to bounce, but to illuminate a closer subject) preflash is cancelled promptly.

How about just dialing in +1 flash compensation in macro flash shots. My sb28 does the same thing. I compensate elsewhere, why not here?

How about using a flash meter for some testing!

Try setting for "Slowsync", this will help eliminate hot spots and take into account existing light for exposure. also, try giving a minus compensation to the flash so that it's more of a fill than main.

Distance of flash to subject does not matter except to minimum and maximum distances. Is the problem only with this lens?

Not sure if this answers your question, but:

 

I also use the 105 micro AF-D, but on a Nikon FE with no flash, so my setup is all manual. I shoot 100 or 50 speed slides using a tripod early in the morning or on cloudy days. I also find that when shooting pictures of light colored flowers they are underexposed.

 

Could it be that because you are shooting a light colored object, the meter is making too light of an exposure calculation? For instance, if you point a spot meter at a white object in the photograph it should turn out grey. Because the light colored flower takes up most or all of the frame, you would need to apply exposure compensation to expose the flower properly whether you are using flash or not.

 

Nick

I happened to stumble over this entry (I'm a Canon user so out of my territory here) and it very similar to a problem I have with my Canon Macro set up:

<p>

I've been taking macro shots with a mixture of canon bodies and both Sigma and Canon Macro lens with a ring-flash.

<p>

When shooting in natural light the TTL exposure works fine (at all magnifications). With the flash I find I have to dial in about +1.0 to +1.5 stops to the flash exposure compensation to get the exposure right. This is true for Velvia, Kodak Elite Chrome EC and Agfa Scala (B&W). Its not a subject colour problem either. Its not the specific camera, not the lens, not the film. Don't see how it can be the flash as the camera is determining cut-off point. It isn't the effect of the changing effective aperture at 1:1 as TTL will take care of that.

<p>

So why isn't TTL working correctly? Pictures are fine - its just a drag having to always dial in compensation or reset film speed.

<p>

I can't help wondering if there is some peculiarity of TTL flash metering that is affected by macro lens?