F100 + SB28 underexposure

[jem]

Hello

 

I've just recently picked up an F100 and an SB28 (having had before

this an F80 and a SB50DX). I've just run a few rolls through it,

mainly indoors in small rooms using a Stofen Omni-Bounce (usual

family and friends festive shots).

 

All my outdoor exposures seems fine (good dense negs and trannies are

spot on too - either straight ambient exposure or as ambient with -

1.7 fill).

 

However all my flash as main light exposures are under, at least a

stop, sometimes as much as 2 stops it would seem.

 

Now, I have been using the flash as follows: Matrix TTL, camera set

on matrix metering, flash at 45 degrees, flash set to -0.3 as

perviously I'd found my sb28/f80 shots to be a tad hot on tranny).

 

In general I do the following (for basic friends and family shots):

 

Outdoors / enough ambient light for fill flash - matrix

metering/matrix ttl flash, flash set to -1.3 or -1.7. This is

generally just fine, probably I'd rather use only -1.0 or less for

the fill but it always seems too hot with this setup.

 

Indoors (flash as main light) - Well here I am fumbling around a

bit. But generally I use either bounce flash off the ceiling with

the little white card out, or the Omnibounce if in a smaller room.

Flash set to -0.3 as I said as I have had hot results in the past. I

have used matrix metering (generally 1.5+ stops under) or straight

TTL (one stop under) or A mode (one stop under). Generally I am

using Kodak Portra 400NC, and I am dragging the shutter, to say

1/15th or so in well lit rooms, to get a bit more ambient light into

the background. Typically I rate neg films one stop under their

speed (ie Portra 400 I rate at 200), so I'm effectively over exposing

one stop - thing negs are not what I was expecting!

 

I have just tried a test with my Sekonic 508 flash meter set to the

same ISO as my camera and these are the results (f100/sb28 manual

mode 1/60th f8, flash meter on shelf unobscured, with dome out)

 

Flash straight on (camera metering always matrix, no film in

therefore ISO 100, 2m from flash meter):

 

Matrix TTL : f4.0

TTL : 8 on the nose

A : 5.6 +7/10

 

Flash + Omni at 45 degrees

Matrix TTL : 5.6 (+6/10)

TTL : 5.6 (+8/10)

A : 5.6 (+3/10)

 

Flash straight up, card out (ceiling 3.5m high)

Matrix TTL : 5.6 (- 0.3 indicated on sb 28)

TTL : 4.0 (8/10)

A : 4.0 (6/10)

 

 

OK, so the results are not too bad with the omni, although still

definitely under. With the flash straight on why is the matrix ttl

mode giving me f4, ie two stops under, consistently??? And with

bounce, given I have 7.5 metres to play with as claimed on the back

of the sb28, I should be able to get f8 on my subject (ie the meter) -

and at the very least it should indicate underexpsure, not just

underexpose by two stops and not report it. In many cases I was

*defnitely* in range, but still getting underxposure by one to two

stops...

 

So, what am I, or the camera, doing wrong? If something was wrong

with the f100s ttl flash meter, why does the fill flash work so well?

 

Any help appreciated - non studio flash is a black art to me at the

moment!

 

Thanks for reading all this!

 

Jeremy

[umd]

<p><i>Now, I have been using the flash as follows: Matrix TTL, camera set on matrix metering, flash at 45 degrees, flash set to -0.3 as perviously I'd found my sb28/f80 shots to be a tad hot on tranny).</i></p>

 

<p>What are your subjects? By tilting flash at 45 degrees you are cancelling preflash and if your subject deviates from medium gray camera won't be able to detect it and give underexposure if its brighter than medim gray, and overexposure if darker.</p>

[hal_bissinger]

Jeremy, first off I will repeat what has been said here at least 100 times- you can't meter TTL flash. Your results will be inconsistent and meaningless. On top of that you had no film in the camera and since TTL flash metering relies on a certain reflectance provided by the film, on two counts your tests are meaningless.<p>

Suggest you try this as a legitimate test- <br>

- Flash in 3D Balanced Matrix fill flash mode<br>

- No flash comp.<br>

- Flash straight ahead, no bounce card.<br>

- Your favorite transparency film at the rated ISO. <br>

- Shoot indoors as you did before. <p>

Now lets see what the results are. Keep in mind that cameras can be different and that's certainly going to be true of the N80 and F100.

[jem]

** Umit - my subjects are people, so not too far from medium grey. Maybe they could do with half a stop more to lift the skin tones. But does normal (non matrix) TTL flash *really* take into account the colour, or rather as I suspect, just the reflectance of the object? ie at its simplest, it just looks for 'f8 worth of light' back from whatever it has hit, and then shuts off the flash.

 

** Hal - Sorry if I've missed the obvious or oft-repeated but having read about flash here on Pnet and in books etc, I haven't actually heard anyone say that you can't meter TTL flash.

 

(Actually it doesn't make sense to me either - f8 * iso 100 is a fixed amount of light is it not, so with standard film in anyway, you should be able to measure TTL flash like any other type of flash surely?? I admit the lack of film in is an issue and will try again with film in in a minute, but since I've jsut done a quick search and not found an answer as to why you can't meter TTL flash could you give me the one line summary or point me to a previous discussion/reference - would be much appreciated!)

 

Given all that, I *did* have film in the camera, I was in small rooms with white ceiling, I shot on camera matrix, flash matrix ttl -0.3, Omni at 45 degrees, Agfapan 100 and Portra 400 NC, and got underexposure. Worse with the Agfa I think but that maybe cause I was using it when it was actually night, not dusk as such.

 

Thanks for your thoughts,

Jem

[umd]

<p><i>does normal (non matrix) TTL flash *really* take into account the colour, or rather as I suspect, just the reflectance of the object?</i></p>

 

<p>No, only matrix balanced fill flash does that, straight TTL assumes medium gray; btw caucassian human skin is NOT medium gray, usually one stop brighter. Hal is indded right in his statement, the aperture camera gives in matrix metering mode (you must be using program or shutter priority, right) is irrelevant from the flash, it is only for the ambient light.</p>

[jem]

Umit,

 

I think Hal meant you can't measure TTL Flash even with a flash meter? Obviously the camera's meter is measuring the ambient light not the flash...but I'm using a Sekonic 508 flash meter (with the dome out) to meter the flash in its cordless flash metering mode. So I'd have thought you can indeed measure, say f8 when you focus on the flashmeter and set off the sb28 on A or TTL mode at f8.

 

Also, I don't even think matrix fill flash takes into account colour - I think it just tries to make a better balance between ambient light and the flash lit subject (ie if tries to match the exposure of subject as determined by ?focus point? and the background ambient light.

 

I'm using manual mode exposure on the camera.

[jem]

Ok - I've had more time to look now and it seems that pre-flashes /short flash duration mean that my Sekonic 508 may not be giving me accurate readings.

 

But the fact still remains I have 12 rolls with under exposed shots on them , with the only difference being the swap on anf F80 to a new f100 - so, my basic question is how should I set up the flash/f100 for bounce or omnibounce situations at night?

 

I notice Ken Rockwell says his f100 typically underexposes flash by about a stop (or more accurately the three he has had). Is this an f100 thing, i.e. a possible or fault, or something the f100 jsut does differently?

 

Thanks for any further help

 

Jeremy

[umd]

<i>I don't even think matrix fill flash takes into account colour</i><br><br>

 

It takes the subject reflectivity (brightness) in to account (not the colour) by using the distance information from a D type Nikon lens and reflected (returning) 'preflash' light. Preflash is not fired at straight TTL. Preflash is promplty cancelled if you tilt flash head and will more or less mess the calculation of subject reflectivity if you use a diffuser. One exception I know of is the SB-80Dx which has its own supplied diffuser, that don't mess anything because of its known transmittance. I think same should be true for SB-28 if it has a supplied one.<br><br>

 

Although what you say seems logical about the flashmeter, it is not; because TTL (or 3d matrix balanced fill, don't matter) depend on reflected light from the film, firing flash in one of these modes without film won't give any predictable result. It will on A mode though.

[Jerry_]

If the camera's meter is on Matrix, you will generally confuse it by using exposure compensation on the camera. Indoors, the SB-28 is not a "Flashbulb" and accordingly, you need more light than expected for a good exposure. Too much light can be corrected in printing, no-enough-light is not so forgiving.

 

A good rule of thumb - shoot a test roll before the results are "required" as you want good negatives for your project.

[anson_jones]

In a nutshell, f8 is WAY to small. You are bouncing off a rather high ceiling, you have already reduced the effectiveness of the flash by about two stops (by using the diffuser), and you are bouncing the flash across the heads of the subjects. Point the flash straight up. Put the camera in Program mode, slow synch, for a few shots and see what kind of readings you get. Just try it.

[patrick_davies1]

I experimented with bouncing and diffusers for about a year. (F-100 + SB-24). Ceiling bouncing was good in certain circumstances, but I didn't want to rely on it. LumiQuest softboxes and reflectors work o.k., too, but the light loss did cause problems at times. I still use these things when I know the conditions will be good for them. I've had varying degress of under/over exposure. One of my theories is that the bounce flash lights up more than just your main subject and TTL cuts off the output when it measures the whole scene as average, which will make your main subject appear a stop under. My gut tells me that Matrix TTL was always doing what it's supposed to do, but not what I want.

 

I have had success with the F-100 with SB-28 on a flash bracket. (Press-T). No diffuser to reduce the output. The flash is just over the lens. Red eye is eliminated. Shadows are thrown slightly behind the subject and are typically not noticeable. I don't know if these are problems you're trying to correct with softboxes, reflectors, etc. Just getting the flash up a little bit made my flash pictures noticeably better. The disadvantage of the bracket is that it's conspicuous and makes the camera/flash a larger thing to wield around.

 

I shoot porta 400 NC at 400 with the flash bracket set up. Sometimes 320. I typically run in program mode, but I'll sometimes run in manual mode and let the TTL drive the correct amount of light for my aperture.

[hal_bissinger]

<i>...But the fact still remains I have 12 rolls with under exposed shots on them , with the only difference being the swap on anf F80 to a new f100 </i><p>

I'm not discounting that there may be a problem with your F100 here but the only way to find out is to rule out "operator error". If you still are underexposing I would send it to Nikon service for evaluation.

[charles_miller]

I use an SB-28 on an N90s with an MF-26. I find that -2/3 flash compensation is good for outdoor fill flash, but NOT for indoor fill flash. I don't know the reason(s), but I cancel flash compensation when shooting indoors with incandescant lighting, and the results are much more to my liking.

 

Infrared spectrum from incandescent lights can fool some meters, and would cause balanced fill flash to calculate incorrectly. I don't know if this is a problem with the F100 or not. But in the past, some of the most expensive hand held light meters have been very sensitive to infrared, and they gave readings a stop too high when operating in incandescent lighting. So I wouldn't be shocked if the F100 TTL meter has this problem too. I truly don't know for sure what the problem is, only that the same situation you describe also exists whith the N90s.

 

As to why balance fill flash works so differently when the flash is tilted for bounce from the ceiling, I can only guess that this an entirely separate problem.

[scott_carlsson]

Wow, plenty of rocket science goin' on here. I would get rid of the Stofen or whatever, put the flash on a bracket, leave the Matrix on, flash straight ahead, manual with shutter at a 1/30th and fire away.

 

I've got the F100/SB28 and it's mostly very good, if anything it's a bit hot. I usually flip the wide diffuser down for all focal length shots that are less than 10 feet or so away. I've got the Stofen and for the most part I didn't see enough difference in the shots to have to worry about the 1-2 stop loss.

 

You know it's a peanut size light source however you dress it up.

[jem]

Ok thanks for all the answers.

 

Anson - admittedly in my tests the ceiling may be a bit high but as a rule the ceiling was lower, maybe 2m say. Hence f8, iso 100, bounce flash or omni should be possible with an sb-28 shouldn't? (Subject about 2/3 metres away, 85mm lens and flash zoom)

 

Patrick - I've been meaning to get a bracket, will definitely give it a go. The basic plan is to : decrease hot spots on the faces, fill up these small areas with light by a combination of directing the flash in more directions (either bounce or omni), reduce red eye, reduce shadows and or hide them behind the subjects. Just a general shooting technique for indoor parties/functions. Basically, what a good wedding photographer does at reception I guess?

 

Hal, I'll try a straight on basic test tonight.

 

Scott - you're right, and that sounds like a decent simple technique, perhaps with a bracket added to get the flash more directly over the lens (vertical shots) and generally further of the axis.

 

Cheers again for all your thoughts, Jeremy

[david_h._hartman]

If you want the flash to be the main light try using standard TTL not balanced flash. Perhaps it�s trying to balance with the available darkness?

 

There is very likely nothing wrong with your camera. It�s most likely doing exactly what it was designed to do. I�ve heard this before.

 

One other though is what is the reflectance of the film you are using? Is it lighter than average? I would guess a combination of errors have all gone toward under exposure.

[david_h._hartman]

<em>"There is very likely nothing wrong with your camera.

Its most likely doing exactly what it was designed to do. Ive

heard this before." dhh<br>

</em><br>

This doesnt read right. What Ive hear before is the

problem of under exposure with the Nikon F100 and flash.

[jem]

I will try standard TTL, even A mode, when using it as a main light. However it does seem to habitually underexpose, and worsse with either bouncing or the stofen. I too have heard of f100 under exposure when using flash, which is why I asked. It should be pretty simple, really. Just flash on TTL to avoid trying to balance out dark areas, drag the shutter, and away you go, eg

 

f5.6 on camera and flash, shutter on 1/15th maybe, and in small rooms with lowish ceilings, the Omni to bounce the flash around more and generally fill up the space with light. The camera should meter the flash on the main subject as determined by focus point and turn off the flash once f5.6 has been reached.

 

But this doesn't seem to be working (with a variety of film types and speeds). In the end, the rest of the info just confuses the matter.

 

And none of this changes the fact that the flash should *indicate* the under-exposure if it occurs, ie if the flash is not powerful enough.

[jem]

In case anyone is still listening, I have since had the camera serviced unedr warranty - the repair sheet says the TTL system was re-calibrated. It is now clinically accurate when using straight on TTL flash. It still seems to like +1 on the flash when doing bounce or Omni at any distance above 4 feet or so.