Flash and mirror

Folks,<br>

  I had three shots left on a roll of 200(?)ASA Sensia that

I had left from a trip to Paris and decided that an experiment on 3D

matrix balanced flash was in order.<br>

  In a room with tungsten lighting a chair stood against a

wall with a 1/2 length mirror sitting on the seat cushion lying

against the seat back at approx 15 degrees from the vertical. On the

right side of the mirror I put a white cotton tee-shirt and on the

left a black shirt. Below the mirror I put a gray item which escapes

me now.<br>

  The lens used were my trusty 50mm f1.8 AFn non-D and the

much-rubbished 28-80mm 3.5-5.6 AFD, all mounted on an N/F80. The

flash used was the Metz 53 with 3402 module (3D etc).<br>

  Sitting on a coffee table about 1.5 meters away from the

mirror the head of the flash was just visible at the bottom of the

mirror. I focused on the edge of the mirror and then moved the

central focus point so that it was exactly on the flash head's

reflection. The three shots were all taken at f/8, 1/60s. The first

was straight TTL with the 50mm, manual mode. The second was with the

28-80 set at 50mm, 3D Matrix on the flash and A-mode on the camera.

The third was with the 50mm, matrix on the flash and A-mode again.

  I got the results back yesterday morning and was

perplexed - the shots were all under exposed by a large amount. The

first shot in TTL was the darkest, the second in 3D Matrix yadda-

yadda was the least dark then the Matrix-balanced one was half-way

between the two.<p>

<table border="1">

<tr><th>Lens</th><th>Flash Setting</th><th>Result</th></tr>

<tr><td>50mm AFn</td><TD>TTL</TD><td>Darkest</td></tr>

<tr><td>28-80 AF-D</td><TD>3D Matrix balanced

TTL</TD><td>Dark</td></tr>

<tr><td>50mm AFn</td><TD>Matrix balanced TTL</TD><Td>Darker</td></tr>

</tr>

<table>

<p>

  In each case the flash head appears as a totally clear

spot in the centre of the frame and in no case was the white garment

anything other than a slightly less opaque patch on the otherwise

very dark film.<br>

 

  I had thought that even in the most basic flash mode the

fact that the camera was not focused meant that the central flash

sensor was ignored or at least down-graded but that seems not to have

happened.<br>

  Perhaps the camera tried to prevent too much contrast

from appearing on the frame by biasing for the most reflective areas.

This I could accept from straight TTL but the much-vaunted 3D Matrix

Balanced (Fill) Flash was only a little 'better' than Matrix, which

was itself only a little better than straight TTL.<br>

  The camera knew from the D-lens that it was set to 1.5

meters (say) but that only mattered a little. It did seem to count,

but I wonder how fine-grained the distance reading is from the

lens.<br>

  So my questions are:

<UL>

<LI>Should I have kept the camera focussed on the mirror frame and

not recomposed?

<LI>Has anyone else performed similar tests?

<LI>Am I asking too much from the camera?

<LI>How good is the distance reading from the lens? Is it accurate to

particular distance or does it gate the scene e.g. (1m, 2m, 4m, 8m

etc) so that the results are at most 1/2 stop out?

</UL>

<br>-P.J.

<<The first was straight TTL with the 50mm, manual mode. The second was with the 28-80 set at 50mm, 3D Matrix on the flash and A-mode on the camera. The third was with the 50mm, matrix on the flash and A-mode again. >>

 

You're changing too many variables. Change just one variable per exposure.

 

<<I got the results back yesterday morning and was perplexed - the shots were all under exposed by a large amount. >>

 

The presence of a mirror creates a hole into which photons can disappear by the bucketful, doesn't it? Why did your test incorporate a device that will direct so much of the flash power in a useless direction? You did prove one thing: a mirror in line with the flash gun can deceive your meter in any mode.

 

<<The camera knew from the D-lens that it was set to 1.5 meters (say) but that only mattered a little. >>

 

Really? A mirror prependicular to your shot doubles the "subject" distance if you focus on the frame but shoot into the mirror. Since the mirror in this case was angled 15 degrees from the vertical, the effective distance is more then double the focused distance. One would expect that to foul up the 3D matrix calculations of proper flash output (although the preflash ought to have compensated).

 

<<Am I asking too much from the camera? >>

 

The camera doesn't give a hoot, so no, you're not asking too much of it. But your test appears to incorporate too many variables. You can conclude that, despite multiple combinations of lens, flash setting and focus point, your test fooled your meter every time. You cannot yet draw conclusions about WHY it did so... needs further testing with (to repeat) just one variable changing per shot.

 

Have fun,

Ah, a response!

 

<< You're changing too many variables. Change just one variable per exposure. >>

 

The change in camera ambient metering mode and lens was to make the camera/flash combination use the various flash metering modes. The aperture and shutter speed were the same for all shots.

 

<<Why did your test incorporate a device that will direct so much of the flash power in a useless direction? You did prove one thing: a mirror in line with the flash gun can deceive your meter in any mode. >>

Because Nikon and Nikon users make great claims for what the 3D Matrix Balanced flash system will do. I wanted to do a simple(!) test to see. It's one of the reasons I haven't dropped Nikon for Cannon long ago.

 

The camera knew from the D-lens that it was set to 1.5 meters (say) but that only mattered a little.

 

<<Really?>>

 

Yes.

 

<<A mirror prependicular to your shot doubles the "subject" distance if you focus on the frame but shoot into the mirror. Since the mirror in this case was angled 15 degrees from the vertical, the effective distance is more then double the focused distance. One would expect that to foul up the 3D matrix calculations of proper flash output (although the preflash ought to have compensated).>>

 

Because the lens was focused at the mirror edge the subject-to-flash distance should not be doubled. The mirror itself took up less than 20% of the image and the reflection of the flash head was just visible at the bottom of the mirror. The rest of the scene was within 30 centimetres from the plane of the mirror and took up 80% of what the camera saw.

 

<<The camera doesn't give a hoot, so no, you're not asking too much of it. But your test appears to incorporate too many variables. You can conclude that, despite multiple combinations of lens, flash setting and focus point, your test fooled your meter every time. You cannot yet draw conclusions about WHY it did so... needs further testing with (to repeat) just one variable changing per shot.

 

Have fun,

 

-- Jim Gifford (Silver Spring MD USA) >>

 

Yes, Jim, I'm going to do more tests, but I'm asking if anyone can explain how to get the perfect (or even borderline acceptable)exposures in the same circumstances without using a flash meter or manual flash and a measuring tape.

Hi,

I did a similar test in a dark room with a Nikon N50 (supports only 2D matrix balanced and center weighted TTL Fill-Flash) and pointing a SB-28 towards a mirror.

 

The results were more or less what you described (dark pictures with one bright spot).

 

In addition, I turned then the light on, pointed the flash-head towards the mirror and the result was a well exposed picture with good contrast.

 

A further test with the room-light off and the flash-head tilted towards the ceiling (diffusor card pulled out) lead to a good exposed picture too.

 

So it seems, that the bright reflection on the mirror in a dark room blinds the electronics.

 

-A.O.

1st problem; You sed Sensia and I think that is clor negative (I'm probably

wrong here), so I presume you are judging the prints and not the film. what

does the film look like?<P>

2nd problem: Yes you should have kept the camera position locked down,

<P>

3rd problem: If the camera is looking into a mirror the distance calculations

are going to be thrown off, especially if you are pointing the camera at

differing angles to the mirror.<P>So your tests are extremely inconclusive and

prove basically nothing.<P>A more rigid methodology would be to remove all

ofthe variables; the mirror, put the camera on a tripod and shoot slides. Write

down all of your settings. make sure the flash has enough time to fully recycle

between shots.

Why the hell did you use a mirror and what was the purpose of your experiment anyway???

<<1st problem; You sed Sensia and I think that is clor negative (I'm probably wrong here), so I presume you are judging the prints and not the film. what does the film look like?>>

 

http://www.fujifilm.com/JSP/fuji/epartners/Products.jsp?nav=0&parent=PRODUCT_CATEGORY_238119&product=1000520

 

<<2nd problem: Yes you should have kept the camera position locked down,>>

 

How could this have made a difference?

 

 

<<3rd problem: If the camera is looking into a mirror the distance calculations are going to be thrown off, especially if you are pointing the camera at differing angles to the mirror.>>

 

Re-read the original question. I focused on the frame of the mirror.

 

 

<<So your tests are extremely inconclusive and prove basically nothing.>>

 

Yeah, yeah...

 

<<A more rigid methodology would be to remove all ofthe variables; the mirror, put the camera on a tripod and shoot slides. Write down all of your settings. make sure the flash has enough time to fully recycle between shots.>>

 

The whole point was to shoot into a mirror! Take a look at the title of the question!

Being on a tripod makes no difference to the reflectivity of the mirror or the other items in the scene.

I gave you all the settings!

Yes, the flash had long enough to charge! The 'OK' light was still lit after firing and the in-viewfinder lightening symbol was solid.

Hal,<p>

 

<b>Why the hell did you use a mirror and what was the purpose of your experiment anyway???</b><p>

 

 

I said this to Jim Gifford when he asked the same question:<br><i>

Because Nikon and Nikon users make great claims for what the 3D Matrix Balanced flash system will do. I wanted to do a simple(!) test to see. It's one of the reasons I haven't dropped Nikon for Canon long ago.</i><p>

Shooting into a mirror would seem to be the most extreme test of the Nikon flash system, no?

It is an extreme test, and the Nikon literature would have one believe that the flash system can handle it, but it can't. If I don't pay attention, and shoot with a mirror straight in front of me, I get under exposed film with N80 & F100 with the N80 built in flash, or an SB28. The only way to be sure of getting this exposed right is with the flash in manual mode. (I have no idea why you would want to do this, since it produce a crummy picture and probably lots of lens flare.)

<i>The flash used was the Metz 53 with 3402 module (3D etc).</i><p>

Ahah! Just caught that one! Have you ever read anything I have written about third party flashes? That thing probably isn't even doing 3DBMFF although it may claim to. Try this again with an SB-80DX and forget the mirror!

Hal Bissinger wrote:<p>

<i>The flash used was the Metz 53 with 3402 module (3D etc).</i><br>

<b>Ahah! Just caught that one! Have you ever read anything I have written about third party flashes? That thing probably isn't even doing 3DBMFF although it may claim to. Try this again with an SB-80DX and forget the mirror!

</b><br>

<ul>

<li>Due to other commitments I haven't been able to keep up with your writings but I hope to be able to reacquaint myself with your postings as soon as decorum and stomach-lining augmentation allow.

<li> Given that the three exposures were different (though still underexposed) suggests that 3DBMFF is happening, just not giving me the results that I was looking (and hoping) for. Metz have a more than decent reputation and work without fault when used with the dedicated moudles for other brands of camera. I can only imagine that they have crippled their latest Nikon module out of a sense of misplaced fun, the lovable little circus freaks.<br>Next time I will contrast the results with the built-in flash on my F80, though next time I will probably not use just the last three frames on a roll.

<li>I will await your sending me a SB80DX with huge relief. I thought that my budget would never allow me to buy such a beast! To think that I saved myself the cost of a mesely few hundred rolls of film by buying my (second hand) Metz flash!

<li>As to not trying the experiment again, perhaps my innate fixation on seeing the boundaries of the automation in my equipment will get in the way again. How knows? Perhaps prayer is my only recourse to resist the temptation, though it's had to find a live chicken in the shops these days.

<LI>I am intrigued though. Would using a SB80DX pass such a difficult test?

</ul>

<p>

-P.J.

Bruce Rubenstein wrote:<p>

<b>It is an extreme test, and the Nikon literature would have one believe that the flash system can handle it, but it can't. If I don't pay attention, and shoot with a mirror straight in front of me, I get under exposed film with N80 & F100 with the N80 built in flash, or an SB28. The only way to be sure of getting this exposed right is with the flash in manual mode. (I have no idea why you would want to do this, since it produce a crummy picture and probably lots of lens flare.)</b><p>

Thanks for your response Bruce.<br>

It looks like the test was too stringent for the system. Oh well. Now I know.<br>

I don't expect to take flash shots in similar circumstances again. I did take a <a href="http://pj.mc-kenna.com/Photography/Album/USA2001/Florida/Orlando/ScienceMuseum/SelfportraitFlash.jpg">picture</a> (90KB) using the built-in flash while inside a three-sided mirror arrangement and it came out alright, despite using my 50mm AFn. I am reasonably confident that the camera considered itself 'in focus' at the time the photo was taken which might have had some bearing on the result. More investigation or a visit from Nikon's Men In Black might help.<p>

-P.J.