Different measurements with spot vs incident metering on same target?

[Diapositivo]

I read the answer by Sekonic as: you cannot compare a reflected reading with a incident reading with a lumisphere, because they collect different light.
(But you should be able to compare a reflected reading with an incident reading with the disc, because they collect the same light).

Or: if you angle your lens, your grey card, and your incident light meter with disc in such a way that they collect the same light, they should give the same reading.

For a document reproduction work, with front light or with let's say two side lights, the two systems (spot measuring and incident measuring with disc, not with sphere) should in theory give the same result, I do reckon.

 

[xtolsniffer]

A couple of years ago I was bored one day and got a whole bunch of cameras out, plus a couple of hand-held meters and decided to calibrate them all against one another using a grey card (three cards in fact from different manufacturers). Most cameras agreed to within a third of a stop when using centre-weighted or spot against the grey card, but I always and consistently found that an incident reading at the card position was about one to two thirds of a stop off the reflected reading both from the meters and from the cameras, with the reflective readings underexposing relative to the incident reading i.e. f8 at 1/20th for incident and f8 1/30th using centre-weighted metering in a Nikon F4. The result was that I stopped using grey cards as I couldn't be bothered to work out why they were not the same and now I just bracket a half a stop either side if I'm using 35mm slide film, otherwise half a stop is within the exposure tolerance of most negative films for my levels of quality control and is probably about the tolerances of the apertures in most lenses anyway. I'm much happier now...

 

[markbarendt]

Interesting shutterboy,

Thanks for the effort and for sharing. Hearing words from a differing perspective helps, even if it is to see what should have been obvious. That response from Sekonic gives me some interesting clues to how to judge a scene and better use my incident meter.

First is the point light source idea. Imagine for a moment a campfire lit scene in the wilderness someplace. While not a true point source it is approaching that idea. The lumishpere would be expecting more light and generally not be getting it so the meter, I would think, should suggest a camera setting to allow more exposure than the lumidisk which can still "see" the whole fire. Just did a very crude test of this thought (Streetlight) and got exactly that result, the lumisphere asked for more exposure.

That nuance is really cool, even if I already knew it worked this way, I had not applied the thought in the same manner.

If all you have to light a scene is a street light or a campfire the lumidisk, pointed at the source rather than the camera may be a better way to peg exposure.

In the context of this thread it also shows that no defect is required to have an incident meter indicate the same or differently than a spot meter on a known target, regardless of how the target of the spot meter is getting it's light the incident meter is seeing a much different picture.

For me that calls into question Dunn and Wakefield's thought that there is no practical difference between using an incident meter and a gray card.

Again thanks!

 

[markbarendt]

One really cool thing about using the lumidisk pointed at the "main" light source is that it may return a considerably faster shutter speed suggestion while still being very accurate about placement to avoid underexposure. More things to try, woot!

 

[juan]

I once tried an experiment with the BTZS software. I measured a scene with an incident meter (one reading in shade - a second in sun) and entered the values. Then I measured the same scene with a spot meter and entered Zone II and Zone VIII into the software. I was surprised to see the indicated exposures were about two stops different - I believe the reflected values were the lesser exposure. I tried this with 5 or 6 scenes and the results were consistent. I thought the BTZS algorithm was supposed to work the same for indicent and reflected readings. I can't explain it.
juan

 

[trondsi]

One stop off seems to be similar to my testing of an incident meter recently, albeit in bright light. Using the incident meter pointing at the camera my shots were underexposed. See the thread in the lighting part of the forum.