Different readings, flashmeter in reflected/incident modes

[wiltw]

Irregardless is not a word. Regardless is a word.

Similarly, a blue loveseat is NOT the same brightness as a gray card by any wild flight of imagination.

Gray cards are carefully crafted to provide uniform reflectance at all wavelengths of light.
That is impossible with a "blue" surface by definition.
Using a blue reference surface makes the readings highly variable based on the meter sensors and filtering.

- Leigh

Picking at nits...if I aim the spotmeter at the gray card, and then aim it at the blue loveseat, the reading nevertheless is 'the same' level of intensity...good enuff for government work!

BTW there is such a word. It may not be perfectly proper, but it exists and has existed for nearly a century

 

[Bill Burk]

wiltw,

I think it would be a simple test to take a spotmeter reading off the reference light source... and then aim another light at the reference light source plane... adjust that new light until an incident meter located at the reference plane looking at the new light reads the same combination of f/stop and shutter speed... then position a gray card next to the reference light source and see how they look side-by-side.

However they appear, will be a topic of conversation.

 

[Leigh B]

then position a gray card next to the reference light source and see how they look side-by-side.

As I mentioned previously, and as is shown on page 2 of the Kodak instructions...

The angle of the card to the light and the meter is critically important.
By changing the angles you can get almost any reading you want.

- Leigh

 

[Bill Burk]

It's true the angle is important, wiltw has demonstrated that changing the angle can give almost any reading.

So I followed the Kodak instructions and moved my gray card so that it is in accordance with the angles recommended by Kodak: 1/3 from camera to perpendicular, 2/3 from card to light source.

Then I moved the light until the incident reading on the gray card side matched the spotmeter reading of the reference source.

Then I took a single shot and in Photoshop, I superimposed the results.

The gray card is the square in the middle. It appears lighter.

As a corollary experiment I moved the light until the spotmeter agreed between reference and gray card.

I had to move the light away farther. Then an incident reading indicated a lower amount of light, as might be expected when you move the light farther away.

Sekonic L758-DR used for both readings. Tungsten photoflood bulb. ISO 100, the meter reads 30 f/11.6 at the reference source with spotmeter (reference source is nominally 100 foot-lamberts). Then when spotmeter agrees gray card reads 30 f/11.6, the incident reading was f/11.2 with the dome retracted and meter aimed at light source, f/11.1 when aimed at camera. With dome out, it read f/11.0 whether meter aimed at camera or at light source.

 

[Chan Tran]

Irregardless of constant light or flash illumination, the metered results will be similar in relative terms...in this case the photo shows the two meters used, and the ISO 400 1/30 f/stop measured with each meter type. In the case of the reflective reading it measured with -0.3EV of the incident reading whereas (per the great perpetual debate about 18% vs. 12.5% gray card) the reflective light reading should have measured more light, not less.
(And the blue cloth upholstery of the loveseat happens to be the same brightness as a gray card.)

As stated already by a number of us, the results of the two meter types are largely 'in the ballpark' of one another, and are absolutely not 2EV different as observed by the OP.

I suspect that the blue seat is darker than an 18% gray card. If it's 18% reflectance then the reading should be reversed that is the spotmeter should have higher reading than the incident meter. If it's 18% reflectance then the reading would be the same only if you use the flat diffuser.

 

[Bill Burk]

I believe my experiment supports Kodak's recommendation that the 18% gray card reading should be increased by 1/2 stop for subjects of normal reflectance.

 

[Chan Tran]

Bill your experiment showed essentially no difference between a flat diffuser and the dome. I think under the same lighting condition the flat diffuser should read higher than the dome by about 1/2 stop.

 

[darkroommike]

Irregardless is not a word. Regardless is a word.

...

- Leigh

The word was first used in 1795, Webster's thinks it a word. Good enough for me.

Irregardless of other posts, the OP was measuring apples and oranges. An incident light meter or metering dome reads illumination and reflected light meter measures reflectance. Both are valid methods of operation but should not be compared to each other except in cases where known standards are metered to calibrate or check calibration.

 

[wiltw]

Bill your experiment showed essentially no difference between a flat diffuser and the dome. I think under the same lighting condition the flat diffuser should read higher than the dome by about 1/2 stop.

It is difficult to generalize!

I just did this set of readings in my home on a sunny cloudless day (so not very much variability to the level of illumination)...in six locations within three rooms of my home. In all cases, two gray targets were laid on the floor or on a table, and an incident meter laid on top of them both. The gray card has a matte finish to present little angular reflective sheen, as does the EZ Balance fabric target. Four light readings were rapidly taken at each location, within about a 15 second period in order to reduce any likelihood of change in illumination.

• The reflective readings were generally brighter than the incident readings, as expected per the axiom that 'middle gray is simply the middle of the tonal range, it is NOT the meter calibration brightness' ...but we also see that in one case the incident reading was brighter than the EZ Balance reading!
• The relative position of flat disk vs. hemiphere was equally split...in HALF the cases the hemisphere was brighter, and in half the flat disk was brighter!
• The spread between lowest reading and brightest reading at any position varied from a low of 0.6EV to a high of 1.7EV, but generally the incident reading was dimmer than the reflected light reading -- but not in every case!
  • In two cases, the reflected light reading was LOWER than the incident reading with hemisphere, which is contrary to usual expectations -- yet both were coplanar.

IOW, light readings are a GUIDE to exposure (this is a statement seen many times in print, but taken to heart by all too few users). The spreads between readings are wide! One should see better correlations than 0.75EV if readings should be treated as 'gospel'

 

[Chan Tran]

If the light meter only get me 0.75EV then I don't need it. My guess generally is within 1EV.

 

[wiltw]

If the light meter only get me 0.75EV then I don't need it. My guess generally is within 1EV.

...so you use no meter reading?!

 

[Chan Tran]

...so you use no meter reading?!

I went on the darkside and don't use meter any more. Take it back! I use the meter to measure light but not for exposure any more. I wish that I can afford the real light meter.

 

[Mr Bill]

I just did this set of readings in my home ...

Hi, thanks for all the data points - I know it's quite a bit of work. I spent some time going through the results, thinking that there would be some likely explanations for the wide variation, but I'm not really seeing anything. (One result I would point out, though, is the "incident hemisphere" reading in col F - it seems to be out of place, and I wonder if that might have been a typo - should be 8.9 instead of 7.9.)

Something that particularly bothers me is the differences in reflective readings between the "spot card" and the "EZ Balance." From what I follow, these are two different grey card materials, one read with a spot meter and the other with a different (?) meter. But given that the geometry seems to be the same, I don't see how the readings can change their relationship (sometimes the "spot card" is ~ 1/2 EV brighter, other times it is about the same.) So something looks fishy to me - is it maybe related to a specular reflection, or perhaps a spectral response difference, or even blocking some light with the non-spot meter.

All in all, though, some interesting data demonstrating the difficulty of good, reliable readings. As a note I've spent a lot of years working with high-volume portraiture (in dedicated studios), where I've been pretty involved in the procedures and setup. At one time we had lots of issues with grey card results, etc. Eventually we set up our studios with a "procedure" for shooting grey cards - they are clipped to a stand at a specific height over the "subject position," perpendicular to the camera (which is at a certain height). This means that nothing is close to the grey card to reflect unwanted light on to it, and the camera won't see spectral reflections. We also used mainstream lighting styles, where the lighting (main and fill) is mainly frontal. Under these conditions, it was very predictable where the grey card densities (read on film) were going to end up based on an incident meter reading. Nevertheless, we never (ok, very seldom) let a new studio go "live" without evaluating test film in the home office.

So my "substantial" experience using incident meters in mainstream studio configurations is that they are very reliable in setting up camera exposure, and the results closely track with grey card densities under the specified shooting conditions.

But when one varies from this, no guarantees - it goes to the knowledge of the photographer to get proper readings according to what they want. Let me make one example. One sort of test lighting we had used was with overhead rail systems with suspended lights. The track for main light was on an arc around the "subject spot." So I can move this light anywhere from near the camera to straight out to the side of the subject, with the light-to-subject distance staying constant. So essentially you would say that the (human) subject exposure does not change. Yet if you took "standard" meter readings, or measured grey card densities, they would vary drastically, EVEN THOUGH THE PROPER SUBJECT EXPOSURE DOESN'T CHANGE, at least not substantially. Again, the knowledge of the photographer is important when we vary from the mainstream frontal lighting configurations.

 

[wiltw]

Something that particularly bothers me is the differences in reflective readings between the "spot card" and the "EZ Balance." From what I follow, these are two different grey card materials, one read with a spot meter and the other with a different (?) meter. But given that the geometry seems to be the same, I don't see how the readings can change their relationship (sometimes the "spot card" is ~ 1/2 EV brighter, other times it is about the same.) So something looks fishy to me - is it maybe related to a specular reflection, or perhaps a spectral response difference, or even blocking some light with the non-spot meter.

Oh, gosh, apologies <brain fArt>, I re-checked my data and during conversion between speed+aperture to EV numbers, I screwed up in posting post 34 results and conclusions! But the data point which you questioned is indeed unchanged!
Here is the corrected data collected...

So the results still prove that you cannot generalize...

• the reflected readings of the gray card were brighter than the incident hemisphere, but not always...one hemisphere reading matched the gray card reading.
• And the reflected readings of the EZ Balance were brighter than the hemisphere, but twice the hemisphere was brighter!
• And half the time the EZ Balance readings were brighter than the gray card, but half the time they were dimmer!
• And generally the incident flat disk resulted in dimmer results than the hemisphere, but once they were very similar
• and lastly, the hemisphere readings seemed to be more than 1/3EV darker than reflected light readings off the gray card, contrary to the 12.5% vs. 18% debates

... so much for trying to generalize.

All in all, though, some interesting data demonstrating the difficulty of good, reliable readings. As a note I've spent a lot of years working with high-volume portraiture (in dedicated studios), where I've been pretty involved in the procedures and setup. At one time we had lots of issues with grey card results, etc. Eventually we set up our studios with a "procedure" for shooting grey cards - they are clipped to a stand at a specific height over the "subject position," perpendicular to the camera (which is at a certain height). This means that nothing is close to the grey card to reflect unwanted light on to it, and the camera won't see spectral reflections. We also used mainstream lighting styles, where the lighting (main and fill) is mainly frontal. Under these conditions, it was very predictable where the grey card densities (read on film) were going to end up based on an incident meter reading. Nevertheless, we never (ok, very seldom) let a new studio go "live" without evaluating test film in the home office. So my "substantial" experience using incident meters in mainstream studio configurations is that they are very reliable in setting up camera exposure, and the results closely track with grey card densities under the specified shooting conditions.

But when one varies from this, no guarantees - it goes to the knowledge of the photographer to get proper readings according to what they want. Let me make one example. One sort of test lighting we had used was with overhead rail systems with suspended lights. The track for main light was on an arc around the "subject spot." So I can move this light anywhere from near the camera to straight out to the side of the subject, with the light-to-subject distance staying constant. So essentially you would say that the (human) subject exposure does not change. Yet if you took "standard" meter readings, or measured grey card densities, they would vary drastically, EVEN THOUGH THE PROPER SUBJECT EXPOSURE DOESN'T CHANGE, at least not substantially. Again, the knowledge of the photographer is important when we vary from the mainstream frontal lighting configurations.

Indeed.
As Bill pointed out earlier, I had done a shot series showing the effect of angle on the perceived brightness of a gray card, as seen by the lens yet exposed with a single shutter+aperture combination, with photos posted on APUG in another thread. So consistency of method, such as you set up, can be important in a very consistent setting (the studio).

 

[Bill Burk]

and lastly, the hemisphere readings seemed to be more than 1/3EV darker than reflected light readings off the gray card, contrary to the 12.5% vs. 18% debates

Whether it's contrary depends which position you take.

I take the position that 12.5% is a better stand-in gray for exposure calculation. And your experiment supports that.

Your experiment shows
The hemisphere believes the illumination is less... exposure meter will recommend a greater exposure.
The gray card is too light, it makes it seem the illumination is greater... exposure meter will recommend a lesser exposure.

My position is that to avoid underexposure, the gray card to choose, (if you want Incident and Reflected light readings to be closer to each other more frequently)... would be in the neighborhood of 12.5%

 

[Bill Burk]

Yet if you took "standard" meter readings, or measured grey card densities, they would vary drastically, EVEN THOUGH THE PROPER SUBJECT EXPOSURE DOESN'T CHANGE, at least not substantially. Again, the knowledge of the photographer is important when we vary from the mainstream frontal lighting configurations.

Great story! Thanks Mr Bill

 

[wiltw]

Whether it's contrary depends which position you take.

I take the position that 12.5% is a better stand-in gray for exposure calculation. And your experiment supports that.

Your experiment shows
The hemisphere believes the illumination is less... exposure meter will recommend a greater exposure.
The gray card is too light, it makes it seem the illumination is greater... exposure meter will recommend a lesser exposure.

My position is that to avoid underexposure, the gray card to choose, (if you want Incident and Reflected light readings to be closer to each other more frequently)... would be in the neighborhood of 12.5%

Bill, I tend to agree. Yet, once in a while I do get 18% gray card reflected light readings which AGREE with incident hemisphere readings, and on those occasions the 12% vs. 18% debates seem immaterial.
That brings me back to my earlier statement, "it is hard to generalize"
..."18% gray card reflected readings tend to result in readings which show more light than incident hemisphere readings, except when they match"

 

[BMbikerider]

One point that does not appear to have been brought up is, when taking a relected light reading of a subject you are getting an average reading from the whole scene, unless you go right up to the main subject. Even then it may not be strictly correct. If it is a portrait, the general average caucasian skin tone is about 1 stop less than the meter would suggest (F11 instead of F8). I mention this to illustrate my next point more clearly.

When taking an incident reading out of doors in daylight we can just turn around and take an incident reading pointing the meter towards the camera because the general diffraction of light that emanates from 93,000,000 miles away is more or less even everywhere.

In the studio where flash is the main source of light, to take an incident measurement we have to go to where the subject is and point the meter at the flash. This simply because of the inverse square law and the way the light is concentrated as it leaves the flashgun. With multiple flashes, I have in the past taken a reading from each and averaged out the reading. You will not be far out.

 

[Chan Tran]

WILTW! I think your readings of the gray card were inflated due to specular reflection. Since you use Minolta meters it should match the reading of the flat disk when measuring an 18% gray card. Yes 18% if you use Minolta meters.

 

[elerion]

Sorry for getting back to the point where I started.
I know about the 18% reflectance, etc,...
I got another flashmeter, a dedicated one, that works on incident mode. It reads perfectly, and exposures are just spot on.

So, there may be something wrong with my lightmeter.
It works fine with ambient light, and exposures are fine if I take flash measurements in reflected mode.
BUT, on incident mode, I get too much overexposure.
Could it be because the dome is not strong enough to withstand a flash light pulse directly?
Any Lunasix F user which can confirm?

 

[wiltw]

WILTW! I think your readings of the gray card were inflated due to specular reflection. Since you use Minolta meters it should match the reading of the flat disk when measuring an 18% gray card. Yes 18% if you use Minolta meters.

Specular reflecton -- from a gray card with a matte sprayed surface to avoid such reflections from surface sheen?!

I can tell you that the room illumination during my test readings included only window light which was entering the room laterally, yet I stool about 70-90 degrees above the floor/table upon which the gray targets were positioned flat, so that any possible specular reflectivity would be occurring at the angle of reflectance (about 30-45 degree to the side) -- the same as the angle of incidence.

So you need to guess again about any cause of reflectivity due to surface sheen.

As for your statement about reading which matches that of the flat disk...in my sample readings graph only ONCE EVER was any reading same as the flat disk reading -- one hemisphere reading = one flat disk, and none of the two reflective surface readings matched.

 

[elerion]

No one ever measured incident flash with a Lunasix F? Am I that weird?

 

[MattKing]

That is how I use my Lunasix F when I am metering flash.

 

[wiltw]

wiltw has given us a lot of good information about the gray card.

And our discussions about how far 18% gray card differs from the calibration of meters, is on the order of 0.33 EV.

Two stops is a significant difference, so it sure sounds like there is a problem with the test OP is doing.

Here is a new observation made about metering vs. actual exposure! It involves the use of d*gital camera meters and exposure, so I hope I am not being too offensive to someone in recounting these observations.

I took Canon 40D, 5D, and 7DII cameras, aiming each at a uniformly lit and uniformly colored wall. The 40D and 5D meters agreed, but the 7DII meter indicated the need for +0.33EV more exposure than their readings.
I exposed one frame on each, and in reviewing the histogram of each shot, l found that the 40D and 5D both placed the exposure so that the histogram peak was at -0.33EV from perfect center of the histogram. The 7DII put its histogram peak EXACTLY AT the histogram center.

IOW, different meter readings may not disagree about the measured reflectance, but may differ in the expected exposure result! If we corrected both 40D and 5D reading by +0.33EV to land their peaks at histogram center, their meters agreed exactly with the 7DII.

 

[elerion]

That is how I use my Lunasix F when I am metering flash.

I get a 2-3 stop overexpose (and unusable images) using incident light flash measurement.
But I get nice exposures in reflected mode.
Does it work fine for you in both modes?
Could I be missing something?
The meters works fine in ambient light. But I tried with different surfaces, and incident flash always read a much less light intensity that leads to overexpose.