Underexposure in camera using light meter Sekonic L-308S

[DennisS]

I just purchased the Sekonic L-308S for use in my home studio - essentially to calculate the light ratios between the Key, fill and hair light.

Yesterday, I measured the lights separately and then with all of them (facing the lumisphere direct to camera lens and under my chin) to get a final reading of f/2.8 with ISO100 and 1/200 secs shutter. I took a self-portrait making sure I had a grey card next to my chin facing the lens.

I could see immediately from the LCD that it was underexposed. I pulled in the photo into LightRoom CC and used AUTO TONE which bumped up the exposure by a stop.

I then exposed the reflective sensor on the light meter (without lumisphere) and pointed it at a grey card (XRite Color Checker Passport) and this matched exactly with my camera reading of the same grey card in the same light.

So it appears the light meter itself is accurate but I cannot explain why I am getting an underexposed image. I was hoping that I could rely on the measurement and just dial in that same number into the camera.

Camera: Nikon D5100 (Crop).

[Manfred M]

With an incident light meter, the Lumisphere has to be pointed AT the light source, not at the camera. You are supposed to be measuring the light falling on the subject, i.e. the incident light.

The gray card is going to be fine for setting your white balance, but nothing more. If you are using the reflective head (or your camera's light meter), then the gray card will work for a REFLECTIVE light reading.

Once you have set your key light (main light), then repeat the same steps for the other lights in your setup. I personally would suggest you start with one studio light (with a simple reflector for fill light) and master than before going through the complications of working with your fill lights and other secondary light sources (kickers, hair, rim, background, etc).

[pnodrog]

I would have thought if you pointed the Lumisphere at the camera rather than at the light source you would end up with an overexposed shot. Something seems a bit odd.

[DennisS]

Yes, that is exactly what I had done when measuring each light - pointed lumisphere directly at light. Also just used the grey card to see what the RGB values would be in LR. In any case, I ran through the effort again - first with Key and Fill, then with only Key. Camera set to f/2.8, 100 ISO and 1/200sec, manual mode - same used to measure to get f/2.8 on the key and just under a stop on the fill.
1. f/2.8 @ key and f/1.6 @ fill - underexposed by about a stop (for kicks - measurement pointed at lens - f/3.2)
2. f/2.8 @ key only - underexposed by a stop and third.

Grey Card RGB - 45% for all 3 channels on the key side. Looking at this I believe since I am dark skinned (that is what I am basing the exposure), my face appears underexposed. I bumped up the exposure in LR using my face as reference. What RGB values should I expect in LR for a grey card with proper exposure?

Keylight - monolight w/softbox (camera left)
Fill - speedlite (off camera) - w/softbox (camera right)
Trigger - Yongnuo RF-605N

[PhotomanJohn]

Let me throw in my two cents. There is often some confusion as to when the meter should be pointed towards the light and when it should be pointed back at the camera. Both are correct depending on the lighting situation. Let me use the following example.

Assume we have two lights, main and fill, one on each side of the subject. If we want to measure the outputs of each light separately then the meter should be placed at the subject and pointed at the light we want to measure. If the lights are both fired during the measurement then the lumisphere should be placed in the retracted position or a flat diffuser should be used to prevent light from the other light from affecting the measurement. Say we then measure the other light and adjust its output or change its distance from the subject to provide one stop less illumination at the subject.

If the light are widely spaced such that the main light falls on one side of the face and the fill on the other without any overlap, we then have a lighting of ratio of 2:1 and the exposure should be set per the reading of the main light.

Now lets move the lights closer to the axis of the lens such that a portion of each shines on the center part of the face like the nose. We now have a lighting ratio of 3:1 since the center area is receiving light from both flashes. In this case a reading taken with the meter facing the camera would provide a better measurement of the highlight since it would sum the contribution from both flashes.

Generally, some judgement is needed to effectively use a light meter in multiple light situations and how it is used will vary.

I hope this helps.

[DennisS]

As per John I did swap out the lumisphere for the lumidisc but did not see much of a difference. I will keep experimenting to capture more data points but in the meanwhile this may help to understand my issue - the right image is after I hit the Auto tone - the values including histogram are representative of the After shot - bumped up by a little under +2/3 @0.80 - as shot was @0.12:

[Black Pearl]

The shot on the left looks to be just about perfect - the blacks are black and you've detail in the lighter squares.
The (corrected) shot looks completely over exposed with washed out (grey) blacks and all the detail burned out of the highlights.

I'm not sure what you are trying to achieve?

[george013]

I don't know if this will help, but I might think of 2 differences between measuring with camera or with an external lightmeter.
First one is that the camera is measuring through the lens and thus is taking in account the looses through the glass and a correction of the AoV for close distance.
Another one is that lightmeters doesn't have to be calibrated in the same way. Figures are going around for a averagel reflection between 12% and 20%.


George

[russellsnr]

https://www.youtube.com/watch?v=Lbv69Um_FcE

[Manfred M]

This video is definitely not the way I do my lighting setups. I use the Sekonic L-358, just like the one in the video. I have the Lumisphere up in all of my readings

I only have one light on at a time. I start with my key light turned on and get a reading and will set that up to give me the right amount of light at the f-stop I am planning to shoot at. I then turn it off and move to my fill light and repeat my steps, then I move to the hair lights / rim light and do the same. If I am using any other lights, ditto.

I then turn on all the lights and take the shot.

From a lighting ratio standpoint; I if I am shooting the key light at f/11 and I want a lighting ratio of 2:1, I will set the fill light until I get a reading of f/8 (i.e. it requires 1-stop more light than my key light to properly illuminate my subject), so if I shoot at f/11, this means the fill light is delivering 1-stop less light than the key = 2:1 ratio. etc. etc.

[DennisS]

The (corrected) shot looks completely over exposed with washed out (grey) blacks and all the detail burned out of the highlights.

Actually the black and white points on the auto corrected photo are correct - nothing is blown out.

All I am trying to do is dial in the suggested reading from the light meter into the camera and take the shot. Obviously it is not as cut and dried and there are some other factors in play here that I am trying to come to grips with. Ideally I would like a proper exposure reference that I can work with rather than rely on my judgement or LightRoom auto tone. I was thinking taking a shot of a grey card may work but then I need to know how to translate the RGB values for the grey card in the uncorrected photo to a reference - like 128,128,128, 112,112,112 or some other 18% or 12% values. I understand gamma comes into play also when making such comparisons.

[Manfred M]

Dennis - can I assume you are using your camera in manual mode?

I started shooting studio flash about five years ago and have never run into the type of problem you are reporting. Could you post the image of yours that isn't exposed properly? I find with studio lights my exposures and white balance are bang on and is one area in photography where I have no issues when I shoot jpegs, as they are usually very good SOOC.

[xpatUSA]

. . . . I need to know how to translate the RGB values for the grey card in the uncorrected photo to a reference - like 128,128,128, 112,112,112 or some other 18% or 12% values. I understand gamma comes into play also when making such comparisons.

118. I forget where I got that from but I've always used it.

[DennisS]

Yes Manfred - full manual. I will post the image later tonight but glad to hear you find this unusual. It is more than likely user error but not exactly sure what that is as yet.

[Manfred M]

118. I forget where I got that from but I've always used.

If Wikpedia is right, then you're close Ted. This article suggests 119, 119, 119.

https://en.wikipedia.org/wiki/Middle_gray

[xpatUSA]

If Wikpedia is right, then you're close Ted. This article suggests 119, 119, 119.

https://en.wikipedia.org/wiki/Middle_gray

Also, in some discussion over on LL:

Mid-gray (L* = 50) can be converted to gamma 1.8 or 2.2 values by use of Bruce Lindbloom's companding calculator. The normalized pixel values are 0.3907 for a gamma 1.8 space and 0.4635 for a gamma 2.2 space. The corresponding values in 8 bit are 99.62 and 118.2 respectively. Lightroom reports values in normalized notation for gamma 2.2.

If you create a mid-gray in Photoshop by filling an sRGB image with midgray (L* = 50), the sRGB pixel value in Photoshop is 119, close to the calculated value of 118.2.

I used to frequent Bruce's site a lot, might have got 118 from there, and then there's the gamma thing - usually stated as 2.2 for sRGB when many of us know it ain't.

Also, the ColorChecker (mini-24 patch) leaflet data shows patch 22 as RGB= 122,122,121 for a lightness L* of 50.867 . . . under an illuminant of D50, mind.

[DennisS]

Here are the latest - same settings - full manual, f/2.8, 100 ISO, 1/200 secs and Key @f/2.8 and Fill @ f/2.2 - both @ 45 deg angle to subject and same height. The first 3 images here are SOOC. The meter was pointed at the lights during measurement - key is at camera left. I also turned off the other light during measurement. Lens is a Nikon 60mmn f/2.8 macro. I have used color picker tool in PS for RGB values. All seem to be a stop under:

Will post the same adjusted for black and white points in next post:

[DennisS]

and corrected for Black and White points:

[Manfred M]

I think I understand what you are doing, although I'm not quite sure why you are doing this.

A couple of questions to start.

1. How far from your subject is your key light located? To be shooting wide open with studio lights and to have an aperture that large strikes me as a bit strange. You are not pushing a lot of light out for this to be happening.


2. What light modifier(s) are you using and how large are they? Are you using grids on them?


3. Are you metering all of the shots with your flash meter at exactly the same distance from the light source? I assume your meter is reading exactly f/2.8 at 1/200th at ISO 100? Or is the reading showing a decimal stop as well?




If I have the time I'm going to pull one of my lights out tomorrow to do some testing.

[xpatUSA]

and corrected for Black and White points:

Difficult to see numbers in these images because they are posted quite small. The Panda one is 639x443px for example. Could you post the screen captures somewhere around 1500x1000px?

In the color checker shot, the mid-gray patch #22 is about 1/3 EV under-exposed in terms of Lab lightness L*, i.e., Log(2) of (39.9/50.9) = -0.351 EV, for what it's worth.

Good luck with the quest.