## Incident light meter burning question

The place for questions about shooting with Blackmagic Cameras.
• Author
• Message

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Ok, so this has been bugging me for awhile now.

I have one light source. I perform an incident meter reading at the subject which is 10 feet from the light source. My camera is set 10 feet from the subject. I set my camera per the incident measurement but the light has to travel another 10 foot to get to my camera and should only be 1/4 the intensity of the light at the subject. The meter has no idea how far my camera is from the subject. How can I possibly get a good exposure?

Steve Fishwick

• Posts: 986
• Joined: Wed Mar 11, 2015 11:35 am
• Location: United Kingdom
Hi Walter, an incident light reading is always relative, i.e. the f stop it indicates will be accurate (if the meter is accurate) at any distance. Incident light readings are normally, in the digital age, still relevant for lighting ratios, e.g. between key and fill.

Leon Benzakein

• Posts: 1275
• Joined: Thu Aug 15, 2013 3:40 pm
Point of departure.

Television: Lighting/Cameraman, O.B. Camera Operator, Experience in EFP, EPG and ENG , Grip, Lamp Operator
Film: Grip, Lamp Operator
Theater: Lighting Designer, Light board Operator, Stage Electrician, Stage Management

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Steve Fishwick wrote:Hi Walter, an incident light reading is always relative, i.e. the f stop it indicates will be accurate (if the meter is accurate) at any distance. Incident light readings are normally, in the digital age, still relevant for lighting ratios, e.g. between key and fill.

Thank you Steve. I believe you but it seems to defy physics. Light, according to physics diminishes over distance by the inverse square law. According to physics as I understand this the first reading at 10 feet from the light source should be 4 times brighter than at the camera where the light has to travel another 10 feet. Even at 100% reflection e.g. 18% Grey card being 50% reflection, the distance the light travels to the camera is twice the distance that it travels to the subject in the example I gave. I can see where the meter would be accurate using the light of the sun as the light has traveled a kabillian miles to the first measurement point and an additional 10 foot would be negligible but I just can't wrap my head around not loosing light as the light travels from the subject to the camera when the light source is close.

Steve Fishwick

• Posts: 986
• Joined: Wed Mar 11, 2015 11:35 am
• Location: United Kingdom
What I mean to say Walter, is take your incident light readings in regards to lighting ratios, then use the 'spot' meter that is in all cameras to read the average, correct balanced exposure.

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Leon Benzakein wrote:Point of departure.

Really good videos for sure. That's pretty much how I use my light meter except instead of using a 90% reflective surface I use a 50% reflective surface i.e. 18% grey card and set that on my waveform at 50 IRE. I have a little more testing to do but to date I find that if I set my camera using my Sekonic L-478DR my camera exposes hot by about a stop per my preference. I need to double check that on the waveform in Davinci Resolve to see exactly where the 50% grey card sits in DR.

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Steve Fishwick wrote:What I mean to say Walter, is take your incident light readings in regards to lighting ratios, then use the 'spot' meter that is in all cameras to read the average, correct balanced exposure.

Got it!!! I can do that without messing with my head. That makes perfect sense.

mickspixels

• Posts: 240
• Joined: Tue May 17, 2022 10:29 pm
• Location: Western Europe
• Real Name: Michael David Murphy
wemrick1 wrote:Ok, so this has been bugging me for awhile now.

I have one light source. I perform an incident meter reading at the subject which is 10 feet from the light source. My camera is set 10 feet from the subject. I set my camera per the incident measurement but the light has to travel another 10 foot to get to my camera and should only be 1/4 the intensity of the light at the subject. The meter has no idea how far my camera is from the subject. How can I possibly get a good exposure?

I used to ponder this mystery and came to the following conclusion. The inverse square law only applies strictly for point light sources. The light reflecting from your subject will most likely not act as a point source so the law does not apply and the fall off in light intensity for the now diffuse source will be significantly less than for a subject acting as a point source (maybe a small mirror).

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
mickspixels wrote:
wemrick1 wrote:Ok, so this has been bugging me for awhile now.

I have one light source. I perform an incident meter reading at the subject which is 10 feet from the light source. My camera is set 10 feet from the subject. I set my camera per the incident measurement but the light has to travel another 10 foot to get to my camera and should only be 1/4 the intensity of the light at the subject. The meter has no idea how far my camera is from the subject. How can I possibly get a good exposure?

I used to ponder this mystery and came to the following conclusion. The inverse square law only applies strictly for point light sources. The light reflecting from your subject will most likely not act as a point source so the law does not apply and the fall off in light intensity for the now diffuse source will be significantly less than for a subject acting as a point source (maybe a small mirror).

Just got done testing this using a C200 waveform and BMPCC 6K Pro against the Sekonic. Both cameras measured light loss as compared to the incident reading. The loss was not exact per the inverse square law but significant. The cameras also differed a full stop which was confirmed on the waveform in DR. One camera registered middle grey at 45.5 and the other at 55.5. Watching YouBoob and reading GooGoo I see a lot of claims of “perfect exposure”. I’m thinking this is fantasy. The exercise does give insight into the tolerances I can expect. Next up will be measuring anticipating noise.

John Paines

• Posts: 5252
• Joined: Tue Jul 28, 2015 4:04 pm
wemrick1 wrote:Just got done testing this using a C200 waveform and BMPCC 6K Pro against the Sekonic. Both cameras measured light loss as compared to the incident reading. The loss was not exact per the inverse square law but significant. The cameras also differed a full stop which was confirmed on the waveform in DR. One camera registered middle grey at 45.5 and the other at 55.5. Watching YouBoob and reading GooGoo I see a lot of claims of “perfect exposure”. I’m thinking this is fantasy. The exercise does give insight into the tolerances I can expect. Next up will be measuring anticipating noise.

Unclear here what or how you're testing, but for purposes of exposure, the light hitting an object doesn't diminish with distance. That's why people take incident light readings (they're reliable). The surrounding exposure may change as the shot gets wider, but the object's exposure should be the same. Discrepancies between the meter and the cameras is a completely different issue. Those discrepancies are to be expected. That's why you typically calibrate the meter. And middle-grey can be different values for different cameras. For BMD cameras it's usually around 38 ire.

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
John Paines wrote:
wemrick1 wrote:Just got done testing this using a C200 waveform and BMPCC 6K Pro against the Sekonic. Both cameras measured light loss as compared to the incident reading. The loss was not exact per the inverse square law but significant. The cameras also differed a full stop which was confirmed on the waveform in DR. One camera registered middle grey at 45.5 and the other at 55.5. Watching YouBoob and reading GooGoo I see a lot of claims of “perfect exposure”. I’m thinking this is fantasy. The exercise does give insight into the tolerances I can expect. Next up will be measuring anticipating noise.

Unclear here what or how you're testing, but for purposes of exposure, the light hitting an object doesn't diminish with distance. That's why people take incident light readings (they're reliable). The surrounding exposure may change as the shot gets wider, but the object's exposure should be the same. Discrepancies between the meter and the cameras is a completely different issue. Those discrepancies are to be expected. That's why you typically calibrate the meter. And middle-grey can be different values for different cameras. For BMD cameras it's usually around 38 ire.

It’s just a lot for me to wrap my head around. Want to see what exposures put middle grey at 50% ire in Davinci Resolve. I’ve read that 38% is BM middle grey a 2/3 stop beneath the rest of the world. Not sure what that buys me. Measuring for 50% I get up to 1.33 stops difference between the meters. I have a spot meter coming tomorrow and I’ll continue to test.

John Brawley

• Posts: 3857
• Joined: Tue Aug 21, 2012 7:57 am
• Location: Los Angeles California
Mid grey in log with BMD is 38.4%

When metering a grey card there is technique that can vary the reading by more than a stop.

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Finely found the answer. Took a lot of searching. From the light source the photon never changes intensity with distance however the photons will disperse and the inverse square law is a function of how many photons hit the target. Further away less photons. An illuminated surface will not act in the same manner. So distance from subject will not affect exposure. Now I need to make a bunch of profiles for my Sekonic to match my cameras. What a pain, lol. More testing, sheesh!!

John Griffin

• Posts: 1289
• Joined: Sun Aug 23, 2015 4:18 pm
Waveforms, histograms, Zebras are all much more accurate ways of exposing digital sensors than incident light meters.

Howard Roll

• Posts: 1965
• Joined: Fri Jun 03, 2016 7:50 am
wemrick1 wrote:I’ve read that 38% is BM middle grey a 2/3 stop beneath the rest of the world.

Rest of the world?

Rec709 is 50%.
BM Film, DJI D Log, and Arri log C are 38%.
Sony S Log, Canon C Log, and Red Log3G10 sit around 33%.

The positions of mid grey (18% reflectance), and white (90%) are going to depend on the gamma curve of the space. The majority of log curves are going to put mid grey below 40%.

Baseline exposure is 25 foot-candles at f1.0 when exposure time and ISO are reciprocal. This is camera and colorspace agnostic.

Good Luck

Steve Fishwick

• Posts: 986
• Joined: Wed Mar 11, 2015 11:35 am
• Location: United Kingdom
John Griffin wrote:Waveforms, histograms, Zebras are all much more accurate ways of exposing digital sensors than incident light meters.

Whilst that may be true to some extent, the point of an incident light meter is about lighting ratios, not from camera reflective readings, which all the above are still different methods of. Although very rarely used now they still can have their place in lighting sets.

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Howard Roll wrote:
wemrick1 wrote:I’ve read that 38% is BM middle grey a 2/3 stop beneath the rest of the world.

Rest of the world?

Rec709 is 50%.
BM Film, DJI D Log, and Arri log C are 38%.
Sony S Log, Canon C Log, and Red Log3G10 sit around 33%.

The positions of mid grey (18% reflectance), and white (90%) are going to depend on the gamma curve of the space. The majority of log curves are going to put mid grey below 40%.

Baseline exposure is 25 foot-candles at f1.0 when exposure time and ISO are reciprocal. This is camera and colorspace agnostic.

Good Luck

Thank you.

Per the experimenting I did last night, I can measure MG (middle grey) with the Canon C200 at 40% waveform and the BMPCC6K Pro at grey between green and pink and they will land in Davinci Resolve at 50% on the color waveform. I can then set the Sekonic to compensate + .5 stop and also hit that DR MG 50% with both cameras. I don't understand the benefit of lowering my exposure to the camera specs.

John Griffin

• Posts: 1289
• Joined: Sun Aug 23, 2015 4:18 pm
@wemrick1 - you are making this way too complicated. There is no perfect exposure that fits all cameras and all circumstances that can be pinned to mid grey.

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
John Griffin wrote:@wemrick1 - you are making this way too complicated. There is no perfect exposure that fits all cameras and all circumstances that can be pinned to mid grey.

Thank you John. I know you are right. It's how my head works. I often have to go the long way around to get the sense that I understand something. I have learned a lot from this thread and this exercise.

Leon Benzakein

• Posts: 1275
• Joined: Thu Aug 15, 2013 3:40 pm
wemrick1 wrote:It's how my head works.

I know how you feel.
Cheers!
Television: Lighting/Cameraman, O.B. Camera Operator, Experience in EFP, EPG and ENG , Grip, Lamp Operator
Film: Grip, Lamp Operator
Theater: Lighting Designer, Light board Operator, Stage Electrician, Stage Management

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Leon Benzakein wrote:
wemrick1 wrote:It's how my head works.

I know how you feel.
Cheers!

Good to know I'm in good company, lol.

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
wemrick1 wrote:
Leon Benzakein wrote:
wemrick1 wrote:It's how my head works.

I know how you feel.
Cheers!

.

Good to know I'm in good company, Cheers!

John Paines

• Posts: 5252
• Joined: Tue Jul 28, 2015 4:04 pm
wemrick1 wrote: I can then set the Sekonic to compensate + .5 stop and also hit that DR MG 50% with both cameras. I don't understand the benefit of lowering my exposure to the camera specs.

Because at the native ISO value, exposing middle-grey at the recommended value (38.4%) will give you maximum dynamic range. If you expose middle-gray at 50% on a BMD camera, you're losing a stop or more in the highlights. You do gain a stop in the shadows, but that's usually not where you want it.

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
John Paines wrote:
wemrick1 wrote: I can then set the Sekonic to compensate + .5 stop and also hit that DR MG 50% with both cameras. I don't understand the benefit of lowering my exposure to the camera specs.

Because at the native ISO value, exposing middle-grey at the recommended value (38.4%) will give you maximum dynamic range. If you expose middle-gray at 50% on a BMD camera, you're losing a stop or more in the highlights. You do gain a stop in the shadows, but that's usually not where you want it.

That makes sense.

Maurizio_Fenici

• Posts: 18
• Joined: Thu Mar 12, 2020 10:20 am
• Real Name: Maurizio Fenici
Hi to all,
in this video is explained similar problem:

Maurizio

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Maurizio_Fenici wrote:Hi to all,
in this video is explained similar problem:

Maurizio

rNeil H

• Posts: 370
• Joined: Tue Jun 26, 2018 9:43 pm
• Real Name: R. Neil Haugen
You have stumbled into a nomenclature problem. As you are equating "middle gray" with a being 50% middle.

It was *never* "50%". Period. Ever.

To get a correct understanding, search for middle gray along with Ansel Adams and Zone System exposure.

And the comments about different cameras having differences with any meter is spot on. I had a 30+ career as a stills photographer, was a bit of a tech wonk. Studied meters versus film responses heavily. Hard data.

Batches of film had their own characteristics and you needed to establish exposure INDEX for any meter, with any specific batch of a film emulsion. These days, each camera/lens combination needs the same.

An incident meter pointing at your main light can give a good reading for your diffused highlights (brightest skin with tone) IF you have tested for that with a specific batch of film or these days, a specific camera/lens combination.

It can give you a good relative data for the contrast between light sources.

And false colors on camera or monitor screens, again when tested, can also give very good exposure and contrast data.

Sent from my SM-S908U using Tapatalk

John Brawley

• Posts: 3857
• Joined: Tue Aug 21, 2012 7:57 am
• Location: Los Angeles California
With digital imaging it’s also a bit misleading to think in terms of ISO and “mid points”

Typically most so called native ISO on digital cameras is really the manufacturers recommendation for maximum dynamic range.

This involves subjective judgement about noise and a subjective judgement about how to apply a curve to the sensor linear exposure….by placing a midpoint you’re also positioning how much headroom and shadow detail you are over and under.

There’s nothing stopping you from manipulating that in a raw workflow.

The simple idea of a Lightmeter that tells you the exposure is a pretty simplistic tool if that’s what you’re using to for.

Light meters are great for measuring a ratio though, or absolute numbers (in foot candles) to make lighting calculations or if you’re there before your crew and want to take a reading without using the camera.

False colour and other tools like zebra really are far more accurate than the light meter because they are specific to what the camera / sensor is actually seeing.

I’ve long felt that the best way to think of exposure on digital sensors is a kind of top down logic.

The point that’s common with all sensors no matter what brand or curve is the clipping point…at what point do they overexpose?

Unlike mid grey which is arbitrary depending on the curve…the clipping point is absolute. If you set both cameras to a clipping point, THEN measure DR down from there, you start getting meaningful results.

This why I try to do an exposure ramp in my own tests.

This is a light shining on a white surface. Using a REFLECTED meter (not an incident) i measure the stop differences as i move the reading closer to the source and put a piece of tape on each +1 stop.

Then on a waveform you will have a clipping point that’s fixed and making it easy to compare where and to also then use your metered in footcandle readings of different exposure areas.

Here’s a HDR test I did where I was trying to figure out how to set exposures based on extreme contrast. It’s a little illogical, but candles are really bright even in low light, and trying to maintain the exposure so the flame still has colour but being able to come up with a look for dailies was the challenge here. You’ll see the exposure ramp in the background on the first shot.

The mid point is irrelevant. I need to preserve the highlight / clipping and I need to be able to still read the actors face in a story suitable way. It doesn’t matter what the incident light meter says here.

(Also the little cube on the chart has a light trap so you also have an absolute black to shoot and look for on a waveform)

Huzzah!

JB
John Brawley ACS
Cinematographer
Currently - Los Angeles

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
John Brawley wrote:With digital imaging it’s also a bit misleading to think in terms of ISO and “mid points”

Typically most so called native ISO on digital cameras is really the manufacturers recommendation for maximum dynamic range.

This involves subjective judgement about noise and a subjective judgement about how to apply a curve to the sensor linear exposure….by placing a midpoint you’re also positioning how much headroom and shadow detail you are over and under.

There’s nothing stopping you from manipulating that in a raw workflow.

The simple idea of a Lightmeter that tells you the exposure is a pretty simplistic tool if that’s what you’re using to for.

Light meters are great for measuring a ratio though, or absolute numbers (in foot candles) to make lighting calculations or if you’re there before your crew and want to take a reading without using the camera.

False colour and other tools like zebra really are far more accurate than the light meter because they are specific to what the camera / sensor is actually seeing.

I’ve long felt that the best way to think of exposure on digital sensors is a kind of top down logic.

The point that’s common with all sensors no matter what brand or curve is the clipping point…at what point do they overexpose?

Unlike mid grey which is arbitrary depending on the curve…the clipping point is absolute. If you set both cameras to a clipping point, THEN measure DR down from there, you start getting meaningful results.

This why I try to do an exposure ramp in my own tests.

This is a light shining on a white surface. Using a REFLECTED meter (not an incident) i measure the stop differences as i move the reading closer to the source and put a piece of tape on each +1 stop.

Then on a waveform you will have a clipping point that’s fixed and making it easy to compare where and to also then use your metered in footcandle readings of different exposure areas.

Here’s a HDR test I did where I was trying to figure out how to set exposures based on extreme contrast. It’s a little illogical, but candles are really bright even in low light, and trying to maintain the exposure so the flame still has colour but being able to come up with a look for dailies was the challenge here. You’ll see the exposure ramp in the background on the first shot.

The mid point is irrelevant. I need to preserve the highlight / clipping and I need to be able to still read the actors face in a story suitable way. It doesn’t matter what the incident light meter says here.

(Also the little cube on the chart has a light trap so you also have an absolute black to shoot and look for on a waveform)

Huzzah!

JB

Hi John,

I did get the answer that I was looking for that being light does not diminish by the reverse square law from an illuminated object. I am in agreement with you regarding the use of the handheld meter. Beyond understanding light ratios I am not finding any great use for it in setting up an exposure. The metering systems on the cameras that I have do a fine job of helping me decide on exposure.

Jeffrey D Mathias

• Posts: 363
• Joined: Fri Apr 28, 2017 3:54 pm
Actually the inverse square law assumes a point source radiating evenly without external forces. Anything we do as lighting only can approximate that, sometimes better than as others. Important considerations in astrophysics. But for setting a camera exposure false color is a great tool. And as John indicates, the clipping point is much better to work from. This becomes critical in setting exposure for HDR. I am frequently disappointed by the many professionally done films in HDR or Dolby Vision. Is it the DoP, the Colorist, the Director... I don't know, but it seems fewer than not get it looking right.
AMD Threadripper 1950x 16-core 3.4 GHz
96 GB Crucial DDR4 2666 ECC UDIMM RAM
AsRock Fatal1ty x399 motherboard
AMD Radeon Pro WX 8200 GPU
Windows 10 Pro 64-bit version 21H2, build 19044.1889

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
Jeffrey D Mathias wrote:Actually the inverse square law assumes a point source radiating evenly without external forces. Anything we do as lighting only can approximate that, sometimes better than as others. Important considerations in astrophysics. But for setting a camera exposure false color is a great tool. And as John indicates, the clipping point is much better to work from. This becomes critical in setting exposure for HDR. I am frequently disappointed by the many professionally done films in HDR or Dolby Vision. Is it the DoP, the Colorist, the Director... I don't know, but it seems fewer than not get it looking right.

Steve Fishwick

• Posts: 986
• Joined: Wed Mar 11, 2015 11:35 am
• Location: United Kingdom
Jeffrey D Mathias wrote:But for setting a camera exposure false color is a great tool. And as John indicates, the clipping point is much better to work from.

If the 'clipping point' is the most important thing then surely false colours are not essential? I have never used false colours, can't look through that, but I am no DP - I have always used, due to my background zebras, which are essentially just about clipping or not.

John Paines

• Posts: 5252
• Joined: Tue Jul 28, 2015 4:04 pm
False color displays every value in the shot from 0% to 100%, with the ratios graphically represented, by color (including the clipping point). It's a fantastic tool, far more useful than zebras. And unlike an external meter, it tells you exactly what the sensor will record and at what level. No guessing or estimating. And all at once, for everything in the shot.

wemrick1

• Posts: 480
• Joined: Sun Dec 05, 2021 12:47 pm
• Location: United States
• Real Name: Walter Emrick
John Paines wrote:False color displays every value in the shot from 0% to 100%, with the ratios graphically represented, by color (including the clipping point). It's a fantastic tool, far more useful than zebras. And unlike an external meter, it tells you exactly what the sensor will record and at what level. No guessing or estimating. And all at once, for everything in the shot.

Can you accurately estimate where noise will occur? I would estimate at ISO 400 20% would be the start of noise. At 800 ISO 30%? At 1200 (lowest in second ISO range) back to 20%? I would love to have a handle on that.