Do you use a light meter?

frugivore wrote in post #17085581
My definition of a perfect exposure is having the brightest part of the scene captured at just under the maximum RGB value (e.g. 4096 in a 12-bit RAW). If the exposure is greater or less than this, I have less fidelity of my scene. And how can a light meter tell me where that brightest part of a scene is if it is measuring the whole scene, instead of various zones of the scene as the camera does? It cannot. The best it can do is figure out the EV of that one scene and hope that the dynamic range is not too small or not too large so that when captured to a digital sensor, neither the highlights get clipped, nor the exposure too low so that the more accurate bits go unused.

Blue text is unfortunately overly simplistic definition that ASSUMES that 'the entire range of brightness in a scene can FIT within the dynamic range of film/sensor'. So what happens when the DR of the scene is 10-11EV and only 8-9EV will FIT?!

• Then the photographer's brain (we are back to the need to THINK...metering is only a guideline!) decides if the highlight details are more important, or if the shadow details are more important.
  • Perhaps the brightest details are the sun's own reflection in the chrome bumper...do you even CARE to capture an image of the sun in the chrome of the bumper?!
    
  • Perhaps the shadow details matter far more than wanted details in the brightest part of the scene.

frugivore wrote in post #17085581
Here's a simple test. Take a plain, uniformly lit backdrop (e.g. white, gray or black seamless paper) and measure it with a flash meter. Take a picture of that backdrop and you should end up with any of them looking to be gray, just like if you were to use just your camera's autoexposure meter with spot metering.

Interesting.

White and black background, incident flash metering:

IMAGE LINK: https://flic.kr/p/onmy​Cw

IMAGE LINK: https://flic.kr/p/oDR1​Xa

IMAGE LINK: https://flic.kr/p/onmh​1L

IMAGE LINK: https://flic.kr/p/oBPh​8s

frugivore wrote in post #17085581
And I trust a camera's meter to determine which parts of the scene are the brightest more than I trust my eyes. Again, I can understand the use of a light meter when using studio lights, but otherwise I think the camera does a better job.


Not infinite, but much greater. Have a look at his comparison and tell me that you don't see film retain more detail than digital:

http://photo.stackexch​ange.com …ad-in-digital-photography


Here's a simple test. Take a plain, uniformly lit backdrop (e.g. white, gray or black seamless paper) and measure it with a flash meter. Take a picture of that backdrop and you should end up with any of them looking to be gray, just like if you were to use just your camera's autoexposure meter with spot metering.

Okay, from the first paragraph, you really need to start trusting your eyes and thinking about stuff before you start snapping away. The whole adage about "a great photographer with the worst camera takes a better picture than the average photographer with the best camera" kinda means that you have to start trusting your eyes instead of letting the camera do everything for you and not thinking about it. Just look at the bright scenes, and think to yourself, "Is this scene brighter than 50% grey or darker than 50% grey. If it's brighter, you overexpose. If it's darker your underexpose. You can't just rely on your camera's reflecting light meter do know to do that for you, because it isn't smart enough so it won't.

About the link you sent. I think that's a poor comparison because it's using old technology and things are different now than they were in 2006 when that 6 Megapixel camera was released. That camera isn't 14-bit.

In your last paragraph. I don't even need to do that simple test, because I know that's not what is going to happen. What is going to happen is what I said before is going to happen: black will look like it's black and white will look like it's white. I know you're talking about this without ever having used an incident light meter before. Because if you have, you would never have said something like you did in that last paragraph.

oldvultureface wrote in post #17085858
Interesting.

White and black background, incident flash metering:

IMAGE LINK: https://flic.kr/p/onmy​Cw

IMAGE LINK: https://flic.kr/p/oDR1​Xa

In-camera reflective (E-TTL) metering:

IMAGE LINK: https://flic.kr/p/onmh​1L

IMAGE LINK: https://flic.kr/p/oBPh​8s

My mistake. I was thinking about that meter that can go from incident to reflected (one of the Sekonics where you twist the dome?) and I was thinking about that reflected metering mode. Yes, when measuring incident light, the readings will be the same regardless of the background. My point is that part of the scene can be so reflective whereby, when using an incident light meter to determine the exposure, that part of the scene would get clipped in the camera. Your white backdrop may not have been, but some other more reflective surfaces may. Now with the black backdrop, that might represent some object that was dark and, instead of capturing it in the higher bits, it gets captured in the lower bits where there is more noise and less accuracy.

It seems that the assumption is that once a file is captured, it should not have its exposure adjusted in post. But like gjl711 said, that's one of the strengths of digital. And, IMO, a necessity to get great fidelity of the details in the scene.

frugivore wrote in post #17086022
It seems that the assumption is that once a file is captured, it should not have its exposure adjusted in post.

Who's assumption?

Sure, I'll use a meter to get everything as close to "correct" in camera, but I have no problem moving the Exposure slider in post - that's what it's there for. Understand that I'm talking about minor adjustments at this point though.

Personally I see very little reason to not try to have your general exposure within a range well suited for the subject and lighting you are working with. My view is that if you're trying to adjust exposure more than a partial stop in post then you have missed judged your exposure. (I miss exposure and often rely on trying to recover in post, but I view that as trying to recover from an error, not as an overly acceptable workflow plan one should expect.)

Exposure settings are something you set according to how you plan to process the image, and there is only a narrow range that would be correct for a given amount of light.

Using a handheld meter can mean you can measure the light faster and more accurately than with an in-camera meter.

frugivore wrote in post #17086022
My mistake.

It seems that the assumption is that once a file is captured, it should not have its exposure adjusted in post. But like gjl711 said, that's one of the strengths of digital. And, IMO, a necessity to get great fidelity of the details in the scene.

Who gave you that assumption? If you ever worked with film, then you would know you developed the film for your subject, sometimes for the shadows and sometimes for the highlights, depending on the film you were using. And after you developed for the subject, you would take your negative or positive and you'd expose your paper for the subject under the enlarger. You'd also still dodge and burn, just like you do in Photoshop today. Not a lot has changed in terms of what you're trying to get towards, except you can do all the aforementioned steps much faster now and you can use sliders instead of shaking film in chemicals for a certain amount of time or flipping a light on for a specified amount of time that you calculate.

But guess what! You still need to expose your shot correctly in the camera. And incident and reflectant light meter strengths and weaknesses haven't changed!!!

frugivore wrote in post #17086022
My mistake. I was thinking about that meter that can go from incident to reflected (one of the Sekonics where you twist the dome?) and I was thinking about that reflected metering mode. Yes, when measuring incident light, the readings will be the same regardless of the background. My point is that part of the scene can be so reflective whereby, when using an incident light meter to determine the exposure, that part of the scene would get clipped in the camera. Your white backdrop may not have been, but some other more reflective surfaces may. Now with the black backdrop, that might represent some object that was dark and, instead of capturing it in the higher bits, it gets captured in the lower bits where there is more noise and less accuracy.

It seems that the assumption is that once a file is captured, it should not have its exposure adjusted in post. But like gjl711 said, that's one of the strengths of digital. And, IMO, a necessity to get great fidelity of the details in the scene.

First, it's only your assumption, not anyone else'. Second, they're called highlights for a reason. You either choose to expose for them or your don't. If you use a proper exposure for the scene then your highlights and shadows will show up exactly where they should. In Evaluative mode, the cameras meter is simply trying to make the entire scene average out to mid gray, it's not doing any calculations whatsoever that have any bearing on the DR of the cameras sensor. You'll still get plenty of blown highlights relying on the cameras reflective metering and Evaluative mode, it just depends on how much of the scene is composed of those highlights and their brightness relative to the rest of the scene.

Scatterbrained wrote in post #17086284
First, it's only your assumption, not anyone else'. Second, they're called highlights for a reason. You either choose to expose for them or your don't. If you use a proper exposure for the scene then your highlights and shadows will show up exactly where they should. In Evaluative mode, the cameras meter is simply trying to make the entire scene average out to mid gray, it's not doing any calculations whatsoever that have any bearing on the DR of the cameras sensor. You'll still get plenty of blown highlights relying on the cameras reflective metering and Evaluative mode, it just depends on how much of the scene is composed of those highlights and their brightness relative to the rest of the scene.

This is exactly correct.

Frugivore, try an incident light meter just once, and report back to us after you've discredited all the assumptions you have about the equipment and you've seen for yourself how it works.

drvnbysound wrote in post #17086074
Who's assumption?

Sure, I'll use a meter to get everything as close to "correct" in camera, but I have no problem moving the Exposure slider in post - that's what it's there for. Understand that I'm talking about minor adjustments at this point though.

The assumption is built into the design of a light meter. Yes, you might make minor adjustments. But contrast this to how I would create an image of a dark scene (e.g. black cat on a dark rug). I would make an exposure that would put the brightest part of the scene (which to us would be considered "dark") at the top-most bits of the RAW file. I'd look at it in post and it would appear very bright - almost white. Then, I'd lower the exposure by about 4 stops.

If I had used the incident meter's exposure values, I would not need not do this extra step, but it helps increase the fidelity as I mentioned earlier.

CRCchemist wrote in post #17086188
Who gave you that assumption? If you ever worked with film, then you would know you developed the film for your subject, sometimes for the shadows and sometimes for the highlights, depending on the film you were using. And after you developed for the subject, you would take your negative or positive and you'd expose your paper for the subject under the enlarger. You'd also still dodge and burn, just like you do in Photoshop today. Not a lot has changed in terms of what you're trying to get towards, except you can do all the aforementioned steps much faster now and you can use sliders instead of shaking film in chemicals for a certain amount of time or flipping a light on for a specified amount of time that you calculate.

But guess what! You still need to expose your shot correctly in the camera. And incident and reflectant light meter strengths and weaknesses haven't changed!!!

Digital capture is different than analog (film) despite the fact that we perform similar processes on them. With digital, capturing the signal as clearly as possible is very important. Exposing for the highlights does this. And an incident meter is not designed to do this. A camera's reflected light meter with zones can do this.

Scatterbrained wrote in post #17086284
First, it's only your assumption, not anyone else'. Second, they're called highlights for a reason. You either choose to expose for them or your don't. If you use a proper exposure for the scene then your highlights and shadows will show up exactly where they should. In Evaluative mode, the cameras meter is simply trying to make the entire scene average out to mid gray, it's not doing any calculations whatsoever that have any bearing on the DR of the cameras sensor. You'll still get plenty of blown highlights relying on the cameras reflective metering and Evaluative mode, it just depends on how much of the scene is composed of those highlights and their brightness relative to the rest of the scene.

That's not due to the meter, but how Canon uses the values. I believe the new Nikon D810 had a metering mode that preserves highlights. And that means it's based on the cameras sensor, not some middle great value.

CRCchemist wrote in post #17086440
This is exactly correct.

Frugivore, try an incident light meter just once, and report back to us after you've discredited all the assumptions you have about the equipment and you've seen for yourself how it works.

I have used one. And I understand how it works. It is very useful when used with manual flashes/strobes, but I think that with just ambient light photos, a camera's meter will do a better job, especially when Canon implements an ETTR option for their auto exposure.

frugivore wrote in post #17087286
Digital capture is different than analog (film) despite the fact that we perform similar processes on them. With digital, capturing the signal as clearly as possible is very important. Exposing for the highlights does this. And an incident meter is not designed to do this. A camera's reflected light meter with zones can do this.

Yes, this is almost a true statement. But one must understand that both meters try to achieve 'capture it a mid tone'...the incident meter reproduces only the midtone object at mid tone (in measuring the light striking the scene), while the reflected meter tries to take ANY THING IT SEES and make mid tone of it (which is why we get grayish snow when it is pointed at the snow, or a grayish black cat when it is pointed at something black!). It is ONLY the histogram (not the METER) which allows us to see 'highlights' (the brightest thing in the scene that you mentioned) as falling at the histogram's right edge...but it is NOT THE METER in the camera! The meter in the camera only thinks 'make this thing grey!!!'

frugivore wrote in post #17087286
That's not due to the meter, but how Canon uses the values. I believe the new Nikon D810 had a metering mode that preserves highlights. And that means it's based on the cameras sensor, not some middle great value.

And so does Canon, in their HTP (Highlight Tone Priority). The purpose of Highlight Tone Priority is to effectively increase the effective dynamic range of the camera, by giving you extra protection of the highlights, while keeping the overall exposure the same. In other words, if two identical pictures are taken at the same exposure, one with HTP on and one with it off, the picture taken with HTP on will have fewer blown out highlights.

But HTP doesn't actually increase the dynamic range of the camera. HTP retains extra detail in the highlights by underexposing the image: the image is actually captured at a lower ISO, and the darker areas of the image are digitally increased to bring their brightness to the same level as without HTP. When HTP is enabled, you cannot select an ISO lower than 200.

Nikon’s feature is called 'Active D Lighting'. Both companies have had this feature for about 7 years!!! And most folks find they like their results better when the automagic (ALO, HTP) are turned OFF. And, make a note of this, both can be manually accomplished in post processing...Nothing magic in camera is happening, as you can achieve the same result with any camera made prior to 2007, too, and using post processing adjustments.

frugivore wrote in post #17087286
...but I think that with just ambient light photos, a camera's meter will do a better job, especially when Canon implements an ETTR option for their auto exposure.

...when the object in brightest highlight truly IS what the photographer wishes to preserve -- and it not always of importance, as my 'sun reflection in the chrome bumper' is such an example, of it NOT being important!

I use mine all the time indoors on paid jobs and for outdoor portraits. It is better for me to take a reading and get it right rather than shoot a few test shots and making adjustments while your client is waiting and waiting. Why make them wait longer than necessary. Take a meter reading, set it and shoot.

But Canon's HTP will add noise in the underexposed areas so it is a trade off in quality.

Shooting wrote in post #17090215
But Canon's HTP will add noise in the underexposed areas so it is a trade off in quality.

Never use Highlight Tone Priority because it will hurt the other end of your histogram and you can do the exact same thing when you bring the CR2 file into your computer. It's a feature that is there for newbies.

Wilt wrote in post #17087349
Yes, this is almost a true statement. But one must understand that both meters try to achieve 'capture it a mid tone'...the incident meter reproduces only the midtone object at mid tone (in measuring the light striking the scene), while the reflected meter tries to take ANY THING IT SEES and make mid tone of it (which is why we get grayish snow when it is pointed at the snow, or a grayish black cat when it is pointed at something black!). It is ONLY the histogram (not the METER) which allows us to see 'highlights' (the brightest thing in the scene that you mentioned) as falling at the histogram's right edge...but it is NOT THE METER in the camera! The meter in the camera only thinks 'make this thing grey!!!'



And so does Canon, in their HTP (Highlight Tone Priority). The purpose of Highlight Tone Priority is to effectively increase the effective dynamic range of the camera, by giving you extra protection of the highlights, while keeping the overall exposure the same. In other words, if two identical pictures are taken at the same exposure, one with HTP on and one with it off, the picture taken with HTP on will have fewer blown out highlights.

But HTP doesn't actually increase the dynamic range of the camera. HTP retains extra detail in the highlights by underexposing the image: the image is actually captured at a lower ISO, and the darker areas of the image are digitally increased to bring their brightness to the same level as without HTP. When HTP is enabled, you cannot select an ISO lower than 200.

Nikon’s feature is called 'Active D Lighting'. Both companies have had this feature for about 7 years!!! And most folks find they like their results better when the automagic (ALO, HTP) are turned OFF. And, make a note of this, both can be manually accomplished in post processing...Nothing magic in camera is happening, as you can achieve the same result with any camera made prior to 2007, too, and using post processing adjustments.

I never use HTP myself and do this adjustment in post, and quite frequently with non-flash photos.

But I'm talking about the calculations that Canon/Nikon programmers make with the metered zones of the scene. Up until the D810, highlights were not really taken into account. I think there will be a shift now. Read about the new mode here:

http://nps.nikonimagin​g.com …ions/d810_tips/​highlight/

I am very excited about this.

I don't disagree with you. Specular highlight should be discarded in exposure calculations. A white wedding dress in bright light should not be.

But getting back to my point, a hand held light meter - whether it's used in incident mode or reflected mode - cannot take highlights into account. As such, it will always be at the mercy of the scene's dynamic range. In most cases, that isn't a problem.