0

I travel constantly and visit lots of places where flash is not permitted (like the inside of churches). These places tend to be rather dark. I always shoot on manual. When I try to adjust the exposure my light meter and my histogram often disagree by almost a full stop. I have tried changing metering methods from average to center weighted but that makes little difference. Spot metering really won't help.

My impression is that the histogram in these situations is more accurate but when I crank up the exposure I tend to get small pieces of the photo that are over-exposed.

Any thoughts or help would be appreciated. Thanks!

  • 5
    What do you mean by "When I try to adjust the exposure my light meter and my histogram often disagree by almost a full stop."? A light meter is not showing you a histogram and a histogram is not the same as a light meter. They can't agree because they don't show the same thing. – user1118321 Jan 10 '18 at 6:29
  • 1
    is your lightmeter an extra device or the camera integrated light measuring sensor? btw. like user 1118321 saied: the values of the light meter (independently from kind of device) and the histogram (which is a graph/diagram) shows different things. – Horitsu Jan 10 '18 at 8:00
  • Please ask a specific question rather than just asking for "any thoughts or help." – Caleb Jan 10 '18 at 16:27
1

There are a couple of factors, but there is more to it. To use them, we need to know how light meters work, and how histograms work.

A histogram is not a light meter. A histogram shows the values how the reflected colors come out (NOT how they ought to come out). Colors do not reflect equally. A white polar bear in the snow in bright sun should be pretty bright, and near the right end of the histogram. A black cat in a coal mine should be pretty dark and at the left end. That is how accurate pictures should come out. Arbitrarily shifting all photos to the right end without concern how the colors ought to come out will overexpose some things.

Reflected light meters try to put everything (the average value of everything) in the middle. If you photograph three scenes which are 100% bright white, 100% dark black, and 100% middle gray, the light meter will try to put all three at about the middle. So that's not very precise either. White should come out high, black should come out low, and middle gray near the middle.

And camera histograms are of two types. One is a single gray histogram, which is a math manipulation to show how black&white film would reproduce the colors (called Luminance). It is very inaccurate regarding how the colors are reproduced, and it should be ignored in the camera. Another type shows three RGB histograms, which is how the colors actually come out, and should be the only one you look at. It is accurate (about how colors are reproduced, but does not know how they should be reproduced). See http://www.scantips.com/lights/histograms.html

| improve this answer | |
  • The on-camera histogram is often rather inaccurate about how colors are reproduced. In particular, at least on many cameras, the histogram is collected from the JPEG preview, even when (for example) you're shooting in RAW format. In such a case, the histogram can show what look like problems, even though the data in the raw file is fine. – Jerry Coffin Jan 11 '18 at 1:34
  • There are no camera settings added in the raw data (no white balance, no Vivid, etc. Raw is raw), but the JPG preview and histogram has camera settings in it. This may sort of assume you will set similar values in the raw later, but yes, you can set them different. – WayneF Jan 11 '18 at 4:30
  • The big differences aren't from adjustments as such, but (for one example) from the fact that the raw data typically has at least 12 (and sometimes 14) bits of data per channel, where jpeg has only 8 bits per channel. The jpeg can show clipping of both highlights and shadows that simply doesn't exist in the raw data. Oh, and yes, the camera settings often are stored in the raw file. They don't change the raw data itself, but the manufacturer's proprietary software will read it and initially process the raw data accordingly. – Jerry Coffin Jan 11 '18 at 14:54
  • 8-bits is about precision, not accuracy. Cameras do WB in 12 bits, and then convert to 8-bit JPG. That's fine if the camera WB setting was accurate, but for other than maybe direct sunlight, it simply can't be known accurately. The 8-bit conversion is irreversible, so subsequent large JPG changes are a problem. Raw 12 bit raw data has greater range and accuracy, especially since we can actually see it first. But any WB clipping will be the same in JPG or raw if the same WB shift is performed. – WayneF Jan 11 '18 at 18:59
  • We disagree in principle. Anyone willing to settle for camera WB settings (including Exif) likely doesn't need to shoot raw. :) We often can't really know what WB to set in the camera. Auto WB is not often accurate, but it seems good to use with raw, because it might be at least ball park (better than wrong settings), letting the camera JPG histogram better match the final raw settings we likely will choose when we can accurately set it. – WayneF Jan 11 '18 at 19:00
0

Light meter make some kind of average of the histogram as the exposure is determined by the total amount of light available (see https://www.iso.org/standard/62742.html). The meter's proposed exposure is a one fits all and (althrough some brand may apply some correction) it may be off in some situation. By using a manual setting you may accept to blow some pixels in order to have a more luminous scene, what you describe is part of the Exposing to the right process (ETTR see https://en.wikipedia.org/wiki/Exposing_to_the_right) but it supose that you rework your image from raw afterwards.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.