Highlight Weighted Metering - When is it right? It can be counter productive if matrix metering does not cause your highlight warnings to flash. Highlight Weighted Metering (HWM) is...

selmslie Loc: Fernandina Beach, FL, USA

It can be counter productive if matrix metering does not cause your highlight warnings to flash.

Highlight Weighted Metering (HWM) is right in cases where matrix metering might fail, you see lots of blinkies. But you still need to watch for blinkies, they may be reduced without going away.

I include the RawDigger histograms because they show what happens in the raw channels.

Here is an image where the highlight warnings go away. But the change in exposure at base ISO is small enough to allow easy recovery of the shadows without bringing in any noise.

Matrix Metering - lots of blinkies in the window 1/13 sec

Highlight Weighted Metering - no blinkies 1/80 sec

In the following posts I will show where it works partially and where you don't want to use it at all.

selmslie Loc: Fernandina Beach, FL, USA

In this example the scene's DR is too wide. HWM reduces the blinkies but does not eliminate them. The highlight rendition is significantly improved but the highlights are still blown in the raw file.

Matrix Metering - 1/6 sec

HWM - 1/40 sec

A scene like this probably calls for additional reduction in exposure, maybe EC-1 or EC-2. But this might result in poor shadows and visible noise. Probably two or more combined with HDR processing would be better.

selmslie Loc: Fernandina Beach, FL, USA

Finally, here is a scene where Matrix Metering does not fail because it has a much lower dynamic range (DR).

Matrix Metering - 1/400 sec

HWM - 1/800 sec HWM reduced the exposure!

This example tells us two important things:

1. When the scene's DR is narrow we don't need HWM. Neither the highlights nor the shadows are in danger.
2. However Nikon implemented HWM, it's not a substitute for ETTR.

But what's even more significant is that it shows that ETTR at base ISO in a low DR scene like a landscape is not only unnecessary, it's a colossal waste of time and effort.

User ID

Neat charts. Gold star for neatness.

selmslie Loc: Fernandina Beach, FL, USA

User ID wrote:

Neat charts. Gold star for neatness.

Too much information

Ourspolair

My experience has been that ETTR does not seem to help when the DR is within the DR of the camera. Check your histograms often!

selmslie Loc: Fernandina Beach, FL, USA

Ourspolair wrote:

My experience has been that ETTR does not seem to help when the DR is within the DR of the camera. Check your histograms often!

I have seen that as well. So long as you don't make a gross error in the exposure setting you are probably not going to have an issue.

I think that's what the JPEG shooters have been trying to get across to the raw fanatics.

If you look at the numbers in the landscape image above you will see that Matrix Metering selected a Sunny 16 value (ISO 100 1/400 @ f/8). That's not an accident.

sandiegosteve Loc: San Diego, CA

By highlight weighted metering, do you mean expose to the right, or the mode in some cameras? I've got a highlight mode in two of my bodies that are great for certain situations, like when doing an event and they turn out the lights but I still want to get the subject in the spotlight.

For ETTR, one thing I recently learned is that the in camera JPEG mode does make a difference. "Standard" or "Normal" will drop the highlights and boost the shadows a little for the preview you see on the back of your camera. When I'd get back to my computer, it wouldn't ever be as "to the right" as I wanted. Then, I changed the in Camera setting to Flat/Neutral. The preview isn't as nice, but it is much closer to actual RAW histogram.

User ID

selmslie wrote:

Too much information

High infotainment quotient

camerapapi Loc: Miami, Fl.

An English gentleman by the name of Derek Forss, an Olympus Visionary and educator easily found in You Tube where he often shows his beautiful photography has referred to this in some of the videos I have watched. Because of the high dynamic range of many modern sensors in many cases it is very possible to expose for the highlights and compensate in post opening the shadows. Some exposures will require the help of a noise reduction program.

He proclaims himself "old school" and he uses spot metering in his photography. I use Olympus and Nikon and the latter does not have the HWM technique incorporated in my old cameras. I do use the spot meter often finding that I have more control on my exposures that way.
You have done a great job with your images and your graphics.

Strodav Loc: Houston, Tx

camerapapi wrote:

An English gentleman by the name of Derek Forss, a... (show quote)

Totally agree. I've managed to skirt a lot of tricky situations by dropping the ISO as the expense of shutter speed and/or aperture. Attached is the DR curve of my D850. There is a difference of about a whole stop of DR above ISO 100. Many photographers don't understand the importance of this curve for their cameras. OP, good topic.

(Download)

selmslie Loc: Fernandina Beach, FL, USA

sandiegosteve wrote:

By highlight weighted metering, do you mean expose... (show quote)

Yes, Highlight Weighted metering is designed to address a subject in the spotlight with the surrounding area dark. But you can't trust it on faith. You still have to watch for blinkies and lower the exposure compensation to minimize them or make them go away.

In Standard or Normal mode the camera's JPEG histograms are within 1/3 stop of the raw histograms shown in RawDigger.

The blinkies are linked to the raw data. Here is a simple test that anyone can use to demonstrate the basis for the blinkies.

selmslie Loc: Fernandina Beach, FL, USA

Strodav wrote:

Totally agree. I've managed to skirt a lot of tricky situations by dropping the ISO as the expense of shutter speed and/or aperture. Attached is the DR curve of my D850. There is a difference of about a whole stop of DR above ISO 100. Many photographers don't understand the importance of this curve for their cameras. OP, good topic.

The reason for the loss of about one stop of DR each time you double the ISO is that it also doubles the amplification of the noise.

cbtsam Loc: Monkton, MD

Strodav wrote:

Totally agree. I've managed to skirt a lot of tricky situations by dropping the ISO as the expense of shutter speed and/or aperture. Attached is the DR curve of my D850. There is a difference of about a whole stop of DR above ISO 100. Many photographers don't understand the importance of this curve for their cameras. OP, good topic.

I don't know about "old school," but I'm certainly old enough to have lost the use of a few neurons, and I don't understand what the graph you've attached says about dynamic range. To be clear, I'm not suggesting it doesn't illustrate something important about dynamic range, I'm merely asking for help in understanding what that is. I can see that one axis is Measured ISO, and the other is EV, but I don't understand how measured ISO - presumably by the exposure meter - as a function of EV tells me anything about DR. If you could help out an old man I'd appreciate it.

wdross Loc: Castle Rock, Colorado

camerapapi wrote:

An English gentleman by the name of Derek Forss, a... (show quote)

Olympus uses Exposure Compensation in their digital cameras as one of the normal variables in Program Mode. If used with the digital camera spot mode, it can be used to set up a zone system based off any single spot reading.

One of the very best spot metering systems was in the Olympus OM-4ti. It would allow eight 2.5° spot readings, averaged them, and set the exposure based on the average. And one could even bias that exposure with a highlight or shadow button. Or one could take a single spot of the highlight or shadow and press either the highlight or shadow bias buttons to offset the spot exposure a set amount.

I feel spot metering is almost a lost art these days. The algorithms have managed to be fairly smart about most photographic situations based off the various light levels and where they occur on the sensor.