The Villages Loc: The Villages, Florida

How important is the metering pixel count with digital cameras? Not talking about Megapixel numbers.

For example, the D500 claims a Metering Pixel count of 180,000, whereas other cameras may list 1,000 or 2,000. Do higher count numbers mean significantly better resulting exposures due to more reflective readings? I would think so, but don't really know.

Your insight would be appreciated.

BebuLamar

I doubt the benefit of using a lot of pixels for the meter. Although I do not have a camera with more than 2000 pixels to test so I can't be sure but I doubt there is any significant improvement.

rmalarz Loc: Tempe, Arizona

Well, pixel count is important enough that Google doesn't count characters in a page title but instead counts pixels.
http://www.webshoptimizer.com/page-title-pixel-meter/
--Bob

Rongnongno Loc: FL

The greater the number the more generic the metering, not the more accurate.

The more generic gives the general public a good option in some way it is similar to measuring the incident light vs the reflective light.

The more accurate is spot metering, a few pixels.

Apaflo Loc: Anchorage, Alaska

The only difference is when using Matrix metering. That mode compares patterns from the RGB metering sensor to a database of patterns in order to determine how a scene should be evaluated.

The higher pixel count results in a larger database being useful. A large sensor has no use for other than Matrix metering. If you do not use Mattix metering the size of the metering sensor is insignificant for your purposes.

The Villages Loc: The Villages, Florida

Apaflo - So if I understand correctly, a sensor with a high metering pixel count would provide better results when used in a Matrix mode? One using Matrix for their metering will get better exposed pictures if their camera has a higher count?

Apaflo Loc: Anchorage, Alaska

That is essentially the value of the larger sensor.

It is also worth noting that Matrix metering is "tempermental", and may give you several different readings for a scene that you think is the same every time! It is automation for beginners, and not recommended for anyone past the total auto stage.

The Villages Loc: The Villages, Florida

Apaflo - "not recommended for anyone past the auto stage"

Even if not in an "auto" stage, doesn't the metering system come into play when determining exposure......unless just guessing at the settings? i.e. manually zeroing in on the meter scale?

CO

Nikon expert, Ken Rockwell, wrote something about it. His assessment was it's not the number of exposure metering pixels that matter. It's how well it's programmed that matters. My Nikon D90 had 420 points. My D500 has 180,000. It uses them for autofocus, auto exposure, and auto white balance. When the camera is in 3D tracking mode it's also using those points.

Apaflo Loc: Anchorage, Alaska

If everything you do is as automatic as it can be done, then Matrix metering might be what works best most often. Not always though...

If you want to control exposure yourself, even just partially by setting one parameter and letting the camera do the rest it is best to totally avoid Matrix metering. It is impossible to control Matrix metering in any way. Trying to use Exposure Compensation with Matrix metering is mostly just frustrating. You cannot tell what type of pattern will be matched! It may change with each exposure. Or it may not.

Apaflo Loc: Anchorage, Alaska

Ken Rockwell is not a Nikon expert, as your statement shows.

The programming available with a 180k sensor is vastly more complex than what could be done with a sensor that had only 420 points to work with. Tracking a subject or recognizing a face are obvious examples were a 420 point sensor can't track more than some percentage of those points, but with a 180k sensor the tracking can examine 10 or 20 times as many points as the 420 point sensor has in total!

CO

Ken Rockwell has a lot of knowledge when it comes to Nikon cameras.

I wasn't saying that the 420 point and 180,000 point metering were equally good. I was just mentioning it to illustrate how much further cameras have come. I'm very impressed with the metering of my D500. It's no doubt better than the metering of my D90 was. I think a lot of it does come down to how good the programming is. For example, Nikon's flash metering has been known to be about the best around.

Apaflo Loc: Anchorage, Alaska

Rockwell spews massive amounts of garbage to. It may be that what you attributed to him is not what he said, but it has the same ring to it.

Nikon can only provide better programming if they also provide the better sensor. Adding the same programming to a D90 would accomplish nothing.

CO

What I attributed to him is not what he said? What kind of nonsense is that? I'm not intelligent enough to differentiate? I have a bachelor's degree in mechanical engineering and program 4-axis computer numerically controlled machine tools. Stop with this. Every one of your posts are this way.

I already said the 180,000 point metering of my D500 is superior to the 420 point metering of my D90.

selmslie Loc: Fernandina Beach, FL, USA

Sounds like the pot calling the kettle black. There is nobody that is right 100% of the time, least of all you. However, you certainly have him beat when it comes to spewing incorrect information. He has you beat in presenting useful information that is usually correct.

There is no metering system that is 100% reliable so that the photographer does not have to think. Each method - spot, center weighted, matrix, external reflected, incident and EV table based like sunny 16 - requires the photographer to understand how to use it and when it might fail.

If you can't make matrix metering work then the fault is with you, not with the method.

There is no database in the camera. There is an algorithm that was built using the 30,000 images and a group of professional photographers and their collective opinion regarding the proper exposure for each case.

The whole database and the empirically determined exposures are run through a neural network or a statistical analysis (not sure which one Nikon uses - it's a trade secret) to come up with an optimal algorithm to be programmed into the camera.

Using more pixels in the meter just makes the algorithm a little more accurate but the end result is still a simple exposure value (aperture and shutter speed) for the ISO.

Whenever a new camera is created with a different pattern of pixels to be used in Matrix metering, the analysis is repeated to produce a new algorithm. The formulas used in the algorithm probably do not change (much). They just get their constants adjusted a little.

You may not know the difference between an algorithm and an alligator but either one can bite you in the ass if you don't pay attention.