EOS 5D MARK IV Metering Issue?

[gdonoghue20]

I get different exposure measures between the top lcd on the camera and when looking through the lens. Through the lens is the accurate measure. What would cause the top lcd to be off?

[TCampbell · Accepted solution]

This can be a lot of things... I'd be easier if I were watching you use the camera.

My first question is:  What metering mode are you using?

The 5D IV has four different metering modes... 

1. Evalutative
2. Partial
3. Spot
4. Center-weighted

Evaluative if the default -- it uses metering points spread across the frame.  But the other modes do not... spot tries to minimize the metering area to just the center.  That's for situations where you want accurate metering on a particular subject and you don't want any other particular brighter sources or particularly dark areas to throw off the metering.  The other modes (partial & center-weighted) either don't use the whole area or they bias the metering to favor the center area.

If you are on a tripod and you use one of these modes (so you can be sure the camera didn't move) then the metering should "in theory" be the same (except for something I'll mention in a moment) because if the camera was using spot-metering it should be pointed at the same subject whether you look through the viewfinder or just look down on the top-LCD.

If you're not using a tripod, then there's the question of "was the camera pointed at EXACTLY the same subject when you tested the meter readings) (and I used all-caps for a reason... treat this like a scientific test... there's "exact" or "not exact" and anything less than perfect isn't a valid test -- that's why I mention the tripod).  

You have to remember that this is not an "incident" light meter (like a hand-held light meter such as a Sekonic brand meter) the camera is measuring reflectivity off an object that the lens can see.  White subjects reflect more light.  Black subjects no so much.  So if the lens moves and isn't pointed at exactly the same subject, you'll get a different reading.  This would NOT happen with a hand-held incident meter. Those meters do not measure light reflecting off a subject... you hold those meters in front of the subject surface and measure the light falling on the subject.  It wouldn't matter if you hold the meter in front of a "white" piece of poster-board or a "black" piece of poster-board... the light "falling" on the board is the same so the incident meter will give you the same reading -- it isn't fooled by reflectivity.  A camera's built-in meter IS reflective and IS fooled by differences in subject reflectivity.

Next is whether or not the eye-cup is covered.  

The metering sensor is inside the viewfinder.   Light shines into the camera when the reflex mirror is down.  It bounces off the mirror and up onto the focus screen (the frosted screen on the top of the mirror chamber).  When you look through the viewfinder, you look through a pentaprism (which has mirror-like properties) and you see the image on the frosted focus screen.  But the METER is just above that screen.

So it turns out... if your eye isn't covering the viewfinder eye-cup, then extra light is coming in through the back of the camera and THAT will throw the meter reading.  To get a more accurate reading (and a more consistent reading) you'd need to cover that eyecup if your eye isn't blocking light from entering through the viewfinder.

I do have a Sekonic hand-held incident light meter and I've tested all of my cameras using a neutral gray card.  That means I put the meter on the surface and take a meter reading... then I point the camera at the same gray card in the same light and take a meter reading through the camera.  The result is that the camera generally does give me a reading which is pretty close.  This works for a gray card better than other subjects because if I used a piece of white card-stock (e.g. poster-board) or black card-stock... I'd get very different readings due to the way the reflective meter works.  A gray card is meant to be "middle gray".  But most gray-cards are toned to 18% gray and most cameras are actually tuned to something closer to 12% gray.  So I do expect a moderate amount of variation in the meter readings.  

I never really expect the meter built into the camera to be bang-on accurate.  That's why I own a hand-held incident light meter.  It will be bang-on accurate and pretty much in all situations.  The reflective meter in the camera will merely be "pretty close" and the camera's exposure latitude means I can just tweak up the exposure levels in post post-processing to get what I want.