D850 underexposing in Aperture priority mode

If I use Aperture priority mode, it seems that if the scene is not bright, it under exposes by 1 stop or so in all metering mode. I took the photo of the sunny sky using aperture priority mode in matrix metering, and the histogram is not near the middle but to the left.

I took the same sunny sky right after with D610 with the same setting, and its histogram is near the middle.

Is this normal behavior of D850 or is my d850 in need of service, or am I misunderstanding Aperture mode? I checked that bracketing, exp compensation, and Active D lighting are not on. I attached d850 and d610 photos of the LCD screen for your view. Thank you very much.

Well, it is sunny 16. At f16, ISO 100, the shutter speed should be 1/100 sec. The problem is that the two images involve two different focal lengths.

The difference is 1/3 of a stop. IMO it is not a big deal.

Did you use the same lens for both exposures?

A better test would be to photograph an evenly illuminated plain white or grey surface, de-focused, with the same lens. This will normally put the histogram nice and central. Don't rule out the lens as the cause, I had a Canon lens which for some reason consistently over exposed, on two different bodies.

I believe that the D850 metering algorithm was changed from earlier versions, in order to give a result closer to what's commonly called ETTR - Expose To The Right.

Previous Nikon DSLRs had a nasty habit of blowing highlights in matrix mode. So be grateful!

However, I realise this doesn't explain a difference between exposure modes. It may be a peculiarity of the lens, camera or combination. Or there may even be a fault with the camera. If there is, don't expect much help from Nikon service. It's almost guaranteed to be returned with a simple 'diagnosis' of 'within specification'. And after a delay of many days or weeks if Nikon UK's shoddy service is any guideline.

"This will normally put the histogram nice and central."

I don't think so. The mid-grey histogram point is normally a little to the right of centre. I've never quite worked out what the lines represent, but all of my cameras give a blank or mid-grey metering that falls to the right of the centre line.

There is some fine tuning available on the exposure if you want to dial in a permanent shift. The D850 does have a different meter than the D610 - algorithm aside, the extra information might be making it tweak the outcome, although the images look pretty different anyway to me.

For some reason, I generally got on with the meter on my D700 (or I was less demanding), but I've had a real fight with everything since under any remotely challenging conditions (aka anything with a wide dynamic range, which is a lot of what I shoot). I often try to have the D810 and D850 in highlight-weighted in the vague hope that it won't blow the highlights, so I can recover them in post. It doesn't generally work - although I admit that a lot of the time I'm looking at the histogram on blinkies when chimping, which isn't quite the same as what the raw file holds. Why Nikon can't do an "ETTR, no, really, don't blow the highlight, if I want to blow the highlight I'll pick another mode" mode (at least for highlights big enough to fill a meter area), I don't know - unless I'm very much doing it wrong. It feels like it ought two be way less of an engineering challenge than whatever it is they think they're currently doing. I get that you don't always want to retain detail in a light bulb (sometimes I do!), but refusing not to blow the large view out of a window when part of the scene is indirectly lit inside (I'm at ISO64 and recording 14-bit raw for a reason) in annoying. I'm close to giving up and resorting to spot metering everything and hoping nothing moves.

There is some fine tuning available on the exposure

Check menu B5 on the D610 or B7 on your D850

If you adjust this setting, it doesn't show anywhere obvious. Infact Nikon have a specific warning

"Exposure compensation icon is not displayed when exposure compensation is altered from the default Value. Continue?"

Check menu B5 on the D610 or B7 on your D850

 

If you adjust this setting, it doesn't show anywhere obvious. Infact Nikon have a specific warning

 

"Exposure compensation icon is not displayed when exposure compensation is altered from the default Value. Continue?"

Ta - headache, wasn't up to hunting in the manual. :-) I've never found it consistent enough to be worth tweaking, to be honest - I just permanently expect to tweak manual exposure compensation every few frames.

Well, it is sunny 16. At f16, ISO 100, the shutter speed should be 1/100 sec. The problem is that the two images involve two different focal lengths.

 

The difference is 1/3 of a stop. IMO it is not a big deal.

Thank you very much for the reply. It is more or less .7 to 1 stop underexposure. Can I ask why the focal length matters? I thought it does not make any difference.

Did you use the same lens for both exposures?

 

A better test would be to photograph an evenly illuminated plain white or grey surface, de-focused, with the same lens. This will normally put the histogram nice and central. Don't rule out the lens as the cause, I had a Canon lens which for some reason consistently over exposed, on two different bodies.

Thank you for the reply. Why would the lens make a difference. Both are Nikon lenses. 85mm is D lens, and the other is 24-70mm f2.8G.

I believe that the D850 metering algorithm was changed from earlier versions, in order to give a result closer to what's commonly called ETTR - Expose To The Right.

 

Previous Nikon DSLRs had a nasty habit of blowing highlights in matrix mode. So be grateful!

 

However, I realise this doesn't explain a difference between exposure modes. It may be a peculiarity of the lens, camera or combination. Or there may even be a fault with the camera. If there is, don't expect much help from Nikon service. It's almost guaranteed to be returned with a simple 'diagnosis' of 'within specification'. And after a delay of many days or weeks if Nikon UK's shoddy service is any guideline.

 

"This will normally put the histogram nice and central."

 

I don't think so. The mid-grey histogram point is normally a little to the right of centre. I've never quite worked out what the lines represent, but all of my cameras give a blank or mid-grey metering that falls to the right of the centre line.

Thank you for the reply.

If somehow there is easy solution, I will just bracket the photo when using Aperture mode. I normally use manual mode when taking photos.

If you send it in for service it may get worse.

Thank you for the reply.

Do you think the service by Nikon will make it worse? I live in Los Angeles area, and there is a service center here.

Thank you for the reply.

Do you think the service by Nikon will make it worse? I live in Los Angeles area, and there is a service center here.

Yes! I think Nikon USA tech like to set the camera to expose lower. I sent my Df in for repair and while they repair the problem it came back underexpose as compared to before by about 1/3 stop. And yes I sent it the El Segundo center.

There is some fine tuning available on the exposure if you want to dial in a permanent shift. The D850 does have a different meter than the D610 - algorithm aside, the extra information might be making it tweak the outcome, although the images look pretty different anyway to me.

 

For some reason, I generally got on with the meter on my D700 (or I was less demanding), but I've had a real fight with everything since under any remotely challenging conditions (aka anything with a wide dynamic range, which is a lot of what I shoot). I often try to have the D810 and D850 in highlight-weighted in the vague hope that it won't blow the highlights, so I can recover them in post. It doesn't generally work - although I admit that a lot of the time I'm looking at the histogram on blinkies when chimping, which isn't quite the same as what the raw file holds. Why Nikon can't do an "ETTR, no, really, don't blow the highlight, if I want to blow the highlight I'll pick another mode" mode (at least for highlights big enough to fill a meter area), I don't know - unless I'm very much doing it wrong. It feels like it ought two be way less of an engineering challenge than whatever it is they think they're currently doing. I get that you don't always want to retain detail in a light bulb (sometimes I do!), but refusing not to blow the large view out of a window when part of the scene is indirectly lit inside (I'm at ISO64 and recording 14-bit raw for a reason) in annoying. I'm close to giving up and resorting to spot metering everything and hoping nothing moves.

Thank you for the reply.

I think the metering is not working correct. If the scene is brighter than the sunny sky, then the histogram is somewhere in the middle or to the right, being over exposed. In addition sometimes, in manual mode, the metering is at zero, but after I take a photo, the histogram is to the right not at the center. I will upload some photos showing this.

Check menu B5 on the D610 or B7 on your D850

 

If you adjust this setting, it doesn't show anywhere obvious. Infact Nikon have a specific warning

 

"Exposure compensation icon is not displayed when exposure compensation is altered from the default Value. Continue?"

Thank you for the reply.

But if I fine tune the metering, I think it also applies to manual mode too. I experimented and I think exp comp in B7 menu is also applied in manual mode. So, I think I can fine tune it. I have to find out more to be sure. Thank you.

Ta - headache, wasn't up to hunting in the manual. :) I've never found it consistent enough to be worth tweaking, to be honest - I just permanently expect to tweak manual exposure compensation every few frames.

Thank you for the reply.

As you said, I will probably not tweak the exposure metering as in B7 menu. I will just try to find out if there is any consistency in the exposure.

I spent over $3k, and I think I may have a defective one. Thank you.

Yes! I think Nikon USA tech like to set the camera to expose lower. I sent my Df in for repair and while they repair the problem it came back underexpose as compared to before by about 1/3 stop. And yes I sent it the El Segundo center.

Thank you for the reply.

Then, am I stuck with this possibly defective camera after spending $3K?

Jason, the reason I suggested using the same lens for both tests was basically, to eliminate as many variables as possible. On paper the lens should not make any difference to the exposure, all other things being equal. However in practice these are complex systems and sometimes they don't behave exactly as expected. In my experience it is possible for a lens to cause exposure variations. But I confess I have no specific knowledge of the camera bodies concerned.

Thank you for the reply.

I highly doubt that the lens makes a difference. The 24-70mm f2.8 was working fine with D610. I will do it as a last resort.

Thank you.

Thank you for the reply.

 

I think the metering is not working correct. If the scene is brighter than the sunny sky, then the histogram is somewhere in the middle or to the right, being over exposed. In addition sometimes, in manual mode, the metering is at zero, but after I take a photo, the histogram is to the right not at the center. I will upload some photos showing this.

But... You're matrix metering. Why would you expect it to be centred? Matrix means "apply magic to get whatever exposure the camera has decided would be appropriate", as far as I know. I've always assumed there's some kind of basic neural net involved, since Nikon talk about training it and nobody's ever explained an algorithm to me. Point at snow and matrix metering knows to place it at a high zone, if you're lucky. In your example, there seem to be denser clouds on the D850 image - I'd choose to underexpose a little to avoid blowing them.

If you get a difference in centre-weighted metering, that would be different. Certainly the matrix behaviour may have changed (the meter changed, so it had to, plus the D850 has another mode), but it's always been "probably the right exposure" rather than "completely repeatable".

But I've already said I don't get on with the meter, so I may be completely confused.

Here are additional photos of D850 aperture mode. If the scene is not dark, it is either normal or over exposure as compared to D610, which is normal in all situations.

How do you account for this D850 Aperture priority mode behavior?

Thank you very much.

Thank you for the reply.

I highly doubt that the lens makes a difference. The 24-70mm f2.8 was working fine with D610. I will do it as a last resort.

Thank you.

Well, the vignetting behaviour will differ, and I suspect the difference between T-stop and f-stop for a relatively complex zoom compared with a fairly simple (small number of surfaces) prime might come into it. Could be a complete red herring though.

But... You're matrix metering. Why would you expect it to be centred? Matrix means "apply magic to get whatever exposure the camera has decided would be appropriate", as far as I know. I've always assumed there's some kind of basic neural net involved, since Nikon talk about training it and nobody's ever explained an algorithm to me. Point at snow and matrix metering knows to place it at a high zone, if you're lucky. In your example, there seem to be denser clouds on the D850 image - I'd choose to underexpose a little to avoid blowing them.

 

If you get a difference in centre-weighted metering, that would be different. Certainly the matrix behaviour may have changed (the meter changed, so it had to, plus the D850 has another mode), but it's always been "probably the right exposure" rather than "completely repeatable".

 

But I've already said I don't get on with the meter, so I may be completely confused.

Thank you for the reply.

I understand what you mean. But please look at the 6 photos that I just uploaded. I took the same scenes with both D850 and D610. The D610 histogram is consistently somewhere in the middle, whereas the D850 is not. I took the scenes of low contrast almost even in color to make a point. How do you account for the D850 histogram in this case?

Here are additional photos of D850 aperture mode. If the scene is not dark, it is either normal or over exposure as compared to D610, which is normal in all situations.

How do you account for this D850 Aperture priority mode behavior?

 

Thank you very much.

 

[ATTACH=full]1303876[/ATTACH] [ATTACH=full]1303877[/ATTACH] [ATTACH=full]1303878[/ATTACH] [ATTACH=full]1303879[/ATTACH] [ATTACH=full]1303880[/ATTACH] [ATTACH=full]1303881[/ATTACH]

I think you have a lot of variables to consider.

You started by saying this was aperture priority. With just the D850 and the same lens, focal length and focus distance, pointing at the same subject, and using a simple metering mode (centre-weighted or spot), do you get different exposures in aperture priority compared with applying program shift in program mode, or selecting the same shutter speed in shutter priority? If you shoot the same settings in manual, is the meter centered?

If aperture priority itself is an issue, that's one thing. If everything matches with a controlled experiment, you can start comparing with matrix, different lenses, and different bodies. At least, I'd want to simplify the situation.

Thank you for the reply.

I understand what you mean. But please look at the 6 photos that I just uploaded. I took the same scenes with both D850 and D610. The D610 histogram is consistently somewhere in the middle, whereas the D850 is not. I took the scenes of low contrast almost even in color to make a point. How do you account for the D850 histogram in this case?

Well, the algorithm (or training set) may have changed. I think I'd say it was coincidence that the D610 happened to be centered. In matrix metering, the question is whether the exposure is aesthetically "better" rather than centering the histogram, and I'm very willing to believe the D850 does things differently. Unless you see another inconsistency, I'd just be interested in whether the exposure chosen was worse for my needs.