Live view metering

Sorry to ask a novice question, but I realised I don't know the answer to this across all Nikon bodies...

 

Someone online was claiming that low-end Nikons don't support metering with manual lenses "for no mechanical reason". I could mostly argue with that, but realised I don't know as much as I thought about the low-end Nikons when it comes to live view.

 

Things I think I know:

• Low-end Nikons can't tell the difference between a mechanical AI or F lens (not set to minimum aperture) and a lens with no aperture lever - they don't have the AI follower ring of the high-end Nikons. If the aperture changes between wide open and some unspecified alternative for the shot, the camera can't meter. The solution of moving the aperture lever (as in DoF preview) would make the viewfinder unacceptably dark some of the time. So no open-aperture metering on Nikons with no aperture following tab, and aperture following tabs are mechanically complex and cost money (by D3x00 standards). "For mechanical reasons".
  
• No Nikon DSLR offers stop-down metering. I've no idea why, since various film SLRs do.
  
• High-end Nikons seem to have metering in live view and low-end ones don't (or didn't?) I imagine this is complicated on lower-end ones that don't have independent aperture lever control (ones on which, if you change the aperture from the camera in live view, nothing happens immediately). Metering in live view would make more sense when stopped down (exposure preview mode). I'll give it a go with a fully-manual (no coupling) lens when I get home with my D810 - I believe it works, and obviously the camera can't tell its aperture from "wide open" - but that won't tell me what the D3x00 and D5x00 series do. The manual seems to be uninformative...
  

So, if you please: How correct is my logic for why low-end Nikons can't meter with manual lenses? And what's the behaviour of live view metering with manual lenses on different bodies? (Obviously you can take a shot and then look at the resulting histogram, but that's never quite as convenient.) And if anyone can tell me why stop-down metering doesn't work, that would be educational.

 

I'm assuming, for the record, that Canon don't do focus confirmation with mechanical lenses because they want to switch between their f/2.8 and f/5.6 AF sensors, and absent any electronics they choose (for some reason) "none of the above". Or maybe they (and Nikon) just want you to upgrade to lenses with electronics. But I'm more interested in the Nikon behaviour.

 

Cheers! (Here to learn, if only for academic interest.)

This will be interesting. I know that it is impossible to use live view on my D7200 with a 1000 mm Nikon mirror lens for a moon photo. Whatever setting I use results in a completely burned out white circle in live view that is impossible to focus on - but the end photo is fine (if you guess exposure correctly). With my D810 it is fine - just adjust the exposure in manual mode (in live view) until it is correctly exposed.

 

Why this is I don't know - it never used to be this way - the D7000 was able to meter in live view I'm sure.

Having used a D3200 extensively with manual lenses, I can at least confirm that it does not meter at all ever with any, including presets and aperture-less lenses. I can understand why it would not work with any auto aperture lens, since the light does not change when aperture is changed while you're viewing, and there is no preview button. I think others have discussed why the DOF preview on digital cameras does not allow stop down metering, but in any case there's no such button to argue about. One might have thought the d3200 could meter a preset or manual aperture lens, but I think the short answer is that without an electronic contact, the meter is simply switched off. In any but manual mode, the camera returns a "lens not mounted" error, and in manual mode with a manual lens no meter settings are available. It would seem that if they wanted, they could have made some limited metering possible with non-automatic lenses, allowing the meter simply to read the light through the lens, as older cameras do with uncoupled lenses in stop down mode, but I imagine few low end users would find it useful and confusion is avoided entirely by simply switching the meter off.

 

As for metering in Live view on others, I believe the D7100 and its ilk do meter in Live View, but what they do not do is show the variation in the display. Live View has its own meter just for the display, and does not reflect a changed aperture until after a shot is made, but the change is effective for the exposure. This is true on the low end ones as well on lenses that do operate the meter. Live view will give you whatever aperture was selected before LV was initiated, and will show no variation when aperture is changed. But the shot will be metered, and LV will return at the aperture of the last shot.

 

On low end cameras, this feature serves as a sort of poor man's DOF preview, since one can change aperture out of LV and the re-initiate LV to see the difference.

Ian - thanks. It looks like the D7200 lacks the "Exposure Preview" mode on the D810 (that stops down the aperture) - otherwise the D810 just tries to show you a clear image that you can focus on. I guess the D7200 is permanently in the latter mode; to be fair, that is what you might want if you were trying to take a macro shot with a flash, because you wouldn't be able to see to focus otherwise. As Matthew says, it sounds as though there's metering, but the display is trying to auto-expose - which always goes wrong with a shot of the moon because the black background confuses it (as if you try to matrix meter on the moon through the viewfinder).

 

Matthew: Cheers for the D3200 knowledge, and I agree that part of the behaviour might be just down to Nikon trying to keep the interface simple on a camera designed for the simpler end of the market. (I still don't really know why manufacturers don't just have a menu option for "expert mode" so all the scary stuff can be hidden but the camera doesn't have to be crippled because of it.)

 

I think others have discussed why the DOF preview on digital cameras does not allow stop down metering

 

Do you know what the conclusion was? I've certainly complained about it, especially for the Df, where the "set the aperture twice" thing seems like a horrible workaround to me (if you're actually changing the aperture dynamically, of course). If there's a mechanical reason, I'll be interested. It's always seemed odd, since the F5 (which can stop-down meter) has the same metering hardware (plus, admittedly, the one in the AF module) as the D3/D300/D700 generation - none of which support stop-down metering. Throw in the independent aperture lever control from the D3/D8x0 series and it doesn't feel hard to implement.

 

One minor reason for my switch to Nikon was my feeling that Nikon tend to cripple their low-end cameras by leaving out features that cost money - whereas the Eos 300D I own is crippled compared with the 10D largely due to firmware alone (as I proved by installing hacked firmware). I don't mind my hardware being limited, it's artificial limits that I don't like.

 

I'm reasonably sure I can work out the behaviour of my own camera, but after nearly ten years of shooting Nikon I should probably have a handle on how this works across the range by now!

No Nikon DSLR offers stop-down metering. I've no idea why, since various film SLRs do.

They could, but Nikon blocks the feature. I believe this has been discussed numerous times in this forum (and certainly others) - a search on stop-down metering including both Roland Vink and Bjorn Rorslett should reveal ample information. For starters, here's one: Nikon Df non-AI stop-down metering It is claimed that the reason is the mechanical aperture linkage - I need to take my "linkage-less" F-mount-modified Apo-Telyt down to the local store to try it out on a "consumer-grade" Nikon DSLR. I just tried it on the D500 - with no entry in the non-CPU lens menu (which consumer-grade cameras don't have anyway) - and camera meters just fine (also in live view) while showing the "deltaF" symbol in the display (and no pushing of the preview button since it can't affect anything on a lens that has no linkage to the camera body (except the mount, of course). Naturally, matrix metering doesn't work that way - the camera defaults to center-weighted.

 

On a side note: I just learned that apparently the D3400 cannot even meter with CPU-lenses that have an aperture ring - for lack of the switch that indicates the aperture is set to its minimum value.

Edited March 8, 2018 by Dieter Schaefer

Thanks, Dieter. As you say, and from the thread you linked, it sounds as though they don't do it because they don't want to. I don't really buy the "different meter value" argument: of course the meter is different, you're stopping down the lens and affecting vignetting, which makes it more accurate. I'd buy the "light coming in the eyepiece" argument more if the F5 didn't support it (Nikon could always automate the eyepiece shutter!) And the aperture lever problem should go away by simply moving the aperture lever to its smallest position, guaranteeing that the lever will hit the stop on the lens if there is one. So I still don't see the technical problem. I'll rummage for other threads, though.

 

I don't desperately mind (though it came up because of the new Laowa macro, which mildly tempts me). But at least for the Df where part of the point was compatibility with pre-AI lenses, it seems completely unnecessary. The best excuse I can think of for the Df is that it inherited the D600 shutter, and that doesn't offer the D800's independent aperture lever control. It still feels like there should have been a less hacky fix than setting the aperture twice, though. But I'm not a Df shooter, and (like a recent argument I've been having about US prices not including sales tax on the label) I guess it's not as irritating if you're used to it.

 

I'm going to go and remind myself what my D810 does in live view. I should probably get my tilt-shifts out so I've got something to produce for next Nikon Wednesday anyway.

I don't remember the final conclusion on the discussion of stop down levers, except that the linkage is not purely mechanical, but I don't think it was ever explained why it has to be the way it is, except that it's easier for Nikon to do it that way.

 

With regard to uncoupled lenses, on a low end camera like the D3200 the meter simply will not operate with anything but a chipped lens. Whenever you get a "lens not mounted" error the meter is just plain off. At that point the camera doesn't care what, if any, lens is on. In manual mode it will shoot with any lens or no lens, and will leave the EXIF information blank.

 

On whatever one calls the middle end occupied by the D7100 and 7200, an uncoupled lens with an automatic aperture will operate the meter and function in M or A modes, but will provide the wrong reading unless it's shot wide open, since it's metered wide open, and the meter does not operate on the fly. A manual aperture or preset lens will provide a correct reading. My favorite normal lens on DX format is the 35/2.8 PC, which is quite at home on a D7100. The Non CPU settings really don't do much except to render correct, or nearly correct, entries into the images EXIF info, so in the case of that one, it will tell you the FL of the lens but will likely lie about the aperture.

 

If you had some pet uncoupled auto aperture lens (Not many of those out there, but there are some old lenses such as Vivitars and T4 mounts that do not take well to AI coupling), you could get around the metering issue to some extent on a D7x00, and I expect others, by calculation. You'll get a five stop allowance in A mode with the compensation button, or of course you can do whatever you want in M mode. Take the meter reading and then calculate how many stops you're going from the max, and adjust it.

 

edit to add: on that latter case we are assuming the lens is safe to mount over an AI follower. T4 lenses are, and at least some pre-AI Vivitars are, but not necessarily.

"I know that it is impossible to use live view on my D7200 with a 1000 mm Nikon mirror lens for a moon photo."

 

- Not impossible, because I've done the exact same thing. There's a way to override Live View's auto brightness, but I can't remember the procedure right now.

 

Anyway. Another oddity to throw into the mix: I was using an old 35mm f/2.8 PC-Nikkor on the D7200. The lens has been modified (OK - butchered!) to prevent damage to the AI coupler tab, and effectively 'tells' the camera that the lens is at f/16, regardless of the actual (preset) aperture. Now, amazingly, the camera still gives reasonable auto exposures with the lens. I can't figure this out, because if I set the lens to f/5.6 or f/8 the camera ought to be overexposing by 2 or 3 stops - thinking the lens is going to stop down to f/16, right?

 

That's a real puzzler. Same as why my old Canon 5D used to give deep amber pictures when fitted with an 85A filter in Auto White Balance and daylight! - Just don't ask why I tried that.

The preset lens gives a proper exposure because when you reset the aperture, it actually changes the light that comes to the meter, so the meter compensates, not by calculating aperture as it does when coupled, but by altering the perceived light. It's the same as if you had coupled lens on and then turned the lights off. The meter would compensate.

Matthew - but in RJ's case, the camera believes it's metering wide open, but believes that the AI coupler is telling it that it'll be shooting at f/16, 5 stops down from the wide-open metering value. So it should be trying to set the exposure for five stops darker than whatever it's seeing through the finder. If I understand correctly, Joe reports that when he sets the lens to an intermediate aperture, so the lens doesn't actually stop down to f/16, the exposure is still correct - not 2-3 stops overexposed, as I'd expect.

 

So I'm confused by what he's confused by. :-)

Here's what happens if I mount my F-mount modified Leica lens onto the D500 (without entering any information in the non-CPU lens menu and none present there); the scenario is equivalent to mounting RJ's 35 PC lens since in both cases there is no automatic aperture and no coupling of the aperture lever to anything in the lens. The difference, of course, is that RJ's 35 moves the Ai follower tab whereas my Leica lens does not. But I tried to mimic that behavior by moving the Ai tab by hand to various positions. For any "preset" aperture value, I always got the same exposure no matter what the location of that Ai tab. What actually does happen is that when I move the Ai tab, the display indicates "deltaF" stops from open and shows - briefly - the changed shutter speed (in A mode), then reverts back to the one that corresponds to the actually "preset" one. So moving the Ai tab has a brief influence on the displayed parameter before "something" intervenes and corrects the speed back to what's correct for the preset aperture. What exactly that "something" is I don't know - but the resulting exposure does not depend on the position of the Ai tab.

Maybe it's the lack of a lens aperture actuator lever that triggers the effective 'stop-down metering' behaviour?

 

This could easily be detected by the corresponding actuator in the camera rising beyond the normal end stop position. The camera body then says ''Aha! No lens, or a preset/fixed aperture lens fitted. I'll ignore any other info apart from the light actually getting through the lens." Or words to that effect.... and in binary.

 

Should be easy enough to check by partially de-mounting a non-CPU Ai/Ai-s lens; such that the lens aperture lever lifts away from the camera body actuator. While watching the exposure.

 

I'll let someone else try this juggling act. It'd be just my luck for the lens to jump clean out of my hand and bounce sickeningly across the floor!

I think it's only because the matrix metering system and Nikon decided that no matrix is as bad as no meter. I actually rather do the set the aperture twice on the Df than stop down metering as stop down is clumsy for A mode and works quite well with the Df (setting the aperture twice but I don't tend to change the aperture often when using A).

I wrote this a while ago and forgot to push the button, so in the meantime I tried partially mounting a D lens, and seeing how it behaves.

 

If you simply leave it uncoupled and operate the aperture ring, it gives the correct exposure, up until the point where the aperture ring starts to move the AI coupler. I did not want to demount it too far for fear of dropping something. If one leaves the aperture ring alone and moves the aperture coupler to meet it, it seems also to give a correct exposure, though owing to difficult access and whatnot I did not carefully observe what the display said. In any case, it seems to correct the exposure.

 

So, back to the previous post....

 

A pre-AI 35/2.8 PC does not move the AI tab. The back ring is non rotating, and the only reason to modify it is to prevent actual crash damage, I have the same one (or likely the same, a pre-AI, silver knob, 35/2.8 from the early 1970's). Mine has the ring turned down on a lathe and repainted so it does not look so butchered.

 

I can't try moving the AI tab with the PC lens because I don't have it on hand. But I think what is happening is that the meter first reads the actual light coming in through the lens. When the AI follower is moved, it applies a calculated offset to that reading, but then, as expected, it continues to monitor the incoming light. But whatever light it reads after that initial offset is presumed to be read through the full aperture, so it recalculates for what it presumes is a reduction of the real light, and corrects the exposure.

"But whatever light it reads after that initial offset is presumed to be read through the full aperture, so it recalculates for what it presumes is a reduction of the real light, and corrects the exposure."

 

- That's not what appears to happen though.

 

Let me explain more fully: The pre-AI PC-Nikkor I have has had part of the rear ring removed, but only sufficient to allow mounting without fouling the AI follower tab. The cut-out stops short of the white mounting-index on the lens, otherwise the index dot would have been machined through. This results in the follower tab being moved by a few stops as the lens is mounted.

 

Despite the tab being moved, the lens meters correctly at whatever aperture is set and activated on the lens. In other words - stop down metering applies, and no re-calculation of exposure takes place due to the displaced AI follower.

 

This shouldn't happen. Not unless the camera body has some way of detecting the absence of an auto-aperture lens mechanism.

 

As I suggested above, the easiest way for Nikon to detect this would be to sense an abnormal positioning of the aperture actuator in the camera body. But maybe another method has been used that isn't so easy to detect.

I would assume that the initial offset of the tab has an effect different from moving it later. It tells the camera that a lens of X maximum aperture has been installed, and that the meter is now reading at its full aperture. After that initial offset the AI follower presumably does not move. Any change of the aperture of the preset lens is presumed then by the meter to be a change in the actual light, and the meter adjusts..

 

What I was referring to, and probably not thinking through quite all the way, was Dieter's observation that when you move the AI tab when a manual aperture lens is on, it briefly alters the meter reading and then returns to a correct exposure. That does sound as if the meter might be detecting the absence of an automatic lever, but I haven't had a chance to try this for myself.

Too late to add an edit: right now I'm away from home and don't have much equipment at hand, but it occurs to me that when I get back in a couple of days I might be able to find a lousy quality AI compatible tele extender that's been kicking around for a while, and which I'm unlikely to miss if I accidentally break it. If I can bend the auto aperture tab out of the way on that, it should then move the AI follower when aperture is adjusted, but not engage the auto aperture lever. If that behaves as Dieter's experiment suggests, and readjusts the meter after the aperture is changed, then it's pretty clear that it must be sensing the automatic aperture tab, since we know an AI lens will not do that otherwise.

 

But I still think that in the case of the PC lens it makes no difference, since the AI follower stands still once the lens is mounted, so the initial reading is presumed to be at full aperture whatever its starting position is.

One minor reason for my switch to Nikon was my feeling that Nikon tend to cripple their low-end cameras by leaving out features that cost money - whereas the Eos 300D I own is crippled compared with the 10D largely due to firmware alone (as I proved by installing hacked firmware). I don't mind my hardware being limited, it's artificial limits that I don't like.

 

This is totally off topic, but I've often thought that the F4 was deliberately crippled. It's one of a handful of bodies that has the AI-s button on the mount, but as best as I can tell there's nothing in the camera that the presence or absence of an AI-s lens changes.

 

The F4 SHOULD be able to provide full automation with AI-s lenses at a minimum, and it wouldn't be a stretch for it to work with all "real" AI lenses in all modes with those lenses. The FE(and presumably other multi-mode cameras that work with AI lenses like the FG, N2000, and N2020) can work with AI lenses by stopping down the lens and taking a second meter reading just before firing the shutter.

 

In any case, if I were brave enough to dig into the F4 and do some reverse engineering, I wonder if it would be possible to re-write the programming in it to allow that. Presumably the "brains" are on a ROM chip, and changing it would require pulling the chip, dumping the ROM, modifying it, then finding a "blank" chip(possibly not as much of an obstacle as many of these chips are off the shelf and still used in embedded systems) and "burning" it. The knowledge do so is well above my pay grade, though, especially since my hex editing abilities are basically zero. Of course, I'm also guessing the ROM isn't socketed, which would mean both unsoldering and resoldering the chip to the flex board...

 

In any case, that's just some idle Saturday morning rambling.

I have also wondered about the F4 shortcoming, as it's one of the few cameras that bothers to have an AIS feeler. The only thing I can think of is that it might modify the meter reading slightly to reflect the even spacing of the AIS aperture.

 

I wonder, though, how often one would wish for an AIS lens to be used in P and S modes at the sacrifice of M and A. It's a lot of work to get there.

I have also wondered about the F4 shortcoming, as it's one of the few cameras that bothers to have an AIS feeler. The only thing I can think of is that it might modify the meter reading slightly to reflect the even spacing of the AIS aperture.

 

I wonder, though, how often one would wish for an AIS lens to be used in P and S modes at the sacrifice of M and A. It's a lot of work to get there.

I wouldn't want a camera without M mode. In fact I am OK with the low end DSLR not having meter with AI/AI-S lenses. But without M mode I can't use it.

I agree, but now I'm not entirely sure whether a modification would disable A and M modes, or just switch over to the others when the switch is changed, and since I don't now have an F4 handy, I'm not 100 percent sure what it will and won't do. In any case, I agree that M is the most important thing to keep. A low end DSLR with a manual lens is basically like a DX version of an old meterless F except that you get immediate post-shot feedback.

I wouldn't want a camera without M mode. In fact I am OK with the low end DSLR not having meter with AI/AI-S lenses. But without M mode I can't use it.

I guess it's a curse! I did some party photographs today with the Df and the meter stopped working. It always indicates an underexposure condition. It acts like the lens cap is on even when of course it's not on. Live view the meter works. So I ended up complete the shoot in M mode and no meter. The results are fine as I have said no meter is no big deal but now I have to send my Df out for repair.

Belatedly... yes, sorry, I'd failed to register that RJ's case was an old PC lens. No aperture lever makes sense.

 

I am, generally, unclear why DSLRs couldn't use stop-down metering if Nikon wanted to allow it. I'm certainly unclear on why AI-S lenses (even if you have to tell the camera that's what it's got, rather than having the feeler as in the F4) can't use camera-controlled aperture. I suspect it's under Nikon's "don't care" list. I'm vaguely curious how much room there is available in the BIOS of a modern DSLR; there were rumours somewhere that it's a bit close to a size limit, but that may be misinformation. As I've said before, I'd love to be given a few days to play with Nikon's DSLR BIOS code - or for them to open it up for general use. I get that Nikon sometimes have functionality for which they'd like you to pay for the next model up, but in this case there is no next model up.

I'm certainly unclear on why AI-S lenses (even if you have to tell the camera that's what it's got, rather than having the feeler as in the F4) can't use camera-controlled aperture.

 

I SUSPECT that might be dangerous territory from Nikon's perspective. If someone entered an AI lens as AI-s, chances are their exposures would be wildly off at least at smaller apertures. The alternatively would be to stop down the lens and meter again as on the FE, etc, but then you don't get the claimed aperture.

 

In fact, I don't like using non-AI-s lenses in P and S mode on my FE for this exact reason. Even though the camera will always give the correct exposure(or at least what the meter judges to be correct) what it tells me is f/8 could end up being f/5.6, for example. Of course, since switching to Nikon I've actually moved over to a "Nikon mindset" and generally use aperture priority on AE cameras anyway, but still :) .

 

Incidentally, as a side note, I handed my D800 to someone the other day who I knew uses a D3xxx or D5xxx and asked them to take a picture of me and someone else. Their first instinct was to grab for the mode dial and twiddle it to "portrait", and when they didn't find it they asked me where it was. The camera was in aperture priority, set at f/8(to give the DOF I wanted) and was set to an ISO that would give a workable shutter speed. I told them not to worry about it and just take the picture with how it was set. That lead to a discussion on using the PSAM portion of the dial on their camera...but I think I lost the person a bit and my girlfriend was starting to get a bit impatient even though she knows me well enough to know I can pretty much strike up a conversation with anyone about anything :) . In any case, I've only used "green box" and some of the creative modes a few times just to see what they did(both on my Rebel XS and on my D70). I found it frustrating how LITTLE control I had over the camera-even rolling through shutter speed/aperture combinations is locked out, and you get no choice of focus point(I'm usually a center, lock, recompose guy) or really anything else. About all you can do is change the image quality and set the flash to off, on, or auto.

Reporting back on my manual lens experiment:

 

I used a D7100, which meters with AI lenses, and a 50/2 AI lens which works fine on it. I took some pictures at various apertures, and it worked as it should.

 

I should mention that I also often use a preset PC lens and that also works as it should. That is to say both AI and preset lenses give correct metering in A mode.

 

I then took an "Imado" brand AI teleconverter and bent the aperture lever so that the lens does not engage it. The lens now stops down when the aperture ring is turned but also still moves the AI follower.

 

Since it's an F2 lens with a 2x converter, it does move the AI follower when mounted. Shot wide open, the exposure is correct.

 

When I stop the lens down, the meter adjusts for the lowered light, as one would expect with a non-automatic lens, but it also offsets based on the AI follower. The resulting exposure is overexposed, in proportion with the offset. So it's one stop over at 2.8, and 6 over at 16. I didn't bother to record and test every step, but it was clear that the exposure was going way way off.

 

I don't know what was happening with Dieter's camera, but with mine, at least, it's clear that it is not correcting for the absence of an automatic aperture.