AIS on D300

Encouraged by recent conversations here I tried my old 28-50mm AIS Nikkor on a D300. It worked well in A & M mode, not in P or S mode as expected. When I manually set F8 the data and monitor showed F4. F22 was displayed as F11 and F3.5 was F1.8. The camera exposed properly for what I had manually set. What's goin' on here. BTW, manual focus is not something I would want to return to. I'd post a sample here but, although sharp and well exposed, I mis-focussed most of the time.

Did you set up any non-cpu lens data?

Did you set up any non-cpu lens data?

Nope.

That was exactly the issue, displays properly when properly set. Now I know although I don't think I will be using this lens any time soon, clunking noise when shooting, hard to push/pull zoom ring, mushy aperture ring without good clicks. I've been intrigued by TT Artisan lenses but I see MF is not for my anymore. May be a mold issue to.

Go to SET UP in the menus and choose non CPU lens. Enter the focal length and F stop. Pretty easy really.

(Sorry I was 3 minutes too late.)

PS I never enjoyed MF much on my Nikon Digital DLSR’s but it is a joy with my Z 5 as you are given tools like focus peaking and magnified focus view through the viewfinder.

PS I never enjoyed MF much on my Nikon Digital DLSR’s but it is a joy with my Z 5 as you are given tools like focus peaking and magnified focus view through the viewfinder.

 

Save some $$$ and look into a used Fuji X-T2-3 camera+Nikon-to-Fuji adapter. Fuji debuted those features in 2014 with the X-T1--seriously.

Bad bad Fuji experiences, will never have another.

The whole idea of AI is that it tells the camera how far the lens is stopped down from full aperture, which is what it needs to know for TTL metering.

 

The earliest Nikon mett to TTL metering (before AI) you had to do some tricks to tell the meter the maximum aperture, so it could figure out how far down you were set.

 

So, yes, with AI lenses on a D300, you need to tell it the maximum aperture, so it knows the actual aperture.

The whole idea of AI is that it tells the camera how far the lens is stopped down from full aperture, which is what it needs to know for TTL metering.

 

The earliest Nikon mett to TTL metering (before AI) you had to do some tricks to tell the meter the maximum aperture, so it could figure out how far down you were set.

 

So, yes, with AI lenses on a D300, you need to tell it the maximum aperture, so it knows the actual aperture.

In fact the pre AI system the camera does know the actual aperture used because the aperture scale is built in to the camera. The AI system only know how many stops the lens will stop down but not the actual aperture and thus you have to enter the maximum aperture. It's kind of going back to the pre AI

In fact the pre AI system the camera does know the actual aperture used because the aperture scale is built in to the camera. The AI system only know how many stops the lens will stop down but not the actual aperture and thus you have to enter the maximum aperture. It's kind of going back to the pre AI

 

Whaaat???

Whaaat???

Yup!

The pre AI system when you mount the lens you have to do the indexing which is to line up the internal aperture scale to the aperture scale on the lens. The camera doesn't really use this information for anything except to display the aperture for you to see. The AI system (they called it auto indexing) actually is no indexing at all. The user will look at the aperture directly via the ADR (aperture direct readout).

When Nikon introduced the new Matrix metering system for the F5 and newer cameras the camera needs to know the actual aperture and thus the F5 can't support matrix metering with AI lenses and you view the aperture via the ADR. Newer DSLR's and the F6, you would have to enter the lens data and in effect manual indexing again so that the camera can display the aperture set but also making use of the matrix metering system.

I can't believe that anyone, including me, finds this discussion the least bit interesting.

One of the biggest advantages of digital photography is the instant image review.

 

So, review image, be critical, correct errors, re-shoot... smile.

 

Next challenge.

 

These are dumb, slow to use, manual focus lenses. If you want speed, you use automation.

I have always wondered if Bebu's statement is correct, as there is no indication in pre-AF cameres with no EXIF to consult, and no evidence that the AI system is actually indexing the lens. I did some not very scientific testing, and found that in any mode but matrix, my D7100 does not show any variation of exposure if one neglects to enter the lens correctly. There is a very slight, usually undetectable, difference in matrix metering, usually only at some point where the exposure kind of leaps out of linearity, and entering the correct lens changes the aperture at which this happens. Its pretty subtle even so, but if you do a series of shots in A mode, with and without correct lens data, there will be a difference at some point in the histograms.

 

But the one question I have is that the F4 data back purports to report aperture. I've never had one, and the instructions are pretty byzantine, so it's unclear whether the aperture is blank for non-chipped lenses. If it is not, then it would appear that at least in the F4 some real indexing is taking place. But the instructions are mute on whether there actually is aperture printed for a manual lens. Then again, the F4 also allows matrix metering with AI lenses, so maybe it's unique. I guess I'll never know unless someone has an F4 with the MF23 data back, and has seen what it prints when using an AI lens.

 

e.t.a. this discussion is probably not interesting. It's probably crazy. So sorry.

Well, it's interesting to me. The function of all the mount features is described in various places, some correct, some with "alternate facts". You have pre-Ai with the fork on the aperture ring. After that you have various lumps and bumps on the back of the mount, the coupling notch on the aperture ring, that small circular notch in the surface of the rear flange and linear or unknown aperture movement vs. actuator lever movement. Maybe other stuff I've forgotten. The mystery to me is what body uses what features in what modes. For example, the FG is supposed to work with Ai lenses but I know it didn't expose perfectly in program mode with some of them. I've only owned FG, FM and F3HP, so limited knowledge.

I suspect it is more interesting to those who like using old lenses on new cameras,

or even new lenses on old cameras.

 

The AI ring tells the camera how much less light there will be when the

aperture closes down. When using non-AI lenses on cameras with

a folding AI tab, you can do stop-down metering.

 

I don't yet understand why matrix metering would need more than to know how much

less light there will be.

I suspect it is more interesting to those who like using old lenses on new cameras,

or even new lenses on old cameras.

 

The AI ring tells the camera how much less light there will be when the

aperture closes down. When using non-AI lenses on cameras with

a folding AI tab, you can do stop-down metering.

 

I don't yet understand why matrix metering would need more than to know how much

less light there will be.

 

The matrix meter (one that introduced after the F4) needs to know the actual scene brightness. If a scene is brighter than EV 16.3 @ ISO 100 then it will set for EV 16.3.

The matrix meter (one that introduced after the F4) needs to know the actual scene brightness. If a scene is brighter than EV 16.3 @ ISO 100 then it will set for EV 16.3.

 

I suppose if you know accurately the curve for some film, or maybe an average characteristic of film, that helps.

 

The way I understand matrix metering is as an improvement over older meters.

 

In the early days of light meters, we had ones that average over a scene.

They needed that much light, and it mostly worked.

 

Then we had spot meters, so one could meter on the actual subject.

It is surprising (point a spot meter at different parts of a subject) how well

averaging meters do. Except for cases with a light source in frame,

they work surprisingly well.

 

But what you really want is the geometric mean, or in other words the average

in stops. (Geometric mean is the antilog of the arithmetic mean of the log.)

Metering at some points distributed across the scene, you can compute the

geometric mean. Other than non-linearities of the meter itself, that should

be independent of the scene brightness.

I suppose if you know accurately the curve for some film, or maybe an average characteristic of film, that helps.

 

The way I understand matrix metering is as an improvement over older meters.

 

In the early days of light meters, we had ones that average over a scene.

They needed that much light, and it mostly worked.

 

Then we had spot meters, so one could meter on the actual subject.

It is surprising (point a spot meter at different parts of a subject) how well

averaging meters do. Except for cases with a light source in frame,

they work surprisingly well.

 

But what you really want is the geometric mean, or in other words the average

in stops. (Geometric mean is the antilog of the arithmetic mean of the log.)

Metering at some points distributed across the scene, you can compute the

geometric mean. Other than non-linearities of the meter itself, that should

be independent of the scene brightness.

 

The matrix metering system is a lot more complex that any of the things you mentioned above. Nikon never published it fully so nobody really know how it does it. As for the scene brightness there may be more reason thant what I posted above. But that fact has nothing to do with film curve but rather it deems there is no scene actually brighter than EV16.3@ISO100.

As I recall the F4, N90 and other medium to high end bodies would give you matrix metering as long as you used an autofocus lens like a D lens but if you set it to matrix and used a manual focus lens, AI or not, it would default to center weighted metering. I never noticed much difference in the two and generally don’t worry about it now be it film or digital. That’s the nice thing about the F2. One meter mode, center weighted, and with a little attention you know when it’s wrong and how to adjust.

 

Rick H.

The matrix metering system is a lot more complex that any of the things you mentioned above. Nikon never published it fully so nobody really know how it does it. As for the scene brightness there may be more reason thant what I posted above. But that fact has nothing to do with film curve but rather it deems there is no scene actually brighter than EV16.3@ISO100.

 

If I count right Sunny 15 is EV 15 at ISO 100, so 16 for snow or sand.

There aren't many scenes brighter, though it could be done with artificial light.

 

So below EV16.3 is it linear on log(intensity)?

 

If you add a neutral density filter does it compensate exactly proportional?

If I count right Sunny 15 is EV 15 at ISO 100, so 16 for snow or sand.

There aren't many scenes brighter, though it could be done with artificial light.

 

So below EV16.3 is it linear on log(intensity)?

 

If you add a neutral density filter does it compensate exactly proportional?

If you add ND it wouldn't know so it think the scene as if it's not as bright.

Just ETTR and switch on the 'Blinkies'.... and adjust.

Small nitpick on above: The F4 does matrix meter with true AI and AIS lenses, but not with AI modified. There is an internal post that the camera reads, which is lacking even on a fully converted pre-AI lens. I don't know what, if any, other cameras have this feature. The F100 does not.

 

e.t.a. according to Ken Rockwell, among film SLR's only the F6, FE and FM share this feature.

Edited January 27, 2022 by Matthew Currie