The crazy thing about Nikon metering

[sanmich]

Exploring the possibility of converting non AI lenses to AI I have found that f/2 and f/1.8 have the exact same AI settings.
That means essentially that Nikon disregards a 1/3 step difference in metering...😱

 

[sevo]

The position on the f/1.4 is only 2/3 stop off that on the f/1.8 and f/2, so that offset continues.

 

[sanmich]

have a look here:
https://richardhaw.com/2016/01/21/diy-ai-conversion/

It's really interesting that they manipulated the aperture reading that way.

 

[sepiareverb]

The ISO settings are adjustable by 1/3 stops.

 

[sanmich]

The ISO settings are adjustable by 1/3 stops.

Usually, you set the ISO dial based on the film, not on the lens...

 

[Huss]

Usually, you set the ISO dial based on the film, not on the lens...

And the fact that you'd only do that at wide open.

Interesting observation. I am going to check it on my D750 and F6 with chipped lenses to see if there is any metering differences. They both will indicate a 1.8 aperture in the digital exposure readout when used.

 

[Fraser]

I'm not really sure what you mean f1.8 and f2 are not in the same position and my camera can tell the difference, when converting lenses I always use this table-
(credit to www.http://pindelski.org)
Screen shot 2016-11-07 at 22.12.35 by f4saregreat!, on Flickr

 

[sanmich]

I'm not really sure what you mean f1.8 and f2 are not in the same position and my camera can tell the difference, when converting lenses I always use this table-
(credit to www.http://pindelski.org)
Screen shot 2016-11-07 at 22.12.35 by f4saregreat!, on Flickr

Interesting, but it looks wrong at first glance.
For starters, on my f/1.8 lenses, the cut is at f/8+2/3 (clearly close to f/11) and it's exactly the same on my f/2 lens (not Nikkor).
My 1.4 lens shows a cut exactly at f/8
Then, your table shows the same setting for f/4 and f/4.5
last, there is something that sounds wrong in a progression based on 1/2 steps in your table, while the actual steps are 1/3.

 

[Steve M.]

They should disregard it, as there is no way on god's green earth that 1/3rd stop is going to have any effect on any film ever made that I have been in contact with. I don't know digital, but if something like that affects a shot, it's just one more reason to shoot film. I have accidentally shot some colour films 1-2 stops off and saw no difference in the pics.

Once shot a test of Tri-X at 100, 200, and 400 ISO's and while the blacks were a little blocked up at 100 it still looked good. In fact, I preferred the look for some subjects, and shot it at that speed intentionally sometimes.

 

[bakemono]

So how many ruined shots have you made due to this 1/3 stop error?
I'm guessing that not a whole lot.

 

[sanmich]

They should disregard it, as there is no way on god's green earth that 1/3rd stop is going to have any effect on any film ever made that I have been in contact with. I don't know digital, but if something like that affects a shot, it's just one more reason to shoot film. I have accidentally shot some colour films 1-2 stops off and saw no difference at all.

Once shot a test of Tri-X at 100, 200, and 400 ISO's and while the blacks were a little blocked up at 100 it still looked good. In fact, I preferred the look for some subjects, and shot it at that speed intentionally sometimes.

You are right.
But then again, don't forget slides...
And tell that to the people advocating shooting TX at ISO 320.
And to all of us listening to all the marketing BS about metering precision .. 🙂

 

[radi(c)al_cam]

And to all of us listening to all the marketing BS about metering precision .. 🙂

Well, Minolta got that metering-coupling thing right in … 1966 already, I guess?

That was about a decade before Nikon came with «AI», is that right?

I hear the Nikon fan boys' PROTEST! 😛

 

[sevo]

Well, Minolta got that metering-coupling thing right in … 1966 already, I guess?

That was about a decade before Nikon came with «AI», is that right?

The prong coupling on pre AI Nikon lenses has constant position and hence accurate aperture indexing. So they already had got it "right" before the changes that came with AI. Whatever Nikon did to the mark on fast AI lenses is obviously deliberate, and increased exposure accuracy - the most likely reason is that it is a compensation for the meter entrance pupil constraints caused by the split image/microprism on the stock Nikon focus screens. Which are f/2 constrained, so that the centre of the focus screen does not get any brighter with lenses past f/2. That must have affected the strongly centre-biased exposure on the F2 era cameras (which read their meterings off the screen). If so, the F3 (the only AI era Nikon to have focus screen independent metering) should have a 1/3 stop bias towards overexposure with fast lenses...

 

[radi(c)al_cam]

The prong coupling on pre AI Nikon lenses has constant position and hence accurate aperture indexing. So they already had got it "right" before the changes that came with AI. […]

Yes, but how incredibly cumbersome this Non-AI was and is!

The Photomic T through-the-lens light meter introduced in 1965 worked at full aperture, so the maximum aperture of the lens had to be communicated to the meter by mounting the lens with the aperture ring set to [FONT=Trebuchet MS,Candara,Georgia,Calibri,Corbel,serif]f[/FONT]/5.6, and then turning the ring to first the minimum and then the maximum apertures. (The need for this step was eliminated by the AI system below.)

https://en.wikipedia.org/wiki/Nikon_F-mount#Pre-autofocus

 

[robert blu]

is really 1/3 of stop in the exposure "crazy" influencing the photographic result?
rob

 

[gerikson]

I don't have a f/1.8 Ai'd lens handy but I distinctly remember that there's a perceptible difference between f/1.8 and f/2 when moving the aperture ring. The readout also moves from "f/1.8" and "f/2". This was a 50mm f/1.8 Series E on a D700.

 

[sevo]

I don't have a f/1.8 Ai'd lens handy but I distinctly remember that there's a perceptible difference between f/1.8 and f/2 when moving the aperture ring. The readout also moves from "f/1.8" and "f/2". This was a 50mm f/1.8 Series E on a D700.

That is not the point - AI would not work at all if it was somehow uncoupled between f/1.8 and f/2, and absolute values are encoded in the bottom tab (only regarded by - some - program mode Nikons). The point is that there is a difference in relative value encoding ("current position relative to fully open") between "fast" and "slow" lenses. When mounting a wide open Nikon lens, all lenses AI rings touch the camera AI tab and very slightly move it counterclockwise - but lenses faster than f/2 do so by 1/3 stop less.

 

[Fraser]

Interesting, but it looks wrong at first glance.
For starters, on my f/1.8 lenses, the cut is at f/8+2/3 (clearly close to f/11) and it's exactly the same on my f/2 lens (not Nikkor).
My 1.4 lens shows a cut exactly at f/8
Then, your table shows the same setting for f/4 and f/4.5
last, there is something that sounds wrong in a progression based on 1/2 steps in your table, while the actual steps are 1/3.

I can see your point but it does work, and remember the guide is just a physical/mechanical reference where to file. When using an unchipped lens the camera still has to be told what the maximum aperture is so I don't think the movement is linear, if you know what I mean 🙂
I always just file short and then tweak till it works 🙂

 

[sepiareverb]

So how many ruined shots have you made due to this 1/3 stop error?
I'm guessing that not a whole lot.

Having put a lot of slide film as well as B&W film through many Ai and AiS lenses including at least two different 50/1.8s I never noticed it. Seeing how Nikon lenses adjust in half stops seems logical that there might not be an adjustment between f2 and f8. Do the Zeiss F mount lenses use 1/3 stops like their M mount lenses?

 

[Mark C]

Interesting, but it looks wrong at first glance.
For starters, on my f/1.8 lenses, the cut is at f/8+2/3 (clearly close to f/11) and it's exactly the same on my f/2 lens (not Nikkor).
My 1.4 lens shows a cut exactly at f/8
Then, your table shows the same setting for f/4 and f/4.5
last, there is something that sounds wrong in a progression based on 1/2 steps in your table, while the actual steps are 1/3.

I've been trying to stay out of this one since I don't have 1.8 and 2 AI Nikkors to look at, but it sure sounds like you are saying this whole thread that is supposedly about Nikon metering is based on comparing a non-Nikkor f2 lens. Or am I misunderstanding your comment. Surely some folks here can post pictures of the relevant factory AI coded f1.8 and f2 Nikkors for comparison.