A Question about Stops

[xpatUSA]

Happy New Year, all!

I've been playing with my new toy: the Gossen Starlite 2 light-meter.

However, I notice that the analog scale cursor moves half a step (or stop) at a time, when moving the thumb-wheel.

Meanwhile, my Sigma cameras lack an option to change the step size, being fixed at one-third of a step.

Having said all that, my actual question is: What was the rationale for the industry changing from 1/2-step to 1/3-step and is there a credible reference (other than general opinion) that discusses that?

I actually don't like 1/3 steps much - the pedant in me cringes whenever I see 0.7 EC instead of 0.66666666 ...

[royent]

On my Canon 6D I can change both the apertures and the ISO values to either 1/2 stops or 1/3 stops

[rpcrowe]

Early on in my film days, generally over the second half of the 20th Century, I worked usually in half stops but, after I switched to digital at about the turn of the century, I began using 1/3 stop intervals; first with Canon DSLR cameras and then with Sony mirrorless.

Neither is easier or more difficult but, the 1/3 stop intervals "seems" to allow somewhat finer adjustments than did the 1/2 stops.

However, when calculating exposure in my mind, it is still easier to do the calculations in half stop intervals.

[pschlute]

Cannot answer your question directly I am afraid. All my cameras for the last 25 years have the option to select 1/3 or 1/2 stop increments.

Like you i don't like 1/3 increments much. I find it easier to calculate in my head when using 1/2 stops, so I always set my equipment that way. I invariably adjust the exposure in pp anyway.

[Manfred M]

Ted - I vaguely remember Canon launching that approach many years ago more as a marketing move (finer increments = better) and the rest of the pack seemed to follow over time. True cine lenses do not and so far as I know have never had click stops so in theory are infinitely variable when setting f-stops.

Studio flash seems to have gone even sillier. In the old days, 1/3 ev was a pretty standard adjustment, but a lot of the "pro" gear now allows changes of 0.1 ev. Can I generally see a difference of 0.1 ev? No, not really.

[pschlute]

I guess if you shoot film and especially slide film then 1/3 can have an advantage over 1/2 stop increments. With digital, no.

Ted....that light meter is a very smart bit of kit !

[DanK]

I have no idea why manufacturers started offering 1/3 stop intervals, but I don't find them useful. After all, the maximum improvement in exposure is the difference between 1/3 and 1/2 stop steps. That is, if 1/3 stop intervals got you to the perfect exposure, using half-stop intervals would make your exposure 0.17 stops off. Hard even to see that. Moreover, if by chance your perfect exposure is at a half-stop interval, you will have an error of that size in the other direction if you use 1/3 stop intervals. Yes, the errors are more more frequent with larger intervals, but they will almost always be trivial and in most cases not really detectable. And like Peter, I find the arithmetic simpler with half-stop intervals.

[xpatUSA]

Thanks all for comments so far.

Another irritation in the world of Sigma is that, when using a legacy lens (no electronics), I can't set a half-stop value in-camera so that the EXIF comes out right ...

Good point about the 1/12 EV difference betwixt the two, Dan ... (My Casio does fractions)

Thanks for the historical note, Manfred.

The Gossen is indeed pretty smart, Peter. Smarter than me, currently.

Off-topic, but keeping it hangs in the balance for now while I decide whether the spot-metering is "better" than Sigma's. By which I mean: is the constant and narrower angle of practical use in my mostly snapshot work? I've just estimated that my fixed-focal-length Sigma compact's spot is about 5 degrees ...

[William W]

. . . What was the rationale for the industry changing from 1/2-step to 1/3-step. . .

I think that it was mainly an alignment of manufacturers’ marketing in the sense of each needing to be respected as “keeping up”.

I think this “progression” became a major marketing tag in 135 Format.

Some (most?) early Pentax M42 and Minolta Lenses (as two examples only) had ½ Stop clicks.

Nikon (and others) moved to MC (Meter Coupled) Lenses – AND – allowed ⅓Stop increments of the aperture, via the camera body, not the lens.

This marketing feature did not only progress on the Aperture available: it was quoted on Filter Factors, Film Speeds; Reciprocity Factors; ‘best practice’ for Test Strips; Light Meters, etc.

Once locked in, what manufacturer of any gear would ‘revert’ to ½Stops?

***

and is there a credible reference (other than general opinion) that discusses that?

None that I know. This discussion, is the first for me.

However, as one first step, I think there would be some credence drawn by aligning the time lines of when different photographic gear ‘progressed’ from ½ Stops to ⅓Stops.

WW

[William W]

Haa Ahh!

I have just read the other replies.

I think that Richard's historical note of his own usage, (it mirrors mine) and Manfred's recollection of Canon's initiation at the least support a similar opinion to mine, and, at the most are three individual experiences which add weight to what might be close to a reasonable answer or part thereof.

WW

[rpcrowe]

My exposure using either 1/3 or 1/2 stop increments is usually pretty well within the ball park. So I guess that is really doesn't matter one way or the other - the destination is important not the route at which we arrive at that destination...

[xpatUSA]

<>

Some (most?) early Pentax M42 and Minolta Lenses (as two examples only) had ½ Stop clicks.

Thanks for the further input, Bill. I have two M42 'Legacy' lens, both indeed with 1/2 stops.

Oddly, Wife's Panasonic LX-7 has a "Leica" lens with an aperture ring (fly-by-wire?) on it and the ring has 1/3 stops. My old Sekonic L-398 has 1/3 stop intermediate graduations on it's magnificent scales too.

Nikon (and others) moved to MC (Meter Coupled) Lenses – AND – allowed ⅓Stop increments of the aperture, via the camera body, not the lens.

This marketing feature did not only progress on the Aperture available: it was quoted on Filter Factors, Film Speeds; Reciprocity Factors; ‘best practice’ for Test Strips; Light Meters, etc.

<>

However, as one first step, I think there would be some credence drawn by aligning the time lines of when different photographic gear ‘progressed’ from ½ Stops to ⅓ Stops.

Pardon the snips!

[xpatUSA]

However, as one first step, I think there would be some credence drawn by aligning the time lines of when different photographic gear ‘progressed’ from ½ Stops to ⅓Stops.

One small step for pedant-kind:

DPR's Canon D30 review dated 2000 quotes 1/3 steps, apparently as default, with 1/2 steps selectable as Custom.

[xpatUSA]

One small step for pedant-kind ...

DPR's Canon D30 review dated 2000 quotes 1/3 steps, apparently as default, with 1/2 steps selectable as Custom.

... one giant stride sideways:

I though perhaps there would some guidance from exposure-meters and looked at Doug Kerr's history of "Director" style models.

dougkerr.net/Pumpkin/articles/Norwood_Director_meters.pdf

My circa 1980 L-398 is in there with it's 1/3 steps, as is the previous 1970-1976 model with it's half-steps. Aha!, I go. But then Kerr throws a wobbler with the 1958 Norwood Super-Director which has ... 1/3 divisions on the aperture scale. Yep, 1958.

I think I'll get a Hurter & Driffield 'Actinograph' and be done with it.

[Manfred M]

My old Sekonic L-398 has 1/3 stop intermediate graduations on it's magnificent scales too.

My Sekonic L-358 allows me to set the unit to 1/2-stop, 1/3-stop and 1/10-stop readings via a DIP switch under the back cover.

[xpatUSA]

I have no idea why manufacturers started offering 1/3 stop intervals, but I don't find them useful. After all, the maximum improvement in exposure is the difference between 1/3 and 1/2 stop steps. That is, if 1/3 stop intervals got you to the perfect exposure, using half-stop intervals would make your exposure 0.17 stops off. Hard even to see that. Moreover, if by chance your perfect exposure is at a half-stop interval, you will have an error of that size in the other direction if you use 1/3 stop intervals. Yes, the errors are more more frequent with larger intervals, but they will almost always be trivial and in most cases not really detectable. And like Peter, I find the arithmetic simpler with half-stop intervals.

Reading more in the Literature, some authors actually split 'Zones' into half-zones. Haven't noticed a 1/3 mention yet. Like Manfred's Sekonic, the Gossen goes better than either one! When averaging, Zones are displayed with 0.1 resolution ... try and find that on your camera EC knob ... (um, not that I would use averaging, BTW).

Here's a little jewel for y'all, found about a second ago (Texas second):

"With reflected reading, you need to know the reflectance of an object metered. If it is not a middle grey, you will need to compensate or the pics will be over or underexposed."

I imagine carrying around a table of all known reflectances but I'm sure that's not what is meant.

Comments?

[Manfred M]

"With reflected reading, you need to know the reflectance of an object metered. If it is not a middle grey, you will need to compensate or the pics will be over or underexposed."

I believe that's called stating the obvious.

Fortunately, given the dynamic range and the "fact" that most scenes are pretty close to "average", reflective light metering gets a result that is good enough and any minor exposure tweaking can be done in post.

I guess one could carry a gray card and get a reflective reading off it, but in most cases where that would work, I would just pull out my incident meter and remove all doubt...

[William W]

Reading more in the Literature . . . Here's a little jewel for y'all, found about a second ago (Texas second):

"With reflected reading, you need to know the reflectance of an object metered. If it is not a middle grey, you will need to compensate or the pics will be over or underexposed."

I imagine carrying around a table of all known reflectances but I'm sure that's not what is meant.

Comments?

I think (have strong view that I am correct), the author is referring to "Colour and Density of Colour".

I have a (physical) chart of the "typical" colours which are to be found in typical scenes. I think it is a Kodak Publication.

I will attempt to find it when I arrive back home. I expect it or a facsimile will also be somewhere on the internet.

It lists and gives colour patches of colours such as: "Lush Green Grass = Photographic Grey". "Fire Engine Red = Half Stop Open" (I think). "Road Asphalt = Two Stops Open" (I think). "Concrete = Half Stop Closed" (I think).

The "I thinks" are what I remember and are not necessarily accurate, but provided for a better grasp of the concept.

Such a chart allows for spot or area metering of a part of the scene: In essence I discarded my rote memory of most of the chart, using only two: "Lush Green Grass" and "Caucasian Skin", as I usually had grass outside and I always had the palm of my left hand.

WW

[revi]

Isn't that warning about reflectance a leftover from film days (esp. slide film)? If exposure is critical to within half a stop,
taking into account the reflectance of the target you measure on is important, esp. with (semi-)spot metering.

And there might be situations where measuring incident light gets tricky (e.g. lion under a tree...)

[Rufus]

As far as I can tell my Canon 6D Mk2 has an option for 1/3 or 1/2 stop exposure increments for aperture and shutter speed, but retains 1/3 stop increments for ISO. That seems inconsistent to me.

1/2 stops make calculations easier, but 1/3 stops all round makes counting clicks easier.

My default is to select 1/2 stops and then only change from ISO 100 if I have to, in which case it does not matter whether incremental ISO steps are 1/3 or 1/2 a stop.