Sunny 16 Rule

[xpatUSA]

Re my cameras' ISO differences, it's not just my copies of them, it's as measured by DXO and often appears to explain the amount of compensation I need to use to get good exposures when using Sunny f/16 in bright sunny conditions.

Adding some ISO tech. to what Bill said:

On a general note, the ISO number can vary considerably in it's results between models, something like 10% (+/- 0.14EV) actual exposure for a given illuminance. But, this is on top of the much larger variation due to the methods of determination of the magic number for individual camera models.

There are three methods commonly used - with a fourth used but rarely.

They are:

a) Saturation-based
b) Standard Output Sensitivity
c) Recommended Exposure Index.

There is a 1/2-stop difference in "headroom" between a) and b).

c) is a bit like the time (whatever Chuck Norris says it is).

So it comes as no surprise to me that "the rule" can need tweaking to match a given DSLR model.

[rant] "They" should have left ISO with film and never introduced it with digital, IMHO. Instead, a single knob marked "LOSS" which showed how much the sensor was being under-exposed - just for people who like to shoot in the dark. No confusion there. Turn it up - get a crappy image.

Naturally, Marketing would relabel that to "GAIN" . . . . which has bigger numbers with increasing loss, just like ISO, bringing back the confusion so extant in the world of photography. [/rant]

Pardon my tongue-in-cheek rant: I hate "ISO" and all the expectations and confusion it causes.

[Pippan]

Thanks Ted, I didn't realise there was so much variation in testing methods. It looks like DxO uses a saturation-based method.

Actually your 'rant' made me wonder why highlights get clipped at higher than native ISOs, since the pixel wells are not actually filled. Maybe the limit is hit in the downstream electronics rather than the pixel well.

[dje]

Thanks Ted, I didn't realise there was so much variation in testing methods. It looks like DxO uses a saturation-based method.

Actually your 'rant' made me wonder why highlights get clipped at higher than native ISOs, since the pixel wells are not actually filled. Maybe the limit is hit in the downstream electronics rather than the pixel well.

Yes Greg I think it's either the ISO amplifier or the ADC that clips.

Dave

[xpatUSA]

Thanks Ted, I didn't realise there was so much variation in testing methods. It looks like DxO uses a saturation-based method.

Well, it's the camera manufacturers who choose what method to use to calculate the ISO number for their products, not DxO.

How DxO do their testing is unknown to me because I am no fan of theirs, sorry.

Actually your 'rant' made me wonder why highlights get clipped at higher than native ISOs, since the pixel wells are not actually filled. Maybe the limit is hit in the downstream electronics rather than the pixel well.

Yes, simplifying, there are two kinds of cameras:

a) where the 'raw data' is actually what the sensor saw (my Sigmas - but not all Sigmas) or

b) where the 'raw data' is obfuscated by a Programmable Gain Amplifier (PGA) stuck between the sensor output and the Analog to Digital Converter (ADC) input. Said PGA might, for example, have a gain of 1 at ISO 100, 2 at ISO 200, 4 at ISO 400, etc.

That means I personally can underexpose my Sigmas with ever-higher ISO settings if I choose so to do (which is never) without highlight clipping.

But, in case b), it is the ADC that gets clipped *** no matter what your ISO was set to! Which is why owners of ordinary cameras have to be much more careful than me to avoid highlight clipping.

*** Or the PGA, thanks Dave.

I won't even talk about case c), et subs, where the manufacturers do all kinds of stuff to the real raw data before it even gets into the so-called "raw" file . . .

[William W]

Hi Greg,

I did understand what you meant when you wrote that the exposure can be corrected with a digital file – but I thought that this is one case when being pedantic is necessary.

. . . Re my cameras' ISO differences, it's not just my copies of them, it's as measured by DXO and often appears to explain the amount of compensation I need to use to get good exposures when using Sunny f/16 in bright sunny conditions. . .

Yes. I concur.

I didn’t ignore that, I simply chose not to comment on it. I wanted only to comment on the F/16 Rule.

For example, (many, most, all?) of the Canon DSLRs’ interim ISOs (i.e. the ISO 125 and ISO 160 and multiples thereof) are not “real” IOSs: (there is a long Canon paper on it I can’t find it in my library atm).

Notwithstanding that fact Canon DSLRs’ ISO 100; 200; 400 . . . etc might not be precise either (don’t know) it would NOT have shown up in my testing of my Canon DSLR cameras against the F/16 Rule, when I used ONLY the camera's TTL Meter – because (I assumed) the Sunlight to be the constant. If any of my Cameras’ ISO 200 (which is what I usually tested) was a little bit out, then that error (assuming it to be a constant error for any camera model) would have been consumed by what I (and my San Francisco friend) attributed to that Camera’s "headroom".

***

. . . Re the rest of it, I'll have to digest it and do lots of testing with my cameras tomorrow. I probably, as you say, do not fully understand Sunny f/16 and maybe fail to compensate enough for scenes that aren't properly sunlit. I also take your point on the word 'exposure' in post-processing--poor choice of words, it isn't changing exposure just as altering ISO isn't altering exposure. Interesting that you say Sunny 16 is for the tropics (obviously I live between those two Tropics), I was unaware of that.

If you want to “test it” then you will require more time than just tomorrow (because you are measuring the comparative of the Sun’s illumination) and you will need at the least a Photographic Grey Card or “Lush Green Grass” lit by Front Lit Sunlight to fill your Camera’s Frame and you will need to set your Camera’s TTL such that it does NOT have any metering computation algorithm (probably doesn’t matter if there is an algorithm in the cameras TTL, provided the comparative tests always use the same Camera Metering Mode.

Also understand that you (being in Dar Win) will get less variance between Winter and Summer than I get down here and my friend got in California. We calculated Winter Sun needed about an extra hour and an half to get to full strength – for example, if we were attaining an accurate “F/16 Rule” at about 0900 in mid-summer, then we would need to wait until about 1030 in mid-winter.

I think that it is important to understand the concept of the F/16 Rule or at least appreciate the concept of the use of the F/16 Rule which I am using and also understand the TWO meanings of the term: “Correct Exposure”, as I am implying them.

The first meaning of “Correct Exposure” is purely technical: the chosen exposure can only be “correct” for one or more sections of the scene which is/are illuminated by the one same intensity of light. I often find students grasp this concept if they think of it in the same manner as thinking about the technical concept of Depth of Field – there is only ONE Plane of Sharp Focus in the scene, but there are DEGREES of the un-focused areas – similarly there is only one ‘correct exposure’ area(s) and then there are degrees of shadow and perhaps areas which will be blown out.

Then there is the Artistic meaning of “Correct Exposure” and this is when we examine and choose the exposure to use to make the whole image render our Artistic Vision of the scene – and that might include areas being blown out and it might include areas of deep shadow with little or no detail.

Perhaps an example will be easier:


“Stage Door”, Sydney circa. 1976

Let’s look at the scene that I previously used as an example image: In a technical sense I chose to use the “correct exposure” for the tiles on the wall – that was easy to do using the F/16 Rule and also it was what I wanted to do, because if I didn’t “correctly expose” those tiles, then I would have a big hunk of white overexposed mush at camera left.

We can now discuss the (different) Artistic question: “Is the exposure that I chose, the ‘correct exposure’ for the Scene and the main Subject?”

Well I think the answer is “yes” because that exposure fulfills the vision that I had for that particular photograph – two key artistic points being the anonymity of the Subject is secure in shadow and I don’t have a large mush of blown white but rather a purposeful juxtaposition to the Main Subject of an ordered system of neat, white, sunlit tiles aligned with but out of the area of the Main Subject’s space.

***

Another example:

“Mother and Child” – Sydney 2009

In this image the (technical) “correct exposure” was assessed for the skin tones on the two Main Subjects. They are illuminated by an hard light, a Street Lamp above and to Camera Left. Both Subjects have olive skin, the Mother’s skin is darker than the Child’s skin.

Considering that I have blown a bit of the white singlet top on the little Girl and that singlet top is illuminated by the same street lamp and is the same distance from that street lamp – then (technically) my “correct exposure” is incorrect: considering my 5D’s headroom I am about ⅔Stop overexposed for skin tones.

If we assess the image’s “correct exposure” in the Artistic meaning - then (hopefully) we would agree that we would not want the exposure of the whole scene increased too much, because if we did, we would have blown out faces and lots of nice detail of an old, large wooden area.

***


“Mimes” Paris 2012

In this third example – the (technical) “correct exposure” was made for the two Main Subjects, who happen to be lit by comparative soft light.

Was I concerned about blowing the (hard light) highlights from the store windows and the car in the background? – Definitely not.

WW

All Images © AJ Group Pty Ltd Aust 1996~2017 WMW 1965~1996

[William W]

When I was using it with color slide film and with consumer-grade zoom lenses and cameras beginning in the 1980s that weren't nearly as good as the equipment I now use, it was not unusual that the exposure would be off by about 1/2 stop. Indeed, I tested the accuracy of my equipment during the ideal Sunny 16 conditions you described and when filling the frame with green grass. Once I had documented the inaccuracy of my equipment, if any, I compensated from that point going forward in my application of the Sunny 16 rule. I doubt that any of that would have been necessary using pro-quality camera bodies and prime lenses, which at the time were far better in many respects than even pro-quality zoom lenses.

Glad you mentioned: filling the frame; Green Grass and consumer grade Zoom lenses – the latter implies the importance of considering T-Stops: I think that I have adequately explained the use of "filling the frame" and the use of “Green Grass” if one doesn’t have a STANDARDIZED Photographic Grey Card.

***

Yes Greg I think it's either the ISO amplifier or the ADC that clips.

Thanks Dave. Locked that information away

***

. . . On a general note, the ISO number can vary considerably in it's results between models, something like 10% (+/- 0.14EV) actual exposure for a given illuminance. But, this is on top of the much larger variation due to the methods of determination of the magic number for individual camera models.

Ta Ted.

Add to that my mention of the interim ISOs not being "real" (at least for Canon DSLRs). You will probably have an explanation for that off the top of your head – I choose not to keep the details of that stuff in my head, because it hurts too much.

***

. . . I hate "ISO" and all the expectations and confusion it causes.

“Use Prime Lenses and Zoom with your feet.”

“Buy a 7D, not a 5D and use the crop factor to get closer to the Subject”

I refuse to bring up the topics of: "Protection Filters" and "Perspective"

Hahahahaahaha!

WW

[xpatUSA]

Ta Ted.

Add to that my mention of the interim ISOs not being "real" (at least for Canon DSLRs). You will probably have an explanation for that off the top of your head – I choose not to keep the details of that stuff in my head, because it hurts too much.

It can certainly get tedious - and I have never owned a Canon and know little about them.

Adding a bit more, Doug Kerr muddied the Canon waters some time back with:

We believe that for these cameras the exposure meter calibration itself is closely in line with ISO 2721. Thus, we must conclude that Canon has rated the “ISO sensitivity” of their sensor systems at about 74% of what would be determined in accordance with ISO 12232—that is,a sensitivity that Canon rates “ISO 100” would probably be rated as ISO 135 if actually determined in accordance with ISO 12232 . . .

http://dougkerr.net/Pumpkin/articles...alibration.pdf

. . . meaning don't bother with adding the traditional half-stop if using a gray card instead of 'sunny 16'

[William W]

. . . meaning don't bother with adding the traditional half-stop if using a gray card instead of 'sunny 16'

Haha.

We can add that to the tedious discussion list too: the 12% / 18% discussion for TTL referencing.

The salient point is, for any reasonably accurate practical purpose: one needs to employ a reasonably sound procedure to 'reference' all the gear used in any one particular Workflow - then (and only then) does the matter of changing the Aperture and the Shutter Speed and the ISO have a continuing relevance.

WW