User talk:GordonMcKinney

From Wikipedia, the free encyclopedia

[edit] Proposed way forward for a Wiki Friendly Photog Calculator

Proposal: To make an open source webpage written in Javascript to tabulate photography data in a form suitable for typesetting in Open Office and Microsoft Office.

Goal: To make the JS easy to maintain for new data sets (i.e. flash guide numbers etc). To ensure the code is open source to allow for customization by future contributors. To avoid some of the pitfalls of existing calculators which narrowly tackle single aspects of photography with confusing presentation.

... Please throw some ideas around :-) --GordonMcKinney 17:09, 9 January 2007 (UTC)

[edit] Cheat Sheet again (Archived)

<snip & archived>

Whichever way you go, you are free to ask others to decide to link your work. It is not appropriate to link it yourself, as I pointed out on WP:EL. Dicklyon 18:03, 8 January 2007 (UTC)

OK, so if I further indicate the ISO 100 nature in the left hand table then it will aid in understanding despite being non standard. As for the exposure value page, yes, that quote plus the EV table both use ISO 100. So I suppose we're talking about a de-facto standard rather than an official standard. As for another user linking to my sheet, could I nominate you after a period of peer review as you clearly have a strong understanding of the subject matter? I'd be happy to add other strobes (ones you think useful) and make any other useful changes. You clearly have a strong passion for optics and photography, why don't we work together. --GordonMcKinney 00:45, 9 January 2007 (UTC)

I'm OK with working together, but I don't think an external PDF cheat sheet is what we need as far as improving wikipedia. If someone else thinks so, they can argue to add it. I don't like the way your table works, either; if you start by looking up the ISO-100 EV, then you have to follow along diagonals, looking both ways, to find a pair of shutter speed and aperture that you like. There ought to be a much easier way, for example using a slide-rule-like mechanism as on most light meters: set ISO against the ISO-100 EV, and then all appropriate pairs of shutter speed and f-number show up adjacent to each other; no scanning needed. Dicklyon 04:19, 9 January 2007 (UTC)

Like these. Dicklyon 04:22, 9 January 2007 (UTC)

That is certainly easier to work with (no scanning) however it requires moving parts, making it harder to simply print and pack. But it is certainly something for me to chew on. I'll try and think of a more innovative way to present the data. The flash tables need a little refresh as they are small and hard to look stuff up on. One question however... my use of a PDF is simply to control the presentation of the data, and since my crib sheet is intended to be a tool I expect it will always be a PDF. My intention is for it to complement the Wiki rather than take precedence over the photography entries. In the end I want this work to be free / open source and helpful to the community. --GordonMcKinney 16:53, 9 January 2007 (UTC)

Follow up, I have now changed my PDF cheat sheet to use the term Light Value (LV) instead of EV. If you have time take a quick peek here Cheat Sheet --GordonMcKinney 22:17, 9 January 2007 (UTC)

Using Light Value rather than EV is a giant step backward; LV as the equivalent of EV at ISO 100 is nothing but urban legend. It never has had any official sanction; someone (perhaps inspired by Andy Eads's 2000 article in Photo Techniques) made it up, posted it somewhere on the web, and other web lemmings rushed to propagate it. If you look at the WP article on Light value, you'll see that it has had so many different meanings that it is essentially meaningless.

I think it's important to distinguish between the definition of EV

and its use as a measure of luminance or illuminance. ANSI PH 2.5-1960, which proposed APEX, equated EV to both the left and right sides of the exposure equation, implying use a a measure of luminance or illuminance as well as a combination of aperture and exposure time. However unwieldy some of us see this, the practice has been almost universal among camera and meter manufacturers for over 30 years, so it is a de facto standard.

It's nonetheless important to recognize that direct relationship of EV to luminance or illuminance has limitations. For measuring daylight outdoor scenes, it usually works fine; in other situations, such as outdoor night scenes or twilight or nighttime scenes that include the Moon, the relationship breaks down. The commonly recommended EVs for such scenes do not correspond to the integrated scene luminance, and except for scenes illuminated by moonlight, reference to illuminance usually makes no sense at all. The relationship between EV and luminance might still theoretically apply if one were to make measurements of individual scene luminances and apply the Zone system to arrive at the exposure. I do this when photographing the Moon, but I use calculated values of luminance because I cannot measure the Moon. I know of no way to measure or calculate the luminance of a street light a mile away. In the general sense, then, tabulated lists of EV simply mean "use these exposure settings in these situations."

That said, I'd still stick with EV for ISO 100 on the Cheat Sheet; although it's technically imperfect, it's in accord with photographic equipment manufacturers' practice. It also avoids perpetuation of web-borne urban nonsense. As I recall, some of Dick Lyons's early comments on the Light value talk page showed the same skepticism. I would make one change, however: get rid of the illuminance values for each EV. Although these values probably are OK for the higher EVs, they don't really apply to some of the night scenes. If you insist on keeping illuminance values, at least be consistent and give them all as klx, with a footnote somewhere for the non-technical reader who may not know what klx means.

The application of EV to practical photography remains a problem; it would be nice if you could set EV on a camera, but except for some Hasselblad lenses, you can't. I certainly agree that the examples Dick cited give the easiest means of application, but a calculator dial is a bit of stunt for a PDF file. I personally use the calculator dial on one of my Pentax spotmeters when applying calculated Moon luminances to camera settings, but this isn't a practical option for most photographers, so a table, despite its limitations, would seem to be the most viable option.

The original issue related to the appropriateness of including the Cheat Sheet in external links to some of the articles. My position is somewhere between Dick's and Gordon's. I think the Cheat Sheet is potentially useful, but I also agree with Dick's comment that it includes so much information that its relationship to WP articles to which it is linked is not always obvious. I think it's probably appropriate for articles that relate to exposure determination, but probably less so for other articles.

In any event, I'm strongly opposed to linking a version that uses LV to mean EV at ISO 100—there simply is no such thing, and we do every Wiki reader a disservice by perpetuating such nonsense. JeffConrad 00:59, 10 January 2007 (UTC)

My head is spinning. I can't seem to escape :) Dick is saying EV is a floating value and therefore can't exist on my cheat sheet as a representation of absolute light levels. LV is a creation which seems to fit a hole which exists between LUX and EV ISO 100. I'm happy with ISO 100 EV but Dick (unless I misunderstand) isn't. So I'm in a mess. Can we discuss a bit further. As for all the extra data... I'm thinking of dropping the handheld shake and focusing on provable data. As for Lux I am flexible and with more real estate on the 1/2 letter (A4) page I could indicate this more precisely. Love to hear more from Dick and Jeff on this subject. --GordonMcKinney 01:22, 10 January 2007 (UTC)

OK, maybe the question is... what is the correct variable name for ISO 100 EV... if none exists can we invent one that is non controversial. Like "Absolute Exposure Value (AEV)" or "Static Exposure Value (SEV)" as EV seems to be a relative variable. --GordonMcKinney 01:25, 10 January 2007 (UTC)

Also, I'll roll back my change to EV ISO 100 or something similar if we can reach a consensus. --GordonMcKinney 01:27, 10 January 2007 (UTC)

Head spinning is good. It means you're listening to the issues. I wrote the below before I read yours, and ran into an edit conflict:

Jeff, good point. But is it any better to use EV to mean EV at ISO 100? I think not. Why does one want EV at all for this purpose? It would work just as well to make up arbitrary symbols for those illuminance levels, wouldn't it? How about letters instead of numbers? Then, EV could be used the normal way. For example, one could make a table to map the situation and the ISO to an EV, and another table to map that actual EV to AV and TV. That is, move the ISO dependence into the table on the cheat-sheet's upper right; things would work much easier, and conform to the standard as well. Dicklyon 01:29, 10 January 2007 (UTC)

OK, so I think we've landed on the problem. There isn't a logarithmic currency to handle absolute light levels. Which I guess is why ISO 100 EV is an unofficial (defacto) standard. So how to crack the nut... can we call it EV100 and add a proposed entry to the wiki? I've never been convinced by LV as it's never been in a textbook that I've read. I prefer to indicate the EV variables are tied to ISO100 (EV100) than to make up a new 'technical term'.

I agree. Use of EV as a measure of luminance or illuminance follows directly from ASA PH 2.5-1960, and EV for ISO 100 is used almost universally by photographic equipment manufacturers, so I see little value in devising a new system. It does need to be mentioned that the EV is for ISO 100 (which was my original recommendation, as I recall). I've used the symbol EV₁₀₀, but this still is coining a new term, so simply stating EV for ISO 100 might be less confusing; again, this what equipment manufacturers do. Strictly speaking, the incident-light or reflected-light meter calibration constants also are required, but this might be going too far for the average photographer. I once used EV_{100 / 14} (for a reflected-light K = 14) in a talk, and pretty much lost everyone.

I don't see much value in using a code of letters (in the manner of the Black Cat and similar exposure guides)—it's just another level of abstraction that a reader must decode.

I think much of the problem is avoided by simply eliminating the illuminance values from the table, which then becomes just a list of recommended camera settings. I'm not sure how to cite the basis for the values given; the reference to the "Treganza average diffuse sky illuminance formula" is pretty obscure, and would seem to require at least a reference to a web site. All links to Tregana at LBL seem to have vanished; the sites I briefly examined appear to be pretty much the CIE standard sky. The topic is a lot to cover, because illuminance is a combination of direst sunlight and diffuse skylight. I'm not sure I could come close to doing it in one paragraph.

In summary: I'd eliminate the values of illuminance, and make certain to state that the EVs are for ISO 100. JeffConrad 02:28, 10 January 2007 (UTC)

Perhaps we could regard EV at ISO 100 as yet additional proof that nature abhors a vacuum. Right or wrong, though, the precedent is long established. Again, we also should bear in mind that not every recommended EV directly relates to the scene illuminance or integrated scene luminance. Ultimately, an EV table says "use these camera settings in these situations," which is all the average photographer wants to know. I use such a table in this manner when I can't adequately meter a scene, such as a nighttime skyline. JeffConrad 02:50, 10 January 2007 (UTC)

Sheet Updated: LUX removed, abbreviation of EV @ ISO 100 written as EVi100 or EV@ISO100. Since abbreviations are commonplace in mathematics and living languages I don't think it will be confused with a new piece of technical jargon. As for the CIE model, I found the definition here. As with all my tables I have tested the results with a light meter and camera in multiple situations to ensure calculation errors and assumptions haven't got the better of me. I still want to enhance the cheat sheet once we reach a level of consensus. --GordonMcKinney 04:36, 10 January 2007 (UTC)

I find the EVi100 confusing—I'd stick with EV₁₀₀ or EV100. The meaning is clear enough if you define it as you did for EVi100.

The CIE sky model you cite is one with which I'm familar; the latest version is given in ISO 15469:2004/CIE S 011:2003, which includes numerous sky types. The clear sky model (probably the most useful) hasn't changed much since 1967, though. I've only read the draft version, for which I did not need to pay. The cited web page still appears to include only sky luminance and not sun + sky, so I'm not sure it's the appropriate reference. The best source still appears to be the ANSI photographic exposure guide, PH2.7-1986, which I used as the primary basis for the table in the Exposure value article. It's long out of print, but as of 9 January 2007, printed and downloadable copies were available from IHS (www.global.ihs.com). JeffConrad 05:31, 10 January 2007 (UTC)

Actually, since you've used EV@ISO100 in both table headers, there is no need to introduce EV100 (or EVi100, or whatever) in the notes. I'd keep it simple. JeffConrad 06:21, 10 January 2007 (UTC)

Take a look at this temp hack suggestion of how you can use standard EV and make it easier to use the chart, too. Ignore the 100 that I accidentally left in the upper-left box. Dicklyon 06:32, 10 January 2007 (UTC)

It's certainly an interesting approach. It avoids the potentially confusing need to adjust a tabular value for ISO speeds other than 100. The shutter/aperture table is the traditional form seen (at least in the old days) in many camera manuals. The Wiki Exposure value article originally had the table in this format; the table was changed to the current format in response to edits and comments by Marc Lacoste. I'm not sure one format is better than the other; each may be easier for certain photographers to follow.

It's often helpful to show several different approaches to a topic such as this one, which initally confuses almost everyone. I think either this or Gordon's original format is workable, and either would provide an alternative to the presentation in the Wiki article. One advantage of Dick's suggestion is that it affords equal protection to aperture-priority types (such as myself, in most cases), but I don't really a strong preference. Personally, I'll probably stick with the calculator dials on my Pentax meters, but that's a mighty expensive approach for someone who doesn't already own such meters. JeffConrad 07:34, 10 January 2007 (UTC)

Food for thought. I like the idea of two entry points into the data. Since I want to juggle my cheat sheet to condense the hand held shake (or remove it). I think this is a opportunity to work the data up using shutter and aperture priority tables. Let me have a crack at an example and we'll see what everyone thinks. Should have it done today (I'm on GMT-6h). --GordonMcKinney 12:50, 10 January 2007 (UTC)

As for the Sky models... I found that set of equations works very well for Texas and reviewing data from old pictures of London, it matches that too. You'll note I subtract two stops to create a margin which I confirmed with my light meter, as the equation assumes 180deg coverage, which isn't always present in a scene. Anyway, I'm hoping the user treats it as a guide rather than gospel. --GordonMcKinney 12:50, 10 January 2007 (UTC)

Here's a draft... my main concern is how much space it takes up without the additional tables required to describe typical light conditions. While this is fine for a Wiki page, it doesn't satisfy the 'camera bag' reference card that I designed my cheat sheet around. As a compromise, what if I turned my exposure table into Wiki HTML and added it to the Exposure Wiki page only? Very open to ideas :-) --GordonMcKinney 14:57, 10 January 2007 (UTC)

The Exposure value article already has a table similar to the expanded tables. I think either your original format or Dick's suggestion would be easier to use in the field, in keeping with your original objective. Moreover, either would provide an alternative to the Wiki presentation, and make a good argument for inclusion of the link. JeffConrad 00:49, 11 January 2007 (UTC)

I think I need to build out my stub table (lighting conditions) using commonly accepted terms. Then I'll rework the layout to use the sliding ISO scale next to the EV numbers as per Dick's diagram. --GordonMcKinney 13:33, 11 January 2007 (UTC)

[edit] Survey Invitation

Hi there, I am a research student from the National University of Singapore and I wish to invite you to do an online survey about Wikipedia. To compensate you for your time, I am offering a reward of USD$10, either to you or as a donation to the Wikimedia Foundation. For more information, please go to the research home page. Thank you. --WikiInquirer 13:24, 3 March 2007 (UTC)^{talk to me}