User talk:JeffConrad

From Wikipedia, the free encyclopedia

Contents 1 APEX 2 Light value 3 Re: DOF article 4 Quotes 5 Depth of field FAC prep 6 Entrance pupil 6.1 Equivalence of equations

[edit] APEX

Jeff, good job on the APEX and Lv stuff. Thanks. Dicklyon 04:40, 20 July 2006 (UTC)

Thanks, Dick. I've made a corresponding edit of the entry for exposure value.

Jeff, why the "criscam" link? I don't see anything about EV there. By the way, I've used the Kyoritsu light box and patterns. I understand one can use it simulating lux levels, testing meter calibration, etc., but I still don't see how it's a usefil artical external link. Looks more like link spam. Dicklyon 21:55, 19 August 2006 (UTC)

Dick, there actually is mention of EV, LV, and K in the C.R.I.S. link, under "kyoritsu test equipment," so I've restored the link. The link was included simply to show that I didn't pull the Kyoritsu references in the test out of the air. I hope this addresses your comment.

JeffConrad 22:05, 19 August 2006 (UTC)

Thanks, Jeff. The other way to do this is to put the link in single-brackets at the end of the statement that needs a citation. The external links sections is for more general further reading and such. See also my comments on the EV talk. Dicklyon 22:14, 19 August 2006 (UTC)

I used the current approach for two reasons: 1) All other in-text references use author-date, and 2) the C.R.I.S. link isn't really a direct reference, and I felt an in-text link that points only to the site might be confusing. JeffConrad 01:59, 20 August 2006 (UTC)

[edit] Light value

This entry still isn't complete (or quite correct). The APEX section in Stroebel, Compton, Current, and Zakia (2000) is more than a bit confused: the text refers to "scene illuminance," and Table 2-3 on the next page uses units of illuminance (with the curious symbol Bv), but the example uses units of luminance, and the defining equation uses the symbol Lv. Although it probably was not the intent, this reference technically uses "light value" as a synonym for both incident-light value and luminance value. Consequently, I think we need to restore mention of incident light. I'm ready to do this, but I'd like to be sure that the usage in Zakia and Stroebel (1993) also is correct. I'm wondering if the table in Stroebel et. al. (2000) was copied from from Zakia and Stroebel (1993). I don't have a copy of that reference--do you (I noticed that you added it)?

Jeff, is this comment you wrote to yourself intended for me? The history says someone else added those refs, but I think I have the 1993 encyclopedia (I'll check in the office on Monday), and the other is available (partly) on google book search: [1]. Dicklyon 03:44, 20 August 2006 (UTC)

Dick, indeed ... I forgot the salutation. I also wasn't too clear in the note--I added those references to "light value" when I removed all discussion of "light value" from the "APEX system" article. What I meant to say was that it looks like you added Zakia (1993) to the APEX system article on 25 June 2006. I seem to be getting as confused as the meaning of "light value" ... JeffConrad 07:01, 20 August 2006 (UTC)

That sounds right, then. That must be the edition I have at work then (I also have a much older hardbound edition, if that's of any interest). What shall I check? Or should I send you a copy of relevant pages? Dicklyon 07:33, 20 August 2006 (UTC)

Dick, given the variety of meanings, it would be interesting to see if the new and old versions agree. I'd love to see the relevant pages, but providing them seems like a lot of extra work. It would seem to me that three issues might be relevant:

1. Did the definition(s) refer to incident or reflected light?
2. Was the definition consistent with APEX (either incident-light value or luminance value)?
3. What symbol was used (I_v, B_v, L_v, or even something else)?

JeffConrad 23:36, 20 August 2006 (UTC)

1. "light level in footcandles divided by 6". That's luminance, reflected, right (candela per sq. foot)?
2. It's consistent with APEX luminance value Bv.
3. Bv.

Basically, they treat Light value as a synonym for Brightness value, and that's all. The entry was written by a J. Johnson. Their entry on the Additive system also says "Light value (Bv)". Dicklyon 17:06, 21 August 2006 (UTC)

Dick, thanks for the info. Footcandles are units of illuminance. The expression B / 6 is consistent with that, and is the same expression used in the table in Stroebel et. al. (2000). In other words, this definition is equivalent to incident-light value; inexplicably, they use the symbol (B_v) for luminance value, when I_v would have seemed more appropriate. What a mess ... If this can confuse the likes of Stroebel and Zakia, I shudder to think what it does to most others.

Even the mention of "illuminance" is potentially a can of worms, because it properly refers to flux density incident upon a flat surface, and is properly measured with a flat (cosine-responding) sensor. The median APEX value of 20.8 for the incident-light calibration constant (and the expression B / 6) also implies a flat sensor. This, of course, usually isn't the best way to take an incident-light reading of a typical scene. The more commonly used "illuminance" measured with a hemispherical sensor requires a different calibration constant and a different expression for relating illuminance to incident-light value. I'm not so sure I want to open the can at present, so I'm disinclined to even mention the sensor type. The same issue applies to the exposure value article, because EV can be a measure of illuminance as well as luminance. I somewhat cover this in an article (in PDF) on the Large-Format site, though this issue is buried in the other content. I could add the link to the Wiki article, but I'm not sure whether it would be more likely to inform or confuse.

Incidentally, it's obvious from the APEX equations that B_v and I_v are numerically equal, so that the same camera settings will result from using either. Nonetheless, the physical quantities are quite different, and I find the accompanying imprecision discomforting. NPOV and "no original research" preclude tagging "light value" as deprecated, but a comprehensive factual treatment hopefully will lead many readers to that conclusion.

I'll try to describe the confusion in a tactful way in the new edit. I think it's worth pursuing this until we get it right because the primary sources (e.g., most of the ASA and ANSI standards that I've cited) are becoming increasingly difficult to obtain. JeffConrad 02:02, 22 August 2006 (UTC)

Suffice it to say that "light value" has had so many meanings that essentially it is meaningless, and would seem an inadvisable substitute for "EV at ISO 100," however inelegant that expression may be. JeffConrad 01:59, 20 August 2006 (UTC)

[edit] Re: DOF article

Hi Jeff!
Thanks for bringing that up, I think this is what you were after? Image:Wolf spider single frame.jpg --Fir0002 22:46, 19 December 2006 (UTC)

I'll add this to the article—I think it will make the benefits of compositing more obvious. I assume it was taken at f/11; do you know the approximate magnification? JeffConrad 00:26, 20 December 2006 (UTC)

Whoops sorry Jeff, I'm used to people replying back on my talk page. Anyway doesn't matter. Image was taken at f/11, not sure what you mean by magnification, but it was almost at my minimum focal distance (1:1) --Fir0002 23:09, 23 December 2006 (UTC)

Magnification is another way of stating reproduction ratio; it's the ratio of image to the object size. If, by your reply, you indicate a reproduction ratio of 1:1, the magnification was 1.0.

Magnification is more common than reproduction ratio in DoF formulae. Knowing the magnification here is informative because it illustrates just how shallow the single-shot DoF is. Look at the formula under Closeup DOF (under DOF formulae); assume, for sake of argument, that an appropriate value for the circle of confusion c with your 20D is 0.020 mm. Then, for a symmetrical lens,

The need for the multiple exposures becomes quite obvious. If, indeed, the magnification was approximately 1, I'll add mention of that to the DOF article. JeffConrad 00:06, 24 December 2006 (UTC)

[edit] Quotes

I don't want to get into an edit war, but I can't see what possible advantage the existing quotes at light value could have. For one, the punctuation is incorrectly located, according to WP:MOS (see also WP:REF for reference punctuation). Two, it makes reading the markup more difficult (necessary markup is one thing; obfuscation is another). Three, the article isn't even internally consistent (normal non-curly quotes were already used in several places). If you feel strongly that the curly quotes must be used (personally, I loathe them, but as there's no consensus either way...), at least consider using the toolbar characters (“”) for readability and fixing the punctuation location. --Fru1tbat 03:56, 26 January 2007 (UTC)

Note: I posted that before I read your comment on the article's talk page, so I concede that I wasn't considering print appearance as a possible advantage... --Fru1tbat 04:10, 26 January 2007 (UTC)

A quick glance at many WP articles appears to reveal a strong preference for neutral quotes. However, this would be considered bad typography by almost any standard. The neutral quote had a relatively brief lifetime—from the invention of the typewriter in 1870 until whenever one considers the date of the typewriter's obsolescence. That the appearance of the neutral quotes is reasonable with the current Wikipedia sans-serif font is coincidence. Were Wikipedia to change to a different font at some time in the future, the appearance might not be as acceptable, even with some sans-serif fonts. With almost all serif fonts, the appearance would be unacceptable.

I'm no more fond of typing (or reading) HTML entities than you are, but they do have some advantages, one of which is that they seldom cause trouble. There probably is something to be said for putting everything in Unicode (as with inputting the characters from the links under the edit box), but there also are some difficulties. It's a bit of a pain to reach for the links, and doing so is not possible when composing the edit with an external editor (as I often do. Wikpedia's editing facilities might kindly be described as limited ...). There ought to be an easier method of inputting opening and closing quotes; 30 years ago, troff recognized `` as an opening quote. Admittedly, Wiki's use of '' to indicate start and end of italics presents a bit of an obstacle.

It looks as though I forgot about WP's use of international rather than U.S. practice for positioning punctuation. However, that article also states

However, insisting on changing to this usage, especially in articles written in American English, is deprecated; there are better and less divisive uses of time.

With all due respect, this would seem to border on a "Pointless edit," which you specifically disavow.

Good point, and I try not to change/fix punctuation unless I'm fixing something more important with the same edit, but the problem with the term "pointless edit" is that it's too subjective. The line you quoted from the MoS was added recently, IIRC, and I think it unnecessarily weakens the guideline. I believe making things like punctuation and grammar consistent across Wikipedia contributes significantly to making it appear more "professional", and that's important to me. That being said, there are certain instances that I consider more egregious than others. The WP:MOS-T section on punctuation is worded with less room for variation, for example.

I have no real objection to your changing position of the punctuation (or changing the HTML quotes to Unicode), but I have better uses of my time. I hate to invest more time preparing for possible code maintenance than I'm likely to actually spend on code maintenance. In the future, I'll follow the WP guideline on punctuation and quotation marks when starting from scratch. In terms of existing material, though, this would seem more suited to a bot. It would not surprise me if such a bot already exists. Many years ago, I had the equivalent of a bot that would put commas and periods inside quotation marks. JeffConrad 06:12, 26 January 2007 (UTC)

I made the edit because I happened across the article and wanted to try to improve it, and I only complained because I felt that reverting did more harm than good, but it won't bother me all that much if it's not "my way". I have a text editor that supports the Unicode characters, and if you want, I could do a quick global replace to at least substitute for the HTML code, but I'd be content to leave it alone. --Fru1tbat 14:02, 26 January 2007 (UTC)

I agree that consistency makes articles look more professional, but I'd put consistency of punctuation (American vs. international style) far behind differences between British and American syntax and spelling. Even more significant is inconsistency of style among multiple authors. It's a minor factor in the Light value article because I've written most of it, and revised the rest for consistency. The Depth of field article is another matter; there have been so many contributors (of widely varying technical and writing ability) that it's more of a hodgepodge than an article. In my opinion, it doesn't deserve an A rating in its present form.

In any event, I've changed the positions of commas and periods in the Light value article to conform to the MoS. It takes more effort to argue than to change it. Although it's not a major issue, I'd rather leave the quotes as they are; I also could make global substitutions, but the Unicode quotes make future edits more difficult for me (some other editor functions don't handle Unicode). JeffConrad 19:59, 26 January 2007 (UTC)

[edit] Depth of field FAC prep

Jeff, I just wanted to say thanks for the great work on the article. It looks to me like it could very easily be turned into an FA-class article in a short time. I'm not the expert on the topic that you are, but all I can see that is deficient at the moment is the amount of inline citations. What are your thoughts, and would you be willing to shepherd the article to the gold star? :) Girolamo Savonarola 22:43, 10 September 2007 (UTC)

Girolamo, I think the article is OK, but I don't know if quite share your high opinion of it. There have been a few additions about which I am not enthusiastic, and there is at least one link to a WP article that I think has seriously flawed technical content. It's also a bit embarrassing when someone feels the need to use Maple to check a fix to a typo (actually, I had botched a copy and paste) in a formula—I probably should carefully review all formulas to ensure that there aren't other errors. I'd also feel better if at least one other significant contributor (e.g., Dick Lyon) has the same confidence in the article.

One question: what is your concern about the inline citations, and how should they be fixed? Some of the hyperlinks are simply to the References section. JeffConrad 23:50, 10 September 2007 (UTC)

Well, as I said, you know the topic far better than I could possibly aspire to, so I'll have to take your word for it. I've already spoken to Dick, and told me that he didn't feel up to the task of leading this to FAC himself and recommended you (although I'd messaged both of you simultaneously). I have no idea what his thoughts on its state are, however.

The inline citations need to be handled, well...inline. Having external links inside the body itself is generally considered bad style - they need to be replaced by ref tags, which can include the links inside there. You can see Wp:cite#Footnotes for more details. Girolamo Savonarola 22:52, 11 September 2007 (UTC)

I've asked Dick for his opinion.

I see only a handful of external links in the text (links to Jacobson's, van Walree's, and my writings—are these the links to which you refer?); it would be a simple matter to remove them (the links are in External references anyway). Links such as (Langford 1973, 81) in Limited DOF: selective focus can't really be changed, though; the article uses the author-date system (“Harvard referencing” to some Wikipedians) of references plus footnotes (via ref tags)—see Maintaining a separate "References" section in addition to “Notes” in the article you cited. Converting the author-date references to ref tags would merge the two, making a real mess of things. The author-date links used work to much the same effect as Template:Harv and Template:Citation, though the formatting of references follows the Chicago Manual of Style (and admittedly, the inline HTML looks a bit ugly to the editor). JeffConrad 00:14, 12 September 2007 (UTC)

Well, I'll leave it to you; it's just something I anticipate may be mentioned by reviewing editors. You might want to also run the article by the general peer review for preliminary comments and touch-ups. As for content deficiencies, as I've mentioned, I'm not equipped to gauge them, but please be bold and tackle what you can see. I look forward to seeing what develops! :) Girolamo Savonarola 00:22, 12 September 2007 (UTC)

[edit] Entrance pupil

Jeff, I don't mind that you took entrance pupil out of my DOF edit, but your reason is wrong. The exit pupil size and position determine depth of focus; but the entrance pupil size and position determine the depth of field; this is most obvious in the object field method. Of course, they turn out equivalent, too, since the f-number is the same whether you base it on one or the other, no? Dicklyon (talk) 01:04, 8 June 2008 (UTC)

Dick, I'm not sure I agree. The exit pupil is the center of projection, and the only view of the aperture seen by the image plane. I think I show this fairly clearly in the section Depth of Field for an Asymmetrical Lens in my paper linked in the DOF article. The ultimate expression commonly used gives the DoF in terms of the entrance pupil (via the f-number), but the entrance pupil diameter multiplied by pupillary magnification gives the exit pupil diameter. If the exit pupil weren't the governing factor, we could ignore the pupillary magnification in closeup DoF equations. Sidney Ray also covers this in the work linked in the article, but doesn't specifically derive the DoF for an asymmetrical lens, so you need to examine material in Sections 14.1.5, 22.2.1, and 22.4.1. David Jacobson implies the same in defining effective f-number in terms of the eXit pupil. Ultimately, the DoF is determined by the aperture diameter, regardless of how we choose to describe it, so we're safe with the current wording.

In any event, I think the article is more accessible to the average reader with the revision to emphasize aperture over f-number. Also, in checking Jacobson's tutorial, I discovered that it has moved (to a version revised in 2007), so I fixed the link. JeffConrad (talk) 03:23, 8 June 2008 (UTC)

Well, take another look at my DOF draft and tell if you don't agree that the outside-the-box is much simpler, and has to be equivalent to any other method that works. The article is fine as is. Dicklyon (talk) 05:52, 8 June 2008 (UTC)

Actually, I don't find it simpler, but it may just be a matter of personal preference and what I am accustomed to. I still don't see how one can avoid mention of the pupillary magnification when the aperture is given in terms of the f-number (or the absolute entrance pupil diameter), because without the pupillary magnification, one simply cannot arrive at the universally accepted closeup DoF equations. Perhaps, though, it is simply as you indicate in your paper

The formula 2SC/md is exact, as long as S is measured from the entrance pupil, but the equations with N may need adjustments.

Van Walree, David Jacobson, Sidney Ray, and I all come up with the same closeup equations, and absent a vast conspiracy, it's unlikely we'd all make the same mistake. I'm sure that with a little mathematical rearrangement, it's possible to show that your equations are equivalent to mine. I shall leave that as an exercise for the reader :-).

In any event, I agree that the article (or at least this section) is fine as it is. For most readers (and most photographic situations), the distinction among aperture, entrance pupil, and exit pupil is irrelevant and quite possibly confusing. JeffConrad (talk) 07:57, 8 June 2008 (UTC)

I'm not claiming the equations are equivalent; I suppose it's possible, if your focal length, f-number, and pupillary magnifiation imply an entrance pupil position, but I'm not at all sure; I am saying, however, that they should give the same answer, for the same physical setup. If my equation is correct and exact as I claim, you can see why I claim that using entrance pupil diameter is the simplest approach (as von Rohr had concluded before me); it leaves you without any worries about f-number, focal length, pupillary magnification, or anything else about the lens. So, let's cook up a real example and see... Dicklyon (talk) 14:53, 8 June 2008 (UTC)

I don't see how our equations could not be equivalent unless one of them is wrong. Pupil magnification indeed implies pupil locations relative to the nodal planes; I illustrate this in Figures 2 through 4 in my paper (the figures aren't strictly geometrically accurate; I moved some things around a bit to make it easier to label the distances, but my analysis doesn't depend on geometrical accuracy of the diagram).

How would we cook up a real-world example? I suspect that with either a symmetrical lens or large camera-to-subject distance, we'd probably get the same answer, because the object nodal plane, entrance pupil, and exit pupil are all in the same place. With a highly asymmetrical lens at closeup distance, our results would probably differ. To be honest, I've never found calculations of much help in closeup work, and pretty much do everything visually. JeffConrad (talk) 03:45, 9 June 2008 (UTC)

Agreed. Still, it would be fun to check the equations on a close-up asymmetric lens problem, and see if they give the same answer. I expect they will, and if they don't then we'll learn something. Maybe one of us got something wrong? Maybe not? Of course, if they match that doesn't imply anything strongly, but it suggests that maybe we both got it right. Dicklyon (talk) 04:11, 9 June 2008 (UTC)

[edit] Equivalence of equations

The equations are indeed equivalent, as they had to be. From my Figure 4 on p. 21, your distance S is

S = u + x_ep ,

where x_ep is the distance from the front principal plane to the entrance pupil. Using my Eq. (81) for x_ep, and making appropriate substitutions for S and d in the difference of your equations for S_F and S_N gives my Eq. (104), which reduces to Eq. (105) when the subject distance is much less than the hyperfocal distance. Direct substitution into your approximate closeup expression 2SC/md also gives my Eq. (105). Note that c in my Eqs. (104) and (105) is the same as your C.

So is DoF determined by the entrance pupil or by the exit pupil? I guess it depends on how you approach it. The safest approach is probably to say that it's determined by the aperture, as the article currently does. JeffConrad (talk) 07:19, 10 June 2008 (UTC)

I agree there. My preference for the entrance pupil approach is based on the fact that it makes it explicit that everything behind that can be safely ignored; the estrance pupil of a camera can be measured from the outside by looking at it, and if you also get the field of view (say from a test shot or electronic preview), then you have all you need, and can safely ignore focal length, f-number, and pupillary magnification. And the equation is super simple, and very easy to derive. But practically, you can get a good result with the "inside" parameters you're usually told (focal length, format size, f-number), and can usually get away with ignoring pupil magnification except in extreme cases. As for "equivalent", not quite. Your equations imply mine, as you showed; but mine don't imply yours since mine don't have enough parameters to describe the more specific cameras that yours apply to; but given some extra stuff, yes, they come out equivalent. Thanks for checking. Dicklyon (talk) 14:52, 10 June 2008 (UTC)