Jump to content

Talk:Depth of field

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Hyperfocal in the "Same Picture" section

[edit]

Assuming a lens with focal length and f-stop both modified by the same sensor format conversion ("crop factor"), I get the part where the equivalence breaks down as macro magnification becomes a big deal. But I don't understand the part about the equivalence breaking down at "as distance approaches the hyperfocal distance for the smaller format (the DOF of the smaller format approaches infinity)". As I calculate it, assuming also-scaled CoC, the hyperfocal distance should be the same for both formats. What am I missing? Thanks! Matthew Miller (talk) 02:45, 25 March 2011 (UTC)[reply]

What you're missing is that that comment is being applied to the condition stated as "With the same f-number." Afterward, the f-number ratio that makes the DOFs equal is computed, and that works fine at distance, as you note. But the DOF ratio itself doesn't work so well when one of the DOFs can approach infinity while the other doesn't. There might be a better way to put all this to avoid the problem, so feel free to propose a rewrite (be bold, and we'll jump in and fix it we don't like it). Dicklyon (talk) 03:32, 25 March 2011 (UTC)[reply]
Our DoF comparisons (like everyone else’s) make use of approximations at near and far distances; unlike most other sources, even Stroebel, we clearly describe what the simplifying approximations are at the beginning of the subsection. One is that the distance is much less than hyperfocal, so that as the hyperfocal distance is approached, the comparisons are no longer valid. We assume that focal length and CoC both scale with format; however, because the hyperfocal distance is given by
it is not the same for both formats, but rather scales approximately with format. I guess both Dick and I concede that the current wording may require a second read; however, at least we state what most other sources (except perhaps Bob Atkins) simply ignore. JeffConrad (talk) 07:49, 25 March 2011 (UTC)[reply]
Perhaps what’s not as obvious as it seems to me: the comparison becomes invalid when the magnification in either format approaches unity or when the distance approaches hyperfocal in either format. Because we assume magnification is proportional to format size, magnification approaches unity in the larger format first; because the hyperfocal distance is approximately proportional to format size, the distance approaches hyperfocal in the smaller format first. Perhaps we could state this, although to me, what we currently say in the article is more digestible than how I’ve just described it. JeffConrad (talk) 22:30, 25 March 2011 (UTC)[reply]

DOF - Increases vs Decreases

[edit]

Two of my previous edits were reverted on the basis that they are incorrect, namely changing "increases" to "decreases" in the Effect on Lens Aperture section. The way the facts are presented contradict information elsewhere in the article.

Opening Paragraph (bold added by me for emphasis)
"In some cases, it may be desirable to have the entire image sharp, and a large DOF is appropriate. In other cases, a small DOF may be more effective, emphasizing the subject while de-emphasizing the foreground and background."
Effect on Lens Aperture Section (bold added by me for emphasis)
"Reducing the aperture diameter (increasing the f-number) increases the DOF; however, it also reduces the amount of light transmitted, and increases diffraction, placing a practical limit on the extent to which DOF can be increased by reducing the aperture diameter."

This contradicts the first paragraph. From what I understand of DOF (and the first paragraph) is that put simply, a larger (deeper) DOF means more of the scene is in focus. A smaller (shallower) DOF means only a specific range of objects will be in focus. So by reducing the aperture diameter, it brings more things into focus which would mean it INCREASES the DOF. But the Effect on Lens Aperture paragraph states otherwise. Please correct me if I am misunderstanding DOF or am missing something here with aperture sizes.

Also, I was told my edits were unsourced. I did not state any specific facts or impart any new information to the article, I was merely making it consistent. How would one suggest I have sourced that better? Hergio (talk) 18:19, 28 December 2010 (UTC)[reply]

I don't see any such article edit under your contribs. Can you link the diff and quote what you changed it to. The text you bolded above seems to be correct and consistent with what you're saying when you state "reducing the aperture diameter, it brings more things into focus which would mean it INCREASES the DOF", so it's not clear what you would have changed. Dicklyon (talk) 19:28, 28 December 2010 (UTC)[reply]
Edits were apparently made without logging in (using IP address 69.205.119.161). JeffConrad (talk) 02:23, 29 December 2010 (UTC)[reply]
There is no contradiction between the two paragraphs quoted above. The first mentions situations for which large or small DoF may be appropriate but does not discuss how to control DoF; the second describes the effect of aperture on DoF without discussing whether large or small DoF is appropriate. The edits contradicted the well-established inverse relationship between aperture diameter and DoF, were challenged as incorrect, and thus required support from a reliable source. But the main objection was that the change was incorrect, which is why I (and I presume Kleopatra as well) reverted immediately, and asked for an explanation before making the edits again. JeffConrad (talk) 02:23, 29 December 2010 (UTC)[reply]

Hyperfocal distance depth of field using EasyTimeline

[edit]

The chart is an interesting illustration, but it has several issues:

  • There seems to be a confusion of aperture and numerical aperture; as Numerical aperture clearly indicates, at infinity,
and even then, the small-angle approximation breaks down for short-focus lenses.
  • The symbol N is used to indicate aperture (f/N), whereas in this article and similar treatments, it’s used to indicate the denominator of that fraction (e.g., N1: f/8” should be N1: 8”).
  • I doubt that it’s obvious to most readers that the near limit of DoF for infinity focus is Hf (I had to think for a minute), though perhaps this simply suggests that we should mention it in this article.
  • The WLs to Exposure (photography) and Distance seem a bit off the wall for this context.
  • The chart seems to imply the definition of hyperfocal distance used in this article, e.g.,
yet illustrates Hyperfocal distance#Mathematical phenomenon using the same symbol H. The phenomenon holds only for the approximate hyperfocal distance
  • Because a larger version of the chart is not available, the text is quite small and hard to read, especially with the bunching up of the text of the headers in the stub columns (e.g., the first would be much easier to read if it were f1: 50 mm, N1: f/8” rather than f1:50mm N1:f/8”). Slightly larger text would also help, but I recognize that this could make the chart unacceptably wide.
  • A minor issue: quantity symbols H, f, and N should be in italics, as I’ve shown them.

I also wonder if the chart is as appropriate here as it is in Hyperfocal distance. The bulk of the illustration is the effect of successive values of H/x; though it’s interesting, it usually isn’t useful in practical photography.

Some of the comments are largely personal preference, but at the very least, the outright errors should be corrected if this chart is to remain here or in Hyperfocal distance. JeffConrad (talk) 00:34, 2 June 2011 (UTC)[reply]

I’ve addressed most of the easy-to-fix issues. The WLs to Exposure (photography) and Distance remain. I think the first should be changed to Aperture or perhaps removed. JeffConrad (talk) 07:44, 25 June 2011 (UTC)[reply]

Number, type, and quality of images

[edit]

We’ve occasionally discussed a possible excess of images, especially ones illustrating shallow DOF. We’ve also done some housecleaning from time to time, though it’s been a pretty haphazard process. This article is now protected as a result of a silly edit war (or at least the beginning of one) over replacing one image of a cat with another. I don’t think we really need two images of a cat, and it’s not obvious that the proposed replacement is superior to the existing image.

I hope we never imply that a new image cannot be added, or that an existing image cannot be replaced with one that better illustrates a point. But we need a process for managing the image collection–a process that has at least some reasonable basis. Though it might be reasonable for an editor to propose an addition, and especially a replacement, before making the change, this approach will probably be very much the exception rather than the rule, and we have no basis for demanding it. But edit warring isn’t the solution, either, and I would think after a couple of reverts to a change or addition, it would be reasonable for the editor to discuss the issue here.

For any discussion to be productive, we need to have some criteria for what we want in images. Though I feel as if I’m stating the obvious, it seems to me that

  • Any image should serve to illustrate a point in this article, and we should probably have no more images than necessary to illustrate any given point—Wikipedia is not an image repository. In some cases, we may have more images of one type than are needed, but we may also be missing images that would improve the article by illustrating a concept that may not be completely clear from the text alone.
  • The better an image illustrates a point, the more it should be a candidate to replace an existing image or supplement what we have.
  • All else being equal, of course, a good image should be preferable to an inferior one. But I would think there should normally be a certain threshold of improvement to justify replacement of an existing image. We have several very high quality images, including that of the Nikkor lens and the Tachinid fly; I think we also have a few that are less good.

We should also have a process for objecting to an addition or replacement with which we disagree. One approach is to mention the objective here before doing anything else, but that seems overly restrictive. I would think that, at the very least, a revert would offer a brief reason why the change or replacement was not thought helpful, with a request to discuss the proposed change here. In some cases, making comment here as well as in the edit summary may be a good idea. But as obvious as this may seem, the comments in the recent mini-war weren’t very descriptive.

Perhaps it would help if we created a simple road map of images that we think we should have, and compare that with what we do have, to arrive at additional images we want, current images for which we might want a better replacement, and perhaps some current images that really don’t further this article’s purpose. JeffConrad (talk) 08:38, 25 June 2011 (UTC)[reply]


I am not sure how clued up you are with photography, but the one of the boy is an awful photo, as stated previously (I will do the same here) it is poorly lit, of poor quality, and has no points of reference, something which the cat picture does, it is also the only picture to do so. Personally, I find the other picture hurts my eyes, it is far too harsh. I am also cheesed off I was followed by the Gang of 4, who undid a load of my edits. Jamesington (talk) 10:33, 25 June 2011 (UTC)jamesington[reply]

If you look at my previous edits here, I think you’ll see that I understand photography, and DOF in particular, quite well. I agree that the photo of the boy is unlikely to win Photo of the Year. But it does illustrate one point quite well—that out-of-focus highlights have the shape of the lens aperture stop, and that alone may justify its inclusion until we find a better image to illustrate this point. The caption should probably be updated to reflect what I just said.
I assume that by “cat picture”, you mean the one you added; it’s less clear what you mean by “other picture” (the current cat picture?). In any event, I’m not sure that a “point of reference” is essential, and that a picture “hurts your eyes” is hardly justification for replacing it.
A confrontational approach (e.g., references to the “Gang of 4”) usually isn’t the most effective one. I agree with Ron Ritzman’s comment at AN/I: after a couple of reverts, you should have gone to the Talk pages, and those who reverted you should have provided better and less-confrontational edit summaries. I can’t exactly cast the first stone, because I’ve occasionally been a bit short with people in edit summaries, and I restored one of your edits that I would have supported reverting had I paid closer attention to the history. So to some extent, we’re all guilty as charged.
Getting back to the task at hand: as I’ve indicated, I prefer the existing cat image to yours (and would keep the image of the boy for other reasons, at least for now). But if replacement finds consensus, that is how it should be.
I suggested developing a list of desired image types so that deciding which images to add, keep, or replace might seem a bit less capricious. It might also eventually lead to making this article better.
Incidentally, there’s no need to add your name in signing a post—the four tildes include it automatically. JeffConrad (talk) 23:47, 25 June 2011 (UTC)[reply]

No one has made any suggestions, so I offer a few.

Current images

  • Move the image of the camera and butterfly back to the lead where it started out, because it’s most effective as an simple, informal illustration of the concept.
  • Move the image of the boy to the section Acceptable sharpness, and change the caption on to mention the out-of-focus highlights in the shape of the aperture. If a better image showing this can be found, replace this image.
  • Consider removing the two digitally processed images from the lead. It’s not clear how they relate to this article; consequently, they seem gratuitous, especially the second one.
  • Consider replacing the four images in the section Effect of lens aperture with ones in which the effect is more obvious at the size displayed. Two images might suffice, and would better match the figure at the right; the images of the jonquil flower would fill the bill quite well, they also well illustrate isolation of the background, and I'm not sure we want to include them twice. Going to two images rather than four might oversimplify, but the effects of aperture would be more readily apparent.
  • Revise the caption for the image of the circuit board to note the softening in the plane of focus at f/32.

New image

  • Four images of a scene from the same camera position but with different focus settings:

    – One at the nearest object to be sharp, with a large aperture

    – One at the farthest object to be sharp, with the same aperture

    – One at the harmonic mean, with the same aperture

    – One at the harmonic mean, with an aperture that has the DOF extend from the near object to the far object

    I’m not sure where this would go (perhaps the lead), and some additional text might be needed to explain it. The purpose would be to illustrate what’s going on for a reader unfamiliar with photography. We’re currently loaded for bear on some of the more complex aspects, but may fall a bit short on the simpler ones (as do many other treatments). The current macro image could serve as a contrast showing a small DOF.

  • Images to possibly replace the four in the section Effect of lens aperture, discussed above.
  • Better quality image to replace the one of the boy but still show the aperture shape in the OOF highlights.

JeffConrad (talk) 02:42, 28 June 2011 (UTC)[reply]

And here's a better cat image.
At f/2.8, the cat is isolated from the background.
Jeff, I think those are all great suggestions, but I'd like to keep one CG image, like Glass ochem dof2.png, to show that the effect of DOF can be simulated by ray tracing. And I found a better cat image from the Cat article. It works both at thumbnail size and zoomed in. Dicklyon (talk) 14:37, 28 June 2011 (UTC)[reply]
All but the new images (arguably the most important, unfortunately) are easily done, subject to having things lay out reasonably. Unless someone objects, I’ll give it a shot.
I see the two cat images as about equal, though I don’t have strong feelings about them. I’ve put the other one here for easy comparison. Given the recent fracas, I think we should give others a chance to weigh in. JeffConrad (talk) 19:19, 28 June 2011 (UTC)[reply]
Easy part done—see how it looks. I’ve left the cat image alone for now. JeffConrad (talk) 22:12, 28 June 2011 (UTC)[reply]
I think File:Kittyplya03042006.JPG is the better cat picture because it has an out-of-focus foreground (the paw) illustrating DOF better. cmɢʟeeτaʟκ 17:27, 29 June 2011 (UTC)[reply]
I had the same reaction, though as I’ve said, I don’t have strong feelings about the choice. If we decide to keep this image, perhaps we should revise the caption accordingly (we already describe isolation from the background in the images of the jonquil flower). JeffConrad (talk) 00:53, 30 June 2011 (UTC)[reply]
Yes, that one's OK, too. We could mention that the paw is closer than the near edge of the DOF and the body is farther than the far edge. And that the DOF depends on how big you make the picture. Dicklyon (talk) 03:59, 30 June 2011 (UTC)[reply]

cat

[edit]
Jamie's shoe cat
Boy with hedge

If you have seen my name before you know why I am here... The picture of the boy I have replaced with a cat, as it has been edited in photoshop. as it is chopped I dont think it should be highlighting depth of field when it could be an entirley fake image and some guy is just puhing his kid, just like people claim I am doing with the cat. Wikipedia is something anyone can edit right? So then instead of reverting my edits and suggesting similar pics of what I have already added why not just let me edit for gods sake. — Preceding unsigned comment added by Jamesington (talkcontribs) 23:14, 28 July 2011 (UTC)[reply]

Could be, but so what? Is there any reason to dislike or suspect the image of the boy? I think it's good; and I don't see anyone pushing it who is relating to it, do you? Chokity, who took it, placed it about a year ago, and nobody objected to it, so no pushing was needed to get it to stay. What image hasn't been adjusted in Photoshop or some such? A reason to not just let you edit is that you seem to have a WP:COI in really wanting your own cat's photo to appear in various articles, when other editors think it's not the best choice. Why not just contribute where no such appearance of COI is possible? Dicklyon (talk) 01:25, 29 July 2011 (UTC)[reply]
Be assured that Dick and I have nothing against your cat (or perhaps just your cat picture)—as we seem to have agreed above, the primary objective of images here is to illustrate what we discuss in the article—it’s not an image gallery. The photo of the boy well illustrates a concept in the article—that blur spots have roughly the shape of the aperture stop. This information has now been lost for the moment as a result of the games here. JeffConrad (talk) 08:52, 29 July 2011 (UTC)[reply]
My point is it is a horrible image. The background is harshly blotted. And the face is in a harsh light. It is not a nice picture. also shows some ambiguous hedges, as opposed to the cats obvious shoe. — Preceding unsigned comment added by 90.206.178.99 (talk) 22:51, 29 July 2011 (UTC)[reply]
Your point is taken. But your judgement is biased. If someone is to decide that your photo is better, that someone shouldn't be you. I think the image of the boy is good, and effective. Dicklyon (talk) 23:01, 29 July 2011 (UTC)[reply]
I can't believe we are still debating which picture(s) to include. Trying to be as neutral as possible, the boy image illustrates bokeh better while the cat image has better light. But why a shoe?
Why don't we instead list the phenomena we want to illustrate (e.g. bokeh and out-of-focus foreground *and* background), then someone go out and take a photo fitting the criteria, instead of picking some random photo satisfying just some of them. (Also preferably use a neutral subject to avoid allegations of pushing one's cat or kid!) cmɢʟeeτaʟκ 01:12, 30 July 2011 (UTC)[reply]
The boy picture is used for one thing; its caption reads "Out-of-focus highlights have the shape of the lens aperture." It's possible someone could find or make a better image for this concept, but this one is OK. Dicklyon (talk) 01:20, 30 July 2011 (UTC)[reply]
I actually have listed, in great detail, what might be good to see in images, and this seemed to find general agreement. If there are other ideas, by all means mention them. I don’t think the image of the boy is great, and honestly questioned its addition. But it well illustrates one of the points on which we seemed to agree, and I finally changed the caption to reflect the value that the image provides here. I don’t think Jamesington’s image is all that good, either—the cat’s face seems pretty dark to me (and the white patch on the cat’s throat distracts. The composition isn’t all that good, either—with the shoe at the left and the cat almost centered, the balance is off. So I think the current cat picture is better; if it were a tie, it would prevail anyway in the interest of article stability. And the picture of the boy is preferable to Jamesington’s cat picture for reasons already stated.
I suppose a “neutral” image couldn’t hurt, but if we were to insist on it, if might be tough to find many candidates. Because the problem here seems to be with one editor, I’m not sure we need to be too concerned. All other things being equal avoiding pictures of family members or pets is probably preferable—but I don’t think it should dictate.
Finally, Jamesington—would you please sign your posts with four tildes? JeffConrad (talk) 06:13, 30 July 2011 (UTC)[reply]

Near:far -- telecentric case?

[edit]

At the risk of adding complexity to a long and nuanced article, this line caught my attention: "The DOF beyond the subject is always greater than the DOF in front of the subject." While that is certainly usually true, isn't it not the case for an object-telecentric system? My sense is that the typical near:far asymmetry is due to the combination of both NA and magnification decreasing with distance. For a telecentric system, I'm thinking: (1) the entrance pupil is at infinity so the angle subtended by the EP doesn't depend on distance, and (2) the magnification is uniform with distance.

On graph paper, this looks right. I have a nominal point at x=−2, a paraxial converging lens at x=0, a 2-unit-diameter stop at x=2, and an image plane at x=4.

  • If I put another point at x=−1, its marginal rays are parallel to those of the first point (by the fact that the EP is at infinity), so the marginal rays go up to y=1 and then are columnated, making an image spot with diameter 2.
  • If I put a third point at x=−3, its marginal rays are again parallel, hitting the lens at y=3, kissing the stop at (2,1), coming into focus at x=3, and hitting the image plane at y=−1, again making a spot with diameter 2.

If that's correct, then the near:far distinction breaks down for a telecentric lens, No? —Ben FrantzDale (talk) 14:36, 11 August 2011 (UTC)[reply]

Depths of field plot

[edit]

The plot is confusing (well, to me, anyway) and needs at least a clearer explanation, and part of the caption reads "the depth of field is increased by using a shorter focal length or smaller aperture." Shouldn't it be *longer* focal length or smaller aperture? Sergivs-en (talk) 09:02, 14 December 2011 (UTC)[reply]


Same focal length for both formats

[edit]

in this section you can find following statement (take a closer look regarding info about perspective):

... the DOF ratio is the same as that given above, but the images differ because of the different perspectives and angles of view....

a few lines below one can find something different regarding the very same perspective:


... so the DOF ratio is in direct proportion to the format size. The perspective is the same for both formats, but because of the different angles of view, the pictures are not the same.


in my opinion this is of course mistake. the perspective IS THE SAME. angles of view are different because of sensor sizes, but perspective is absolutely the same. we SEE different portion of the picture but from the same perspective.

my explanation: imagine two cameras sharing the same perspective (like in 3D stereoscopic rig) with 50% transparent mirror. ignore inverting the picture, color balance and image flipping which is caused by mirror. both cameras are matched to each other, so both cameras are seeing the same picture. ZERO PARALAX, SAME OPTICAL AXIS. SAME PERSPECTIVE, SAME LENS, SAME FOCAL LENGHT, SAME F STOP, SAME SENSOR SIZE. and obviously - everything is the same. ok.

NOW - we are CHANGING one of the two cameras. this new camera has SMALLER sensor size. everythig else is as before: ZERO PARALAX, SAME OPTICAL AXIS. SAME PERSPECTIVE, SAME LENS, SAME FOCAL LENGHT, SAME F STOP. what is different now? perspective? no. just the portion of the image sensor is seeing! perspective for sure is the same, because we didn't change that.

when trying to explaing this in a different way i am often using a POSTCARD as an explanation. if we see certain landscape with the, say house in the midle of this postcard (this is our camera with bigger sensor size) we cannot change PERSPECTIVE by cutting out outer part of this very postcard, so only house in the middle is now visible (smaller sensor camera). we will see just different INFORMATION on this postcard but from the very same perspective.

that's it.

thank you,

Filip kovcin (talk) 22:19, 29 December 2011 (UTC) filip kovcin, warsaw, poland[reply]

It's contrasting two different ways of comparing same focal length. One case has same subject distance and hence same perspective; the other different distance and perspective, but same field width at the subject distance. Maybe you can fix it to make the cases more clear. The current wording "Another approach..." perhaps doesn't make it clear enough that it's introducing a different situation. Dicklyon (talk) 00:00, 30 December 2011 (UTC)[reply]

File:Glass ochem dof2.png to appear as POTD soon

[edit]

Hello! This is a note to let the editors of this article know that File:Glass ochem dof2.png will be appearing as picture of the day on July 22, 2012. You can view and edit the POTD blurb at Template:POTD/2012-07-22. If this article needs any attention or maintenance, it would be preferable if that could be done before its appearance on the Main Page so Wikipedia doesn't look bad. :) Thanks! howcheng {chat} 23:02, 21 July 2012 (UTC)[reply]

Simulated depth of field
A 3D ray tracing of a glass molecular model, demonstrating the algorithm's ability in simulating depth of field (DOF), which is the distance between the nearest and farthest objects in a scene that appear acceptably sharp in an image. When much of the foreground and background of an image is sharp, its DOF is deep. The opposite is called shallow focus.Image: Purpy Pupple

The formulas for D_F and D_N are wrong.

[edit]

The final step, using the found equations for V_N and V_F in the thin lens formula gives a different result than shown in the wiki page. Solved by hand and using wolfram alpha give this result:

http://www.wolframalpha.com/input/?i=1%2Fx+%3D+%281%2Ff%29+%2B+%281%2F%28%28f*v%29%2F%28f%2BN*c%29%29%29+solve+for+x

idk how wikipedia editing/math works but someone who does should fix this

80.246.200.217 (talk) 14:35, 27 November 2015 (UTC)[reply]

Wolframalpha uses Mathematica. I use equally powerful Maple to solve the six equations as a set, and obtain result easily, see rearranged section 11, and the Maple code in the Maple link. --Gisling (talk) 07:10, 20 February 2016 (UTC).[reply]
And one would not normally express it as IP did, with lens-to-film distance as an independent variable. It's not clear what he thinks is wrong, but his version is certainly odd. Dicklyon (talk) 07:40, 20 February 2016 (UTC)[reply]

Changing Away from 'lens can only focus on one point' concept in Circle of Confusion

[edit]

As mentioned in the change summary, a lens focuses on a plane rather than a point. It's important for an understanding of DOF to recognize that the DOF is with respect to the plane of focus rather than the sensor (film is a sensor) plane; although for the special case of a system with limited adjustments where the lens axis is always perpindicular to the sensor plane, the DOF distances can easily be converted to be with respect to the sensor plane. This conversion has been done for decades by camera and lens makers to achieve convenient and practical use of cameras that don't have lens swing or tilt capability. That this is common practice may lead many of us to stop thinking generally and tend to think in terms of these limited-capability imaging systems that have been in widespread use for over 50 years. Editors please beware of this trap. 4.7.25.146 (talk) 21:00, 4 April 2018 (UTC)[reply]

Removed misleading synonyms for DOF. With my limited resources, I couldn't find a reliable reference that equates "focus range" or "effective focus range" with DOF. It wouldn't be a big surprise if it is contained somewhere in the context of limited adjustment cameras; or in operator's instructions written with marketing or public relations in mind; however, it would add to the misconception that DOF may generally be obtained by simply measuring from the imaging system to objects imaged with acceptable clarity, and subtracting the nearest from the farthest. This will only work when the lens axis is perpendicular to the sensor plane of a flat sensor. 4.7.25.146 (talk) 19:16, 7 April 2018 (UTC)[reply]

Article is massive

[edit]

Right now, the article is extremely long! Over two-thirds of the article is taken up by the sections "DOF Formulae" and "Derivation of the DOF Formulae". These seems unnecessary and possibly be split into their own articles or removed outright. Not even the physics or mathematics articles have such lengthy derivation sections. MarkH21 (talk) 10:41, 8 January 2019 (UTC)[reply]

Removed the formulae sections because Wikipedia is not a textbook (WP:NOTTEXTBOOK). MarkH21 (talk) 06:15, 14 January 2019 (UTC)[reply]
I agree it's massive and bloated, but it would probably be better to cut it by about half than to throw out all that work of many editors over many years. If you think a split to another article on DOF math is a solution, let's consider that, too. But just chucking it all out seems like not the best solution. Dicklyon (talk) 06:23, 14 January 2019 (UTC)[reply]
A massive trimming or split of the formula section are possible solutions that I would probably be fine with. With regards to the derivations, I believe that it should be entirely removed. Its inclusion does not really make sense, Wikipedia is not a textbook nor is it a place to give the mathematical derivation of known scientific & engineering phenomena. Anyone who wants to learn about how to derive the formulae would just refer to the referenced textbooks / sources, as one would do for any other subject with mathematical formulae.
Just to note, whether editors have spent significant time writing material should not be a consideration for whether written material should be kept. That's not to say that the work isn't appreciated or potentially valuable though! MarkH21 (talk) 08:15, 14 January 2019 (UTC)[reply]
Following WP:NOTTEXTBOOK / WP:NOTHOWTO, I have trimmed the "DOF Formulae" section down while retaining most of the formulae, but I'm still not convinced that the section belongs. The derivations section is gone because that really does not belong.
There are still some overly-detailed sections and material more fitting of a textbook/how-to, so I am adding the "how-to" tag. MarkH21 (talk) 13:57, 20 January 2019 (UTC)[reply]
The derivations aside -there's quite a lot of repeated content in the article and it's at quite a high level. I think there is a fair amount of rewriting that could be done and I'm happy to make a start on that if people are happy to look over my shoulder a bit? Joe (talk) 17:26, 24 January 2019 (UTC)[reply]
I took an initial swing. Will come back with more books shortlyJoe (talk) 19:37, 24 January 2019 (UTC)[reply]
So far, the revisions look quite good and are much appreciated! Thanks! MarkH21 (talk) 04:43, 29 January 2019 (UTC)[reply]
Thank you - it makes it much easier to keep going when there's a vote of confidence. :) Joe (talk) 09:54, 29 January 2019 (UTC)[reply]

Maybe we should make an article Depth of field formulae and derivations or Depth of field calculations or something like that to preserve the bulk of the removed material? Dicklyon (talk) 05:18, 29 January 2019 (UTC)[reply]

I think that's a sensible suggestion and I'd be for it. My personal preference is to leave the formulas (and derivations) until the structure of the rest of the article has played itself out (because some formulas might end up inline, and I feel like some aren't relevent to DoF anyway...) but I'm without strong opinion on it. Joe (talk) 09:54, 29 January 2019 (UTC)[reply]
I've now looked over the derivations that were removed properly. I think we should have formula (because it provides the exactness that we can't deliver with a general introduction text), and I think that the derivations help with that; however I don't think these are particularly clear derivations even to someone with a fairly solid maths background - I think that we'd do much better with geometric deviations some text and a bunch of diagrams. I'm still on the fence, but I'm now at "I think we should probably have derivations, but I don't think those derivations are half as good as you'd find on YouTube" Joe (talk) 10:13, 29 January 2019 (UTC)[reply]
Please, let's place the redacted derivations into the suggested Depth of field formulae and derivations or Depth of field calculations article (or the like). There are plenty of links at photography forums such as photo.net, dpreview.com, photography.stackexchange.com, etc., that have linked to this article for years, and now (apparently) suddenly, the information is gone. sbb (talk) 01:30, 24 March 2019 (UTC) 'Edit: While the information might better belong at Wikibooks, the simple fact is that so much of Wikibooks' content is just poor. It has no life, community, or upkeep, from most of my readings there. So what was well-written, and pretty well diagrammed, is now just lost to the ether.[reply]

Todo list

[edit]

Some bits I'd like to do in the future - I thought I'd put them here so a) I don't forgot and b) people have a chance to object.

  • I'd like to improve the explanation of what depth of field and circle of confusion actually *are* so it's obvious to the non-technical reader
  • I'd like to add in some diagrams (hopefully animated) and proper text for how the depth of field changes with focal length, subject distance, and aperture.
  • I'd like there to be a decent section on Bokeh, there should be some nice illustrations of things like fancy heart shaped Bokeh ([1] has the right licence)
  • I'd like to have some notes on the extremes of depth of field (lots of examples in [2]
  • From the above - would be nice to add in some content on Group_f/64 for a bit of colour
  • The 'object field methods' section needs work - Merklinger's work appears to be self-published (interesting to read thought tho - [3] and cited a few times in Google Scholar (but hardly mentioned in Google Books) - and it's been presented as the main opinion with Ansel_Adams being relegated to the 'other people think' section. :s Joe (talk) 09:57, 29 January 2019 (UTC)[reply]
The examples of extreme depth of field weren't actually that good, but group 64 should be excellent. I've also added the heart shaped bokeh. I've done some more rewriting today, but I'm also noticing that a lot of the content isn't about depth of field and really belongs in aperture or one of the other photography/optics articles.Joe (talk) 21:37, 27 January 2019 (UTC)[reply]

I'd keep the Bokeh stuff in its own article. I like Merklinger's work, and it's respectable enough, being cited, though self-published. Dicklyon (talk) 18:13, 29 January 2019 (UTC)[reply]

For sure, I mistakenly wrote bokeh above when I meant 'illustrate that points turn into the shape of the aperture by way of the cool heart shaped ones' - I do think that we should link up a little better with Bokeh as well. Joe (talk) 16:09, 30 January 2019 (UTC)[reply]

Object Field section

[edit]

I'm looking at Depth_of_field#Object_field_methods and having read the main source I've got a few issues....

  1. The key part of the Object field method Merklinger proposed seems to be about making it easy to calculate depth of field - but that's not in the section
  2. The section draws a false contrast. Adam's appears to be saying: "Photos look better when close things are as sharp or sharper than far things" (I don't have that source, but I understand that matches up with his body of work), whereas Merklinger is saying "if you want to read text on a boat that's far away, then you have to make sure it's very sharp - you don't have to work as hard if it's close." - I don't think those two views are particularly in conflict and I feel like they should be presented as such.
  3. It feels like it's in the wrong bit of the article - at this point the reader doens't know what influences DoP, or even that the focus point isn't in the middle - I feel like those bits go first.

I'm unsure how to rewrite (or restructure) to get this right - does anyone have any ideas (or the Adam's source)? Joe (talk) 19:20, 30 January 2019 (UTC)[reply]

I think the object field method is sensible as a scheme in which both Adams's and Merklinger's contradictory opinions can be illustrated and contrasted. It's also used in a neutral way by von Rohr, as described in the Thomas Dallmeyer 1899 book (and see my unpublished survey about this stuff). There's more in Dallmeyer's book than I realized at the time (before I had Google Book Search to help me). You can't count on my unpublished synthesis, but you might find it a useful guide to the history and approaches. Dicklyon (talk) 18:14, 31 January 2019 (UTC)[reply]
The closest thing I see in the cited Adams 1980 book is this bit. Dicklyon (talk) 18:31, 31 January 2019 (UTC)[reply]
No, it's on page 51 more explicitly: [4], as our ref 5 says. Dicklyon (talk) 18:42, 31 January 2019 (UTC)[reply]
Thank you muchly on all counts! The sources and the review a both going to be brilliant - I'm semi-learning this as I edit (I *thought* I knew about it) and it's great to have someone with real expertise looking over my shoulder :) Joe (talk) 09:55, 1 February 2019 (UTC)[reply]
Now I've read your PDF (abeit lightly) I've a much better idea of the structure I think we should have, and in the process I suspect it solves some of our formula problems above. Today I've added a bit of content to introduct object field to the passer by, and I've also moved it - because I think we should have an introduction to the 'standard' approach before presenting an alternative - to make this work I might see how the 'effects' section looks with a similar DoP formula using focal length.Joe (talk) 15:27, 2 February 2019 (UTC)[reply]
Jeff Conrad had a ton of expertise, too, but tended to be more wordy than me over the last 13 years, as you can see by how much text he added. I've mostly just tweaked things, trying to keep the errors out. If you can come up with a cleaner and leaner article, that will be great. Dicklyon (talk) 03:58, 3 February 2019 (UTC)[reply]
On that topic - how would you feel about (once we've done a bit more cleaning up) putting the article forward for GA? I'm mostly thinking of it because it would be nice to have the article looked at by someone who actually isn't familiar with the topic and it's a nice thing to go for :) Joe (talk) 13:58, 3 February 2019 (UTC)[reply]
I've never engaged in such processes, but yes it could be useful. Dicklyon (talk) 17:17, 3 February 2019 (UTC)[reply]
After more cleaning up, I would also support such an endeavor! MarkH21 (talk) 22:31, 4 February 2019 (UTC)[reply]

Recent changes.

[edit]

Did a couple of fairly violent changes just now - thought I'd stop to let other people check them over.

  • I added a basic depth of field formula to the top of the article because I think it helps and because otherwise we kind of ambush people at the end.
  • I folded together the overview and 'factors affecting' sections largely to see how it looked - I think it works quite well and I'm inclined to keep it.
  • I removed a couple of images (we have a lot - and we've lost a lot of wordyiness so there's some traffic) and moved others around.

Comments welcome - going to edit somewhere else for a bit. Joe (talk) 19:54, 8 February 2019 (UTC)[reply]

Nice! It does look better. Quick comment, do you think there's any way to cut the Table of Contents down a bit, i.e. section/subsection/subsubsections to remove? Seems like a very long TOC relative to the article length now... — MarkH21 (talk) 11:31, 10 February 2019 (UTC)[reply]
Thank you! Re:TOC - it's shorter than it was a couple of minutes ago - I did a few relatively small changes recently. I think that Depth_of_field#Other_applications is going to get rewritten into a paragraphy, so that will reduce the TOC - I also think that some of the formula are starting to feel like they want to be in the apropreit sections - so that will reduce down quickly. Joe (talk) 20:07, 10 February 2019 (UTC)[reply]

Contining Violence

[edit]

I've been continuing my near-violent changes. I think we are nearly there thought.

The three things that I think need to be done before submitting for GA review are:

  1. Consistent referencing
  2. Sort out formula section
  3. Thorought proofread.

I got most of the way thought the referencing a while ago - it only needs a bit more work - particular on the difraction section that came back. I did a fair amount of work on the formula section today - getting steadily more unhappy with it as I went. I think we need to get a single consistant mainstream textbook or source and start it from scratch. At the very least we should be using the same symbols thoroughout the section. Will investigate the library and come back in a few days to see were we are.Joe (talk) 17:17, 17 February 2019 (UTC)[reply]

Oh yeah, rewrite the lead - the article has changed a lot so it will probably need an overhaul.Joe (talk) 17:18, 17 February 2019 (UTC)[reply]
It's looking good. But an awful lot was lost; did you collect it some place? Dicklyon (talk) 03:15, 24 March 2019 (UTC)[reply]

Lead, 1st caption, section title and content, POV

[edit]

LEAD
The lead still seems too POVish; but, now the POV is not hidden from the casual reader. Salvaggio et al clearly stated their assumption. The reference and link to "focus stacking" was removed because that article seems to have no reliable basis. Also the 'can't be done by traditional methods' comment was removed as being unsupportable.

1st CAPTION
Reworded the caption

SECTION TITLE AND CONTENT
Changed section title to "Overcomming DOF limitations" to more accurately reflect the intent. Made a few improvements to the wording. Left "focus stacking" in but maybe "image stacking" is a more widely used term (Adobe uses "image stacking") and there doesn't seem to be a reliable reference for "focus stacking".

POV
The article is written from the POV that cameras have a lens axis perpindicular to the sensor plane; that can be very misleading. May want to consider using Conrad, Merklinger, Ray, Wheeler, and others that show what DOF is more generally. The general still applies to the current POV.

208.54.86.201 (talk) 02:00, 8 June 2019 (UTC)[reply]

GA Review

[edit]
This review is transcluded from Talk:Depth of field/GA1. The edit link for this section can be used to add comments to the review.

Reviewer: Kingsif (talk · contribs) 01:56, 17 September 2019 (UTC)[reply]

Hi, I'll be reviewing this with comments posted soon (if just to learn more about f-stops myself) Kingsif (talk) 01:56, 17 September 2019 (UTC)[reply]

Style

[edit]
  • Lead could be a little longer per length of article
  • Please change "techniques/equipment" to 'techniques and/or equipment'
  • Wikilink circle of confusion - the first time it's used. A brief explanation may also help.
  • Some random capitalized words?
  • Is there a need to italicize "circle of confusion" in exactly one instance that it's mentioned?
  • In the Factors section, the first line is just repeating the lead (and copyvio)
  • Should have wikilinks and brief explanation of other technical terms (assume the reader knows nothing of camera tech)
  • Inconsistency with how depth of field is capitalized
  • Factors affecting... does not seem well structured, with some perhaps irrelevant content that would belong in a different section. In fact, the other sections have the same problem.
  • Article doesn't seem to have a good sense of direction.
  • Later sections especially seem to be a quick collection of information from personal knowledge tacked on the end without consideration
  • Fail - re-write may be necessary, with a good structure

Coverage

[edit]
  • Lead does not cover main points, only outlines what dof is. (Also, it's copyvio.)
  • Top section of Factors affecting depth of field seems to cover more what dof is and how to calculate it and functions of it more than factors affecting it.
  • Scheimpflug principle illustrated but not otherwise mentioned
    • There is no need to mention the "Scheimpflug principle" in every article that discusses swing or tilt. It seems to be promoting the annoying misunderstanding of the principle that, swivel of lens or sensor causes swivel of POF. Although it is true that one causes the other, that is not the principle. And, one doesn't need to know the geometric relationship of POF, lens plane, and film plane to understand DOF (or even how to use swing or tilt). - NewageEd (talk) 09:47, 23 January 2020 (UTC)[reply]
  • Fail areas covered in appropriate sections, clear examples of info missing

Illustration

[edit]
  • Good choice of top image
  • The scientific diagram at factors is good and clear, but relies on the description. Not sure there would other wise be a way to show this, though.
  • Nice illustrative use of resultant image with building blocks
    • The building blocks illustration could be improved if the plane of focus was on the middle stack of blocks. Then it would show that depth of field works in both directions; maybe add a couple stacks of blocks. - NewageEd (talk) 09:22, 23 January 2020 (UTC)[reply]
  • Seems like the 4 dial images are all for the DOF scales section, and overflow severely - should put a gallery at the bottom of the section
  • Needs attention

Neutrality

[edit]
  • Yes, good
  • Pass
    • At this stage it is written from the narrow point of view of cameras with no swing or tilt of lens or image plane. Maybe when it is rewritten to include camera movements it will become neutral. I'll start by rewriting the "Camera movements" section. - NewageEd (talk) 09:13, 23 January 2020 (UTC)[reply]

Verifiability

[edit]
  • Has a citation needed tag, that seems to have been added today
  • Good range of RS
  • However, large parts of article unreferenced
  • Likely some OR in 'most cinematographers, though' kind of writing
  • Fail - a cn tag is a cn tag

Stability

[edit]
  • Had some edits today for general improvement. Nothing major, though.
  • Pass
[edit]
  • Nearly the entire lead is copied here
    • Copied to there you mean; that's obviously not a place anyone would copy from, just an assemblage of snips from WP. Dicklyon (talk) 13:03, 21 September 2019 (UTC)[reply]
      • I did check the context; the lead is random snippets of the source text, which is a really long paragraph in answer to a question on a forum about depth of field. It would be unusual to have a lot of someone's own writing with whole sentences of only the WP lead incorporated so naturally at random points... Not saying it has been taken, but it would be odd for the text to have been copied to that source, given its presentation. Kingsif (talk) 14:30, 21 September 2019 (UTC)[reply]
        • This is not a copyright violation or plagiarism. The journal page referenced above, here, are personal notes under the title of "Monster Hunter World white sharpness". The snippet was obviously taken from Wikipedia after 8 June 2019, not the other way around. However, reliable sources are clear in describing depth of field in relation to the POF. The current wording misses this clarity in comparison with the previous wording of 10 Feb 2019; which, appears to accurately represent illustrations and descriptions contained in reliable sources. - NewageEd (talk) 08:52, 23 January 2020 (UTC)[reply]
  • Images seem correctly licensed
  • Fail - please rewrite lead

Overall

[edit]
  • Needs a significant restructure and more citations throughout. Also copyvio and obviously missing info. It looks like the main editor has a lot of knowledge, has worked hard creating it. Main reason for immediate fail is the poor structure. Kingsif (talk) 21:12, 18 September 2019 (UTC)[reply]

Blurry hearts picture

[edit]

The blurry hearts illustration has more to do with the fact that aperture shape affects the shape of blurry spots; but within the DoF, the blurriness is not noticeable. This is looking like a "how-to" photography-art guide. Seems appropriate for an article that discusses aperture shape and artistic blur; which, might not belong in WP (textbook stuff?). - NewageEd (talk) 10:24, 23 January 2020 (UTC)[reply]

Estimate dof in image dimension

[edit]

This is a new logic I have been working on, no publication so far but I will appreciate if somebody is interested in writing an article on this topic. Feel free to contact me, thanks.

he key point is that when we work in the image dimension instead of the scene the equations become very simple. The size of the focus blur varies linearily along a plane in the image

The relation is;

where:

  • a is the numerical aperture
  • h is the "altitude" above the plane. This value can be adjusted if the camera is tilted with the cosine of the angle with the plane.
  • d is the distance of the point from an origin in the focus plane (where s=0)

The demonstration is here

A first conclusion is that the focus is at exactly at the middle distance between the near and far dof limit. Again this is true in the image and along a plane.

Then we have a relation to set the f# for having everything in focus between the focus plane and the near/far limit

  • f is the focal length, h the altitude.
  • k is a constant representing the minimum resolution, k =800 is consistent with the results given with common circle of confusions with classic dof calculators
  • r is the ratio of the image height between the focus and the near or far limit (for instance 1/3 of the height).

There is also a relation taking into account diffraction:

The links are here:

Do not hesitate to give me your feedback, I want to do it the clean way. The equations are disclosed, it is quite simple in fact,it simply uses a different logic Everything is explained in the links. Feel free to contact me !!


Additional property for focus positon vs focus ring: The focus point in the image varies linearly with the focus ring position along a line/plane. When you turn the focus ring, you can see the the position of the focus point in the image moves regularly along the line (but seen in the image).

This is demonstrated here: focus moving with focus ring

Linearity of focus blur in dof scale

We can try to write the equation of the blur s when the position in the dof scale varies (<=> focal plane position).

if we focus at point s0, we can calculate the focus blur s at position s1:

The focus blur simply varies linearily with focal plane position with a gradent = 1/f#

This is explained here — Preceding unsigned comment added by Chrisinparis (talkcontribs) 20:38, 18 September 2024 (UTC)[reply]

Chrisinparis (talk) 18:58, 30 January 2024 (UTC)[reply]

Alternative estimate method

In the previous method we had to estimate the height of the camera abocve the plane. In some cases, it might be difficult to estimate it.

In fact, we can estimate instead any object size located in the reference plane.

There are interesting properties to note

  • The dof length (in the image) on the reference plane remains constant if we move the focus point in the reference plane
  • We can reduce/extend the dof linearly (in the image) by changing the f#

This means that if we can estimate the dof, not necessarily directly the desired dof, we can then adapt it to position and match the dof we want.

Here we will use first a method called the hyperfocal diameter thaht has been described in a dpreview forum thread Here

Then as explained we have the flexibility to focus anywhere and calculate a f# to match the desired dof.

This alternative method is:

Chrisinparis (talk) 22:40, 9 October 2024 (UTC)[reply]