Talk:Vignetting

From Camera-wiki.org
Jump to: navigation, search
This is the discussion page for Vignetting. Click here to start a new topic.


Discussion pages are for discussing improvements to the article itself, not for discussions about the subject of the article.


These edits are trying to make some distinction I don't understand, and have simply made the article more confusing. One cause of vignetting is that the outermost lens elements may be of insufficient diameter, as a tradeoff to reduce weight or size. Is that "natural" or "artificial"? Light does not get tired and dimmer when it travels further, unless some medium is absorbing it. Unless the claim "Lens designers indentify two forms of vignetting: natural and artificial" can be linked to a source I will assume it is something you pulled out of nowhere, and rewrite it.--Vox 13:44, 17 March 2012 (PDT)

I guess what's meant is that the light travelling to the corners of the frame is spread out more (i.e. the light in a given solid angle is spread over a bigger actual area of film at the corner than near the centre) and this is partly because the light has further to travel (the intensity falls as 1/r-squared, and r is bigger at the corners), and partly because the light falls on the film at an angle, so there's a cos-theta to multiply in, and that's a bigger factor at the corner than near the centre). Don't think I'd like to explain that in an article without a diagram...
Might be worth including a ref to centre-filters, that let people like Schneider sell lenses that really vignette quite badly! :)
If you don't understand a topic then don't criticise. Glass has lower transmission that air (it is not perfectly clear), so light travelling further through glass will be attenuated.
Any vignetting caused by physical obstruction, such as a filter or hood, is artificial.
Read any article on lens design and you will soon find discussion of natural vs artificial vignetting. You can quote Erwin Puts "Leica Lens Compendium" if you like.
Don't attempt to rewrite unless you have better knowledge.--DesmondW 15:07, 17 March 2012 (PDT)
PS I have studied optical design--DesmondW 15:11, 17 March 2012 (PDT)
I'm happy if you have access to your old optics textbooks—that is exactly the kind of source that may be cited to support statements which are surprising or doubtful-sounding. Please read Wikipedia's guidelines on verifiability ("The burden of evidence lies with the editor who adds or restores material") and citing sources to understand why it is important. (We are making an effort to tighten up the older, laxer policy under which some of camera-wiki.org's existing articles emerged.)
It sounds like the distinction being made here is "some vignetting is a consequence of geometry alone and cannot be lessened." Can we say just that more clearly? I am very surprised to hear that light is attenuated to any significant degree passing through a (marginally) greater thickness of optically-clear glass. I would need to see this documented. Tests of lenses quite frequently show a loss of more than one f-stop in the corners when used at the widest aperture. This is more than any plausible absorption could account for.
You may be a veteran in your industry, but photography and its related sciences are too extensive for any one person to be an expert in all its aspects. My questions are reasonable ones; and even when the facts are settled there are still better and worse ways to present a clear explanation. As I've said before, a wiki is not the best environment for anyone who reacts defensively to questioning or criticism. --Vox 16:50, 17 March 2012 (PDT)
Vox you seem to dog every contribution I make with often ill informed remarks. Please read what I have stated:
There are three causes of "natural" vignetting viz:
- Longer light path through glass elements which are not perfectly clear
- Dissipation due to angle of incidence
- Reduction in effective aperture due to angle of intersection
You correctly state that most lenses (especially standard) show a light loss of ~1 stop at the edges, due almost entirely to natural vignetting, so why are you disputing what I have said?
I find your attitude completely contrary to the collaborative nature of any wiki and unsuited to an administrator.--DesmondW 17:09, 17 March 2012 (PDT)
you seem to dog every contribution I make I think you are over-generalizing. When you write about cameras or light meters that you have firsthand knowledge about, I take a quick look and make minor edits where I see some rough edges, but I do not question the facts. This is something I do with any other contributor. But when you propose to explain broader principles, all I ask is that you write understandably, and maintain some minimal standard of accountability. I am not an unusually dense person: If I can't understand something, other readers may also find it confusing as well.
I am sorry you find me unfit for my admin role. As one of the originators of camera-wiki.org's fork from Camerapedia, and as the person who actually keeps the bills paid on time, unfortunately I won't be disappearing for the foreseeable future. You are of course free to petition some of the other active editors to have me deposed if you wish. --Vox 21:29, 17 March 2012 (PDT)
Vox, looking around the site I see many articles with sloppy and often downright poor text. I don't make purely editorial changes as it is not my place to do so, but I do suggest that general editorial review would be useful. --DesmondW 04:45, 18 March 2012 (PDT)
When I became involved, the wiki already had 5,288 article pages. You are right that a number of them are poorly written and unverifiable. Finding them all will be a slow process, especially as other mundane maintenance tasks have taken up much time. This is why I would like to avoid any deficiencies in newly-added articles. If there are specific pages you find objectionable, please start a topic on the community discussions page calling for their improvement.
I have begun looking for linkable sources on the topic of vignetting. I do find that the term "natural" vignetting is used, and so I withdraw my objection to that term. Other sources use the terms "optical" versus "mechanical" vignetting, which seems clearer to me. It may be worth introducing the term Cos4θ vignetting, to help emphasize its geometrical origin.
I have not yet found any source which mentions absorption within glass as contributing to vignetting. Can you provide one? Also: In your Erwin Puts reference, what page numbers address vignetting? --Vox 08:55, 18 March 2012 (PDT)
There are simply too many poor articles to raise discussion for every one.
Natural / artificial (Leica terms) Vs optical / mechanical are certainly interchangeable and if you think that the latter are clearer then let's change the article. I refrained from introducing cos4 as rather esoteric, but it is central and perhaps should be discussed (I thought that describing reduction of the effective aperture, the result, was already quite technical). Erwin Puts discusses vignetting on page 80 under the title "Vignetting and the cos4 effect". Attenuation through glass is probably more theoretical than actual. However attenuation of angled light rays through the emulsion layers of the film is significant, as is dispersion. Puts talks of the "bundles of light energy passing through the lens at an angle with longer paths" (not an exact quotation).--DesmondW 11:24, 18 March 2012 (PDT)
Thanks for your changes to the article. I think our discussion here has helped to clarify it. I can add the page number to the Erwin Puts reference or you can yourself.
I am uneasy with the way the word "dispersion" is used here. As I'm sure you know, this has a specific meaning, for the varying refraction of different wavelengths. Could we say "reduction in illumination" instead?
Incidentally, I agree with Dustin McAmera's proposal to mention that in extreme wide-angle lenses, a graduated center filter may be needed to compensate for the high vignetting. --Vox 18:04, 18 March 2012 (PDT)
All suggestions agreed with pleasure, please do make the changes as I shall be away for a week --DesmondW 10:36, 20 March 2012 (PDT)