Commons talk:Featured picture candidates

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

This is the talk page for discussing improvements to Commons:Featured picture candidates.
Archives: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26
candidate list

Hi everyone!

Yesterday I nominated an image for featured pictures for the first time ever. I understood from @W.carter: that it is not allowed to ask people to vote. This is called canvassing, as I understood. I was wondering how often canvassing actually occurs here. Since it is very possible that regular users vote for other regular users' photos, I was wondering to what extent this could be considered canvassing. I have a feeling that as a new user, you are less likely to have your photo declared as featured. But I could be wrong!

So far no regular user has voted for or against my photo. However, this while other photos nominated on the same day, and even photos nominated a day after, have been voted on by users who have also voted on other photos. Is this because I am new and people don't know me yet? Or do people not dare to vote for or against my photo, and wait for someone else to vote for the first time? I was just curious! It's nothing personal!

I would love to hear from someone!

Kind regards,

S. Perquin (talk) 17:25, 5 May 2024 (UTC)[reply]

Hi there S. Perquin! Since I'm in the business of informing newbies, I'll answer you. :-)
First, canvassing is rather rare at FPC, but the mere hint of it can be dealt with quite swiftly.
You being new has nothing to do with people not voting. If you nominated a bad photo, you would get a number opposes very quickly, I can guarantee that. ;-) This is actually quite good for a phone photo. Phone shots are usually not of a quality that is good enough for FPC. So firstly, you should be proud that it hasn't been "trashed" right away, no swift 'opposes' is a good sign. On the other hand, the photo just isn't spectacular enough for immediate "wow!"-supports. To give you the very hard honesty, it's what we usually refer to as a "Meh..." photo. It is good and nice and all, but it doesn't have that extra punch that is expected from an FP. The odd angle doesn't help either. Many regular voters are a bit skeptical to artistic photos and they usually don't get many support votes unless they are especially spectacular.
Like many newbies at FPC, you are a bit nervous and chatty. Try to be more patient and see what happens. ;-) If a photo is just on the brink of being good enough for FP, people like to take time to think it over. The worst thing you can do is being impatient, or withdraw your nom before the voting period is over. If you want to ease into how it is to get your photos reviewed, you really should start with nominating them at COM:QIC first, that will get you more acquainted with how voters think and how the system works. It is strongly recommended that you start there before moving on to FPC.
Btw, you don't need to tell people what camera you used and settings and such, we can read all that from the Metadata section on the file page and on sites that decipher the photo's EXIF data for us.
I can also add that you seem to have bunch of loyal friends who are unfortunately making things difficult for you by not reading the rules before showing up and vote. They are not doing you any favors, I'm sorry to say. If they continue with odd comments in annotations and voting despite not having enough days and edits on Commons, some voters can take this the wrong way and stay away from the photo of someone that can be seen as bringing a mess to FPC. I'm sure they mean well, but take this as a friendly warning. --Cart (talk) 18:36, 5 May 2024 (UTC)[reply]
Dear Cart, thanks for your comprehensive answer! I will be patient and I will also nominate my photo on COM:QIC! I didn't know this existed as well. Thanks again for your reply! I wish you a nice evening! Kind regards, S. Perquin (talk) 18:54, 5 May 2024 (UTC)[reply]

Digitally altered photos

[edit]

For the first part of this discussion see Commons:Featured picture candidates/File:Purple martin in flight (30977).jpg.

Digital alterations of photos have been going on for years now and some have been promoted to FPs, 1 & 2 & 3, but like Frank Schulenburg said they are becoming more and more frequent. This is of course because editing programs are getting smarter, more available, and because these techniques are getting normalized and part of our daily life outside Commons now. We will not be able to, or should, try to stop such photos from being uploaded on Commons and even nominated at QIC and FPC. There is no way of knowing where to draw lines, and putting a ban on such will only result in undeclared major edits.

But I think it is about time to make it clear that significant alterations should be declared on the file pages. That way, voters on QIC or FPC can make case by case decisions about how they feel about the alterations. I propose that we add a new rule to the FPC rules (QIC can open a similar discussion if they like) about disclosing significant alterations. I don't have a good wording for such a rule yet, but it could be something like: Significant digital alterations, like replacing the sky, must be declared on the file page. The original version should also preferably be visible in the image's upload history.

We should also keep in mind that we have no way of knowing how altered or manipulated photos imported from outside Commons are. This is very unfair to Commons photographers who try to do the right thing. Voters often go into Oh-Ahh!-Rapture! over a photo from Flickr or Unsplash that any experienced photo editor can see is heavily manipulated, while they will turn a cold side to a Commons photographer who is honest and declares a slight alteration. I think this too should be taken into consideration when voting.

Let's discuss this. Best, --Cart (talk) 12:06, 21 May 2024 (UTC)[reply]

At a time where AI-created and digitally altered images flood the web, it's important for Wikipedia to ensure the integrity of the visual information presented. Wikipedia is a widely trusted source of knowledge, and heavily edited images can mislead viewers, distorting reality and misrepresenting the subjects depicted. Authentic photos help maintain the credibility of the encyclopedia by providing an honest and true-to-life visual representation, which is essential for educational and informational purposes. – I agree that we won't be able to prevent people from uploading images where the sky has been replaced or where reflections have been photoshopped in. However, I'm eager to find a way to prevent those photos from being used in Wikipedia articles. Maybe we need a red, blinking warning sign that says "Please do not ever use this photograph in an encyclopedic context, because it doesn't reflect the reality and will undermine readers' trust in Wikipedia". Other than that, I like Cart's suggestion above (both the wording as well as the requirement to also upload the original version). In addition, I suggest that we add a requirement that “photo art” is clearly marked in the image description and file name. Best, --Frank Schulenburg (talk) 16:27, 21 May 2024 (UTC)[reply]
There are not specific techniques that should be prohibited. It should be about the extent to which it affects the depiction of the subject. Adding, removing, combining, etc. should all require disclosure, but there's a world of difference between swapping in another shot of the same sky from the same place/time and, say, adding a pretty coastal sunset to a mountainous afternoon shot. One is, by definition, realistic while the other is not. It isn't the technique itself that does this but how it's used. Using content-aware fill or the new Photoshop AI tools to expand the sky for compositional reasons seems to be something folks are perfectly ok with, even though that is creating something completely unreal. Creating additional buildings in a skyline using the same technique, however, would probably cross all of our bright lines because it would significantly change the reality of the subject. Removing a little twig poking in from the side is generally viewed as acceptable but removing part of a map would not be. Sharpening is ok, but using an AI program to shift the focus or to fabricate bokeh where it would be impossible might raise some eyebrows. Stitching a panorama to capture the experience of looking at a space vs. stitching it to make a curved vista look horizontal again could be problematic because of the way it affects the subject. In short, what's important is that the subject is depicted in a realistic way. There are a variety of modifications that we do/can tolerate when they improve the picture without sacrificing any realism, and those modifications should be documented but not forbidden. The idea that swapping the sky for one in another photo from the same place/time would be placed in the same "unrealistic" paradigm as creating 100% fake reflections is, to me, wild. — Rhododendrites talk17:08, 21 May 2024 (UTC)[reply]
Of course there are different degrees of alterations, but if we are to implement some sort of rule about disclosure, we need to keep things simple. It's impossible to make subdivisions based on how much a photo has been altered, I mean who will decide where those lines go. We can only inform that changes have been made and let the those who use the photo (or vote for it) make their own decision about whether it's acceptable or not. We can't exactly set up some sort of censor guild to review what alterations are ok and what are not, and grade the images accordingly. In most cases, common sense will prevail. We only need to examine an image's file page better before slapping it onto an article. Heck, at en-wiki texts are routinely checked so that they are not just copied from sources, there are tools for it. It shouldn't be harder to simply check an image's file page. It wouldn't be such a bad idea to simply make WikiProjects more aware about source criticism in these AI times. --Cart (talk) 17:44, 21 May 2024 (UTC)[reply]
I like the guidelines of the National Press Photographers Association (NPPA) for what is not ok in post-processing with regard to protecting visual integrity[1] (I bet there's more available on the web when it comes to sharing photos that are being used in a journalistic and/or educational setting):
  • "Dramatic changes in color that alter the original color of the scene. For example, changing a gray sky to blue. Color correcting sensor/white balance issues from incorrect camera settings is allowed.
  • Changes made by dodging or burning, adjustments to brightness, contrast, color, saturation, sharpening or clarity that significantly alter content by obscuring, enhancing or diminishing elements in the photograph.
  • Just like during the making/capturing of an image you may not add, move, remove any objects or persons. You may only use the cloning tool – or any other tool – to remove dust spots on the image created by the lens, the camera sensor or dust from scanning physical negatives. You may NOT use the cloning tool – or any other tool – to extend the photograph or expand the canvas of the photograph."
When it comes to ethical guidelines, we might also want to talk about whether it's ok to bait animals in wildlife photography, btw. But that's a whole different conversation ;-) Best, --Frank Schulenburg (talk) 17:53, 21 May 2024 (UTC)[reply]
 Support I agree with adding a disclose rule, specific wording to be discussed. The idea of adding the non-manipulated image to the file's history is interesting, is anyone aware of an upload tool that could automate/aid such a workflow? --Julesvernex2 (talk) 09:42, 22 May 2024 (UTC)[reply]

References

  1. Visual integrity, guidelines of the National Press Photographers Association for the 2023 Best of Photojournalism Competition, last accessed on May 21, 2024
We actually have a template here to put on photos that adhere to the NPPA guidelines. It's called Template:RAW. ;-) I just wonder how many Commons photographers would be willing to put that on a file page. --Cart (talk) 18:15, 21 May 2024 (UTC)[reply]
Hm ... what if I don't have a RAW file, but still stick to the guidelines? Kritzolina (talk) 12:10, 24 May 2024 (UTC)[reply]
There are a number of templates dealing with original images, but since most of them are not quite right for this purpose, I'm sure we can fix a corresponding 'Template:Straight Out Of Camera' for jpegs based on the RAW template. --Cart (talk) 13:14, 24 May 2024 (UTC)[reply]

There are multiple discussions here. One is what the expectations should be for labeling modified images in general, which is better suited for COM:VP.
Another is what the expectations at FPC should be for labeling modified images, and it seems safe to say that FPC would prefer nominators to err on the side of disclosure, and that persistently failing to disclose may result in a community ban from participation here (or at least strong skepticism/opposition).
Finally there are the bright lines drawn about That Which We Must Not Promote, which I've thus far found completely unpersuasive. (I'll note before I complain that while my image sparked this discussion (this time), I don't think anyone would accuse me of regularly putting up manipulated images.) Anyone can oppose for any reason, of course. For some reason, inexplicably, taking clouds from the background of one shot and swapping it into the background of another shot from the same place/time is on the wrong side of a bright line when it comes to what's "realistic". But selective background removals are realistic; extending backgrounds to include space that never existed is realistic; AI enhancements are realistic; blasting subjects with artificial light out in nature is realistic; stitching images is realistic; placing plants and fruit in sterile, all-black reflective lightboxes is realistic. If you're fine with fruit that looks like it grows on the Death Star, chameleons that pose for Vogue, "cleaned up" backgrounds, and fake skies but don't like one real sky swapped for another that's fine, but spare me the pretense of one being clearly wrong and another clearly right. In none of these examples is the subject itself made to look unrealistic -- it's all about presentation and background and differing standards for what makes sense. — Rhododendrites talk02:15, 24 May 2024 (UTC)[reply]

I think we need to limit this discussion to what we can expect from an FPC. General Commons photo guidelines is a too big subject and more suited for other forums. Myself, I have no problem with altered images as long as this is declared in a clear way on the file page. I think this debate might have been infected by the fact that a number of undeclared (but easy to spot) altered photos has turned up at FPC lately. Declaring what has been done is the only way to assess what we a comfortable with or not.
I come from a commercial part of photography where every photo is edited/altered. In advertising, the photo is only seen as the raw material on which the image is built with hours of retouching light, cloning dust, adding missing parts, etc. That's part of creating a good image for a product, anything else might be considered sloppy in that business. So I have no problem with staged, cleaned up, fixed images of say an object/fruit/interior. That is expected. Even to the extent that the first shot is called a 'photo' while the finished product has ceased to be a photo and become an 'image'.
However, for journalistic photos, I would not accept more than cropping or what the guidelines mentioned above dictate. Unless there is a conflict with the rules of WikiMedia, like in the Børsen photo where two logos where removed.
Between those two extremes, is a wide range of fixing up images where I would accept more or less editing, depending on how 'arty' or documentary a photo is. This is the area where most of the FPCs are. Most of these can be dealt with by simply using the retouched template, but for significant changes there needs to be more. Like the original in the upload history. We do that when editing old photos and images, and no one finds that strange. It only helps voters make up their mind.
I don't see this discussion as what should be 'allowed' or not, since that needs to be a case by case decision on FPC, but simply a rule that we need to know what's been done to the image. --Cart (talk) 10:43, 24 May 2024 (UTC)[reply]
  •  Comment There is no absolute right or wrong in photo-editing. We will all draw the line in a different place. We have to trust users who upload their own images until evidence points to problems. Non-Wikipedian images are a different matter. Here we should err on the side of scepticism. Those who nominate images where the photographer cannot be contacted should expect to be challenged. The other issue with disclosure of 'heavy-handed manipulation' is that the image can be used elsewhere without the disclosure information - unless the meta data could be edited somehow to say 'I replaced part of the blue sky with a cloud' or 'I added the moon'. Would it be OK to oppose an FPC on the basis of 'We cannot be sure the moon is as shot' or 'we cannot be sure that the peak of Everest was that bright'? Charlesjsharp (talk) 15:25, 28 May 2024 (UTC)[reply]
  • I think it would be fine to oppose a photo if we think that some things have been photoshopped in an imported photo. My reasoning is that all votes as very subjective, and we can oppose on grounds that something in the photo doesn't look right, even if we can't prove it. When we suspect photoshopping, it's usually because something in the photo looks odd, and such feelings can be voiced in a more diplomatic way. Granted, that takes some thinking and objective reasoning, but such reasons are part of the FPC voting process. So instead of saying outright that we oppose on the grounds of photoshopping, a vote can instead be something like: "The brightness of the moon doesn't seem to match the rest of the photo, and that makes for a bad composition for me" or "Based on other photos of Everest, the light looks a bit odd and unnatural to me." --Cart (talk) 15:51, 28 May 2024 (UTC)[reply]
  • I don't want to call out the photographer - they're great photos in the abstract - but the 2012 third place POTY File:Glühlampe_explodiert.jpg and first place for POTY in 2013, File:Glühwendel_brennt_durch.jpg are edited in ways that remove literally all encyclopedic value from them. The lightbulb has power going into it in both cases. I wouldn't have minded, say, wires attached to the lightbulb, or other tricks, but it's just misleading to do it this way. It's a pity, because, visually, they're incredible.
Adam Cuerden (talk) 15:44, 28 May 2024 (UTC)[reply]
In that case, it's good that Commons in not just a repository for Wikipedia images, but we can also host photo creations that are visually incredible. Even more so since those photos are well declared as to how they were made. People can actually learn how to create similar arty images by looking at the file page. And, reading how a "magic trick" was done will make you better at detecting other "tricks", so that in itself is educational. (Btw, if you know a thing or two about electricity and wiring, it's actually rather easy to set up the same scene and photograph it all in one take. You just solder on two wires to the screw of the bulb, on the back so they are hidden from the camera, and pull them through a small hole in the background. But don't try this at home, if you are not an engineer or electrician and know what you are doing!) --Cart (talk) 16:06, 28 May 2024 (UTC)[reply]
Aye. Wouldn't mind it if there was SOME sign of the electricity. But they were being used as if a cracked lightbulb would automatically have the tungsten filament catch fire, etc. Adam Cuerden (talk) 20:48, 28 May 2024 (UTC)[reply]
I want to second Cart’s statement: “I don't see this discussion as what should be 'allowed' or not, since that needs to be a case by case decision on FPC, but simply a rule that we need to know what's been done to the image.” Banning certain kinds or amounts of editing limits creativity. What we need are honesty and transparency, and both can be achieved by documenting the edits, not by banning them.
IMHO it would be very difficult (or even practically impossible) to formulate general rules of what would be allowed and what would be forbidden for a multi-purpose image archive like Wikimedia Commons. For example, the NPPA guidelines cited above are excellent for their original purpose – news photographs –, but IHMO it would be destructive to use them as a general rule for all uploads on Commons. The ban on cloning out any persons would make it difficult or even impossible to publish some photos here. Often I cannot take a photo of some building or natural monument without some easily recognizable bystanders. The GDPR, as it is often understood here in Germany, makes the publication (maybe even the mere taking) of such photos illegal (there is an exception for press photographers, but probably not for anybody else – we are still waiting for decisive judgements on this topic). Now instead of just giving up and not taking any photo, or asking every single person for a written permission to take and publish the photo, including the possibility to revoke that permission at any time (which is virtually impossible: it would urge us to be ready to delete the photo at any time), it’s much easier to take two or three photos with moving persons and to combine them to a single one, showing no (easily recognizable) persons, or to clone out the persons in post-processing. The NPPA rules would forbid this. This is just an example, but IMHO it shows that while these rules are excellent for a certain use and audience, it would be difficult or even harmful to apply them as general rules to a multi-purpose image archive like Wikimedia Commons.
And there are even simpler examples. For example the dynamic range of many natural scenes still extends the dynamic range of camera sensors. This is why people use HDR, manually combine some exposures or at least do some local editing like raising the shadows of the landscape while keeping the sky a bit darker. The purpose of such edits is simple – people want to give an adequate impression of the original scene within the limits of the small dynamic range supported by traditional 8-bit-per-channel RGB images. There is nothing dishonest or deceptive in such edits as long as you do not exaggerate them. But depending on how you read the NPPA guidelines, such edits may be prohibited. This is another point why I think that general rules are dangerous. We must rather judge and discuss concrete images, not prohibit ways of editing in general.
I would also second Cart’s other statements … e.g. on the recent moon example. There is no need to cancel a FP nomination just because the moon etc. “is photoshopped” (especially when this just means a local adjustment of the brightness like here). The question is rather: what was edited? to what extent? was it done well? is the result still realistic and honest? etc. And, of course, especially when the edits are far-reaching: Are they documented well? Best, – Aristeas (talk) 15:41, 29 May 2024 (UTC)[reply]

Open letter about resourcing of Wikimedia Commons

[edit]

Dear FPC regulars, I am posting this here because it could be of special interest to you.

There is some concern that the Wikimedia Foundation could reduce the resourcing for Wikimedia Commons, probably because many people do not understand the importance and possible impact of Commons to its full extent. Therefore some users have written a short, very moderately formulated essay Media knowledge beyond Wikipedia and invite everybody who agrees with it to sign it in order to show support.

Please take a look at that statement and consider to sign it, too. The same essay is also present on meta.wikimedia.org and can be signed there, too. Thank you very much, – Aristeas (talk) 09:44, 4 June 2024 (UTC)[reply]

Proposal to have icons on the top right of files that became POTY finalists and winners This file was awarded first place in Picture of the Year This file was awarded second place in Picture of the Year This file was awarded third place in Picture of the Year This file was a finalist in Picture of the Year

[edit]

Dear FPC users,

As you all know some featured pictures eventually end up being a Picture of the Year finalist or winner. A proposal has been made on this page to add an icon on the top right of each file page that was assessed POTY winner or finalist, like it's already the case for Featured pictures , Valued images , Quality Images , Wiki Loves Earth winners and Wiki Loves Monuments winners .

So feel free to leave a vote (red or green) on this page in order to see if there is a consensus to implement this proposal.

Thank you for your time and I wish you all a beautiful day -- Giles Laurent (talk) 18:42, 4 June 2024 (UTC)[reply]

 Support I like the idea Wilfredor (talk) 20:56, 4 June 2024 (UTC)[reply]
@Wilfredor: Thank you for your vote but I think all votes should be posted at the same place which is : on this page (click here). So I would be grateful if you could also post your vote there! Thank you in advance and have a beautiful day! -- Giles Laurent (talk) 21:28, 4 June 2024 (UTC)[reply]

How many active delists by the same user should be allowed?

[edit]

This question pops up once and again, so it seems the current wording of the rules is not sufficient and should be amended by an explicit sentence stating how many active delists by the same user should be allowed. An amendment of the rules requires a general discussion, hence belongs to this talk page. Because this time the discussion has started apropos of this delist nomination, I copy the comments from that discussions to this place. Please allow me to ping the same users again in order to hin them at this general discussion: A.Savin, Basile Morin, Charlesjsharp, Rhododendrites, W.carter and everybody else, please continue the discussion here. – Aristeas (talk) 08:16, 6 June 2024 (UTC)[reply]

  •  Comment We often say that images are “nominated for delisting” and AFAICR apply the “only two nominations” limit also to delist nominations. However I would be happy to learn that I am wrong.
Asking some users which were especially interested in rule discussions in the past – @A.Savin, Basile Morin, Charlesjsharp, Rhododendrites, W.carter – and everybody else with FPC experience: Would you say that the “only two active nominations by the same user” limit applies also to delist nominations (as I have assumed here) or would you say that one can nominate as many images for delisting as one wants?
I would argue that the limit should also apply to delist nominations.
  • Formal argument: the rules say “The delisting rules are the same as those for FPs, with voting taking place over the same time period”, so if not stated differently the two nominations limit should also apply to delist nominations.
  • Material argument: In past discussions the limit has often been justified by emphasizing (1) that nominations must be discussed carefully (we want quality, not quantity); and (2) that the FP maintenance resources are very limited. IMHO both points apply even a fortiori to delist nominations: (1) they certainly need an especially careful discussion (we all want to avoid arbitrary delisting) and (2) they mean even more work for us maintainers than normal nominations – delisting a FP always requires misc. manual edits.
How do you see this? Thank you very much for your statements! Best, – Aristeas (talk) 05:35, 6 June 2024 (UTC)[reply]
  • As I recall we've had a few discussions about how many delists should be allowed and always come back to the "Delists are the same as normal nominations". Mostly for the reasons that people have such mixed feelings about having delists at all, and that there is so much work in closing them and only a few users know how to do it properly, since it's all done manually.
But since this question pops up from time to time, perhaps we should add a word or two to the rules to clarify this. Such discussions belong on the FPC talk page, not on a nom since it requires input from the community. I suggest you copy these comments to that page and let the discussion continue there. --Cart (talk) 07:53, 6 June 2024 (UTC)[reply]