Commons talk:Valued image criteria

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search
Notice Do not make any change on Valued image criteria page without a clear consensus of the community and / or of the regulars of the project in this most visited talk page.

small comment[edit]

Small comment on criterion 4:

If it is an organism, it should be identified to lowest possible taxon. There are zillions of specialist contactable by the web to get information. E.g. for beetles or spiders you can almost always get to the generic level (with some external help that is.). Species level might involve preparation of sexual organs, and e.g. that would justify an incomplete species level id. A bit of effort from the uploader's side is IMO not too much asked. Lycaon 07:29, 2 June 2008 (UTC)

I agree with you on this point. So I guess we should add another item in the list saying
  • If it is an organism, it should be identified to the lowest possible taxon, which can be visually deduced from the image.
Concerning external help, resources, we should perhaps compile a list of resources (Users here and on other WMF commons projects, links, etc.). Is that something you would be willing to draft, Hans? It could be called Commons:Identifying organisms or something like that. The same list could be of use at COM:QIC and COM:FPC. I guess some of the users listed at COM:TOL could be potential helpers. -- Slaunger 12:24, 5 June 2008 (UTC)
I'm sure I can contribute towards this. The list may get long though... ;-)). Lycaon 12:28, 5 June 2008 (UTC)
Over time yes, but to get started it does not have to be that exhaustive. Great you are willing to do that ;-)) -- Slaunger 12:50, 5 June 2008 (UTC)
I'll start tonight. Lycaon 12:57, 5 June 2008 (UTC)

Comment on criteria[edit]

Criteria number 1 & 2 are comparative - competitive.

  • 1. Is the most valuable illustration of its kind on Wikimedia Commons
  • 2. Is nominated as being the most valuable within a suitably generic scope

Does not assess the inherent value of an image, instead compare it to similar cases. It is impossible impractical to assess a single image to large number of images on broader topics. As a result nominations will mostly be in narrow scopes.

Yes, that is correct. I call this the scope balance. We have tried to descibe that in Commons:Valued image scope. In the test review phase we found that very broad scopes are also very subjective concerning evaluation of value as the relative value will depend pretty mucj upon which Wikimedia project you are using as reference. For instance there is a current nomination within the scope "Livestock". However, the Wikipedias from "Cowistan" and "Goatland" would probably have different preferences with respect to what they would consider the most valuable image of "livestock". -- Slaunger 10:23, 5 June 2008 (UTC)

Criteria no 3 is most important and assess inherent value of the image. This should be elaborated.

  • 3. Must illustrate its subject well
I have seen your thoughts about this and I think they are well-considered and relevant. However, I would rather not like to elaborate to much on this criterion (especially not the technical details) as we have COM:QIC and COM:FPC, which much focus on these aspects. And we also have crit. 1 stating that the image should be the best at illustrating the subject.

Criteria 4,5 & 6 does not assess the image itself but the auxillary information added by uploader and not necessarily reflect the value of image. Unique design of commons allows these to be added and modified later.

  • 4. Is fully described on the image page.
  • 5. Is geocoded, when relevant.
  • 6. Is well categorized, at an appropriate level.
We already has renomination procedures described around in various templates and and category documentation. I have now tried to document a Commons:Valued image candidates/Nomination procedure#renomination procedure based on what we already have. Here it is stated that declined candidates can be renominated if the issues leading to a decline is fixed. I guess this is basically the same as having your alternative states. We also discussed at some stage to have a "hold on" state, but abandoned the diea agreeing to allow for renomination instead. -- Slaunger 10:30, 5 June 2008 (UTC)

--Nevit Dilmen 23:05, 4 June 2008 (UTC)

Do you have a recommendation? --MichaelMaggs 06:00, 5 June 2008 (UTC)
I tried to figure my thoughts about what makes the value of an image here. 3 Seems to be more important to me. Perhaps inherent values of image should be weighted more, elaborated a bit, and be upper on the list?
For one & two a limited number of categories or gallery pages might be suggested for reviewer to compare. Nominating of the competitors might be done by any reviewer. There is also a chance that a better image be uploaded and nominated later. Being useful is different from being used. An useful but unused image should not be declined.
4,5,6 should not be criteria for permanent rejection, if they can be completed later. Instead of declining, a new status such as deferred, delayed, pending info might be added. Once the required info are completed the image might be renominated.
--Nevit Dilmen 07:10, 5 June 2008 (UTC)
I have a similar concern right now with #3. Right now, the description talks only about technical aspects. However, the content of the photo is equally important. For example, my oppose here might not fit under any of the criteria if #3 solely deals with technical aspects. Another example: let's assume the only free image we have of a John Doe is a photo from his childhood, but he is most recognizable as an adult. Should this be a VI in the scope of "John Doe"? There is no language in the criteria preventing this, unless one extends #3 to apply to content as well as execution. – flamurai 21:11, 9 June 2008 (UTC)
I have some sympathy with this. At the moment all we have is a comment on the scope page which reads "There should be something visible in the nominated image which links it specifically in some way with the chosen scope". I would like to strengthen with something like "The image should be reasonably characteristic of the typical range of subjects falling within the claimed scope. For example, the image Image:Sphinx2 July 2006.jpg would not be appropriate as a VI nomination within the scope cats as it is extremely atypical of the normal range of animals falling within that scope. That will be the case even if that image happens to be technically the best of all cat images on Commons." Actually, voters are already starting to impose requirements along those lines. --MichaelMaggs 22:04, 9 June 2008 (UTC)
I agre with your proposal, Michael. -- Slaunger 00:28, 10 June 2008 (UTC)
OK, I have added to scope page, but it can still be tweaked if need be. --MichaelMaggs 06:27, 10 June 2008 (UTC)

Criterion 5[edit]

It has been brought to my attention (I don't know why I didn't see this before?), that criterion 5 contains following phrase "Exceptions include: studio and other non-place-related shots, unknown locations". Those last two words IMO invalidate the whole concept of geotagging: One can of course always argue that location is not known. It is providing a wide open door for avoiding adherence to this criterion. I would love for this two words to be removed from the criterion 5 text. Lycaon (talk) 07:20, 25 March 2009 (UTC)

I think it should be left in, but perhaps qualified: Not every Wikipedian is still around, nor are all potentially valuable images by Wikipedians. In such cases, the knowledge that might once have allowed geolocation is gone. Likewise, if the location isn't known to reasonable precision, then geolocating only serves to mislead, and more descriptive information, e.g. if they had been driven around by a tourguide in Kalamazoo Wildlife Refuge, in the imaginary country of Imagania, and do not know where the stops were, then "In Kalamazoo Wildlife Refuge, Imagania", may be more appropriate than an arbitrary location within Kalamazoo Wildlife Refuge. Adam Cuerden (talk) 07:33, 25 March 2009 (UTC)
I laughed when I saw that. If isn't geocoded, it's probably because it's location is unknown, and since it's unknown, geocoding isn't required—you're off the hook. I know this isn't the intention, but it needs to be reworded if it's kept. Rocket000 (talk) 07:44, 25 March 2009 (UTC)
I see the problem. I think these are somehow words I once formulated. The intention was that exceptions can be made for cases like
  • Images taken at sea - here you often do not know the exact position, instead you should specify some regional information.
  • Historic photos, where the subjects location is impossible to retrieve within reasonable accuracy from available sources, and where the exact location is not crucially relevant for the scope
  • Photos, where the creator cannot be reached, and the exact location is unknown. We may go pretty far in trying to find the location, like here and here, where quite some research was investing it getting a reliable geolocation.
  • I do not think geolocation of animals and species is overrated. I think it is very important. Not only can they be used as input to distribution maps, they can also be used to illustrate regional variances in the appearance of a species. For me, there should be a very good reason for not having the geocode for a animal/plant if you want to support it as a VIC.
I guess, we should rephrase the criterion such that it cannot be interpreted as a wide open door. --Slaunger (talk) 21:42, 26 March 2009 (UTC)
Please see Commons:Valued_image_candidates/Loxodontacyclotis.jpg for the discussion of Slaunger's second link. --Walter Siegmund (talk) 21:02, 20 March 2012 (UTC)
Somehow I overlooked Adams comments, but I generally agree with Adam in his views on this, but only after making a serious attempt and doing some detective work could the geolocation requirement be mitigated. --Slaunger (talk) 21:47, 26 March 2009 (UTC)
FYI, I didn't nominate that frog because of this loop hole. I didn't know about it then. I do think geolocation of animals is overrated (at least it is here). To the normal person, it doesn't matter. In an article, it doesn't matter. To illustrate the subject, it doesn't matter. There's value there, but a different kind. Case in point, I noticed that image's lack of geocoding after I made the nomination. That's how little I pay attention to geocoding most of the time. I'm not arguing for a change in the guidelines or anything. I know that it's very much a part of this project's definition of value. Rocket000 (talk) 05:03, 27 March 2009 (UTC)
I never had the impression you were nominating it because of a loop hole Face-smile.svg. --Slaunger (talk) 07:02, 27 March 2009 (UTC)
Neither did I of course. Lycaon (talk) 08:34, 27 March 2009 (UTC)

(reset indent) I mostly agree with Slaunger on the exceptions, probably but for the third reason. Maybe we can add another one and that is Privacy. I can imagine that if you take a picture in your own garden, you don't want that to be published to the world. But there are several ways to indicate a geolocation without revealing the exact position. Examples for rare species are here and here. Information like in those two examples is still useful for distribution data on those species, while stating that it occurs in a particular country (even a small one like Belgium) may be too vague to have any use. Lycaon (talk) 08:34, 27 March 2009 (UTC)

What I wrote about was the original intent with the unknown locations part of the existing criterion. Privacy is already included as an exception and I think we should keep that part of the criterion as it is. Quote (I have emphasized two exceptions):
"Exceptions include: studio and other non-place-related shots, unknown locations, illustrations, diagrams, charts and maps, situations where the publishing of a location might be prejudicial (for example, privacy concerns, endangered species)."
Yeah, right of course. Have been mixing up things a bit here ;-). Lycaon (talk) 10:21, 27 March 2009 (UTC)
I agree that the manner in which the location has been specified in the rare species examples you give is excellent. Maybe we should explicitly mention this way of specifying approximate location for the exceptions? --Slaunger (talk) 09:55, 27 March 2009 (UTC)
It is my impression that we basically agree that the unknown locations phrase, which is current in the criterion should not be interpreted as a wide open door. Based on our discussions, I therefore propose to make the following adjustment to the criterion (striked out is a proposal to delete, new text in bold is a proposal for an addition):
Proposal 1. Keep the current form, but adjust slightly the wording
5. Is geocoded, when relevant.
All images are expected to be geocoded unless it would not be appropriate to do so. Exceptions include: studio and other non-place-related shots (e.g., at sea), unknown locations, illustrations, diagrams, charts and maps, situations where the publishing of a location might be prejudicial (for example, privacy concerns, endangered species). Where an exact location needs to be avoided, some coarse location data (e.g., regional) should normally be provided in the description field. See here and here for examples.
Proposal 2. Refactor into an itemized form, and adjust the wording (the adjustments are the same, so I have omitted the formatting for deletins/additions)
5. Is geocoded, when relevant.
All images are expected to be geocoded unless it would not be appropriate to do so. Exceptions include:
  1. Studio and other non-place-related shots including:
    • Illustrations
    • Diagrams
    • Charts
    • Maps
    • At sea
  2. Situations where the publishing of a location might be prejudicial inclusing
    • Privacy concerns
    • Endangered species.
Where an exact location needs to be avoided, some coarse location data (e.g., regional) should be provided in the description field. See here and here for examples.
Some comments
  • I propose to delete the rather inspecific normally in the last sentence, as I cannot find any relevat cases, where we would make exceptions for providing some approximate location. But I am in doubt of this. Could privacy concerns imply that not even a regional location should be specified in some cases? It is not something I feel strongly about though, but it seems kinda fuzzy/open-door-like to have there.
  • In refactoring the criterion I had a thought about Endangered species. Would Endangered species (on the location) be more spot on. I mean, you can have species, which are globally abundant, but endangered in the specific location, where the photo is from. Also, as a non-native speaker I am in doubt of the possible nuances between endangered, threatened, and rare. Is endangered the most adequate word to use?
What do you think? --Slaunger (talk) 09:47, 28 March 2009 (UTC)
    • It does strike me that Geocoding is only relevant when the place itself is relevant. If geocoding provides no encyclopedic value, it should not be required. Otherwise, it's bureaucracy for bureaucracy's sake. Adam Cuerden (talk) 17:15, 12 July 2009 (UTC)
      • It can be quite difficult to say when geocoding provides encyclopaedic value. The notion seems rather subjective and introducing it may trigger debate in many candidates. My point of view is: if it is possible and relevant (as already considered) to provide geographic information, then we should provide it, and it is up to the reusers to decide whether it has an intrinsic value, encyclopaedic or not, that they want to take benefit from. --Eusebius (talk) 17:45, 12 July 2009 (UTC)
        • Perhaps if it's easy to provide, but a lot of nominations seem to run into trouble over this where it's said that exact geocoding can't be found out, and they're basically told "do it anyway". Valued images are supposed to point out the best thing(s) on Commons within a scope; not having geocoding is oftentimes much more minor than all the other things we're quite happy to forgive if no other such images exist. Adam Cuerden (talk) 17:57, 12 July 2009 (UTC)

Criteria V[edit]

All images are expected to be geocoded unless it would not be appropriate to do so. Exceptions include: studio and other non-place-related shots, unknown locations, illustrations, diagrams, charts and maps, situations where the publishing of a location might be prejudicial (for example, privacy concerns, endangered species). I nominated a photo of homemade jam, an extremely clear example of studio and other non-place-related shots where geotagging is not relevant. But a user rejected the photo by reference to criteria 5. There is no guidance in the criteria for why. If the user was right, and there are unwritten rules, it must be made clear in the criteria.--Ankara (talk) 20:07, 1 October 2010 (UTC)

See here for guidance. Lycaon (talk) 20:24, 1 October 2010 (UTC)
Thanks for the link. But the criteria should probably not be on the talk page, but in the guidelines? And to me it seems completely crazy to include that information in the image description, it is only relevant for the evaluation of the VI.--Ankara (talk) 21:47, 1 October 2010 (UTC)

Animals in artificial habitats[edit]

Caméléon Madagascar 02.jpg

We have many photos of animals where the background habitat is not natural. Sometimes this is so obvious that nothing needs to be said, but sometimes it is not clear. This leads me to suggest that candidate photos of animals should be more explicit about the type of place where the photo was taken. I was lead to this line of thought by File:Caméléon Madagascar 02.jpg. Here the chameleon carries brown/orange camouflage against bright green leaves. Considering the artificial-looking log on which the chameleon is standing, I concluded that it was in a zoo or other enclosure. Generally speaking, camouflaged animals in the wild do not stand out so from their surroundings; they inherit behavioural mechanisms which keep them pretty well hidden. Anyway, my suggestion is that all proposals of animal photos should make it clear whether the location was a natural or artificial habitat. Macdonald-ross (talk) 08:00, 14 January 2013 (UTC)

Section 4 - Wikipedia links[edit]

I would like to add the sentence "Links to Wikipedia should be included to assist with verification" to Section 4. I think that the most appropriate point would be after the first sentence of section 4.2. The first paragraph would then become

"There should be a full and informative description of what the image depicts, along with any relevant auxiliary information. Links to Wikipedia should be included to assist with verification. Multilingual descriptions should use ...". (Additional text is in italics).

Any comments? Should we also recommend that reviewers assist with translations into their own mother tongue where appropriate? -- Martinvl (talk) 14:46, 31 January 2015 (UTC)

Coarse Geocodes[edit]

May I suggest that a note be placed in Section 5 to assist editors who wish to use coarse geocodes. My proposed text for the last sentence is (new text in italics):

“Where an exact location needs to be avoided, some coarse location data (e.g., regional) should normally be provided in the description field.[Note 1] Regional data can be provided as an alternative to coarse geocodes.
Note
  1. ↑ Rather than quoting locations to 4 decimal places of as degree (giving an accuracy of about 10 metres), using two decimal places will give an accuracy of about one kilometre while omitting the decimal part altogether will give an accuracy of about 100 km.

”

The note will follow Section 6.

Martinvl (talk) 08:55, 18 September 2015 (UTC)

I have a couple of concerns with this idea. How would the reader/viewer know from the dot on a Google map that the accuracy was within a 1 km or 100 km circle, and not the normal tolerance expected around where the dot was? Can the geocoding renderer distinguish between, say 39.000000, -107.104716 (which might be the precise location) and 39,-107.104716 which is (by your reckoning) +/- 100 km on the first coordinate, or would they both look exactly the same to the reader/viewer? DeFacto (talk) 15:36, 18 September 2015 (UTC)
One could always flag to the reader that "For privacy reasons, this coordinate is only accurate to within 10 km". Readers would then know what accuracy they were working with before they clicked onto Google Maps.Martinvl (talk) 16:27, 18 September 2015 (UTC)

┌─────────────────────────────────┘
 A change to the proposed rewording:

“Where an exact location needs to be avoided, some coarse location data (e.g., regional) should normally be provided in the description field. Location data can be degraded by reducing the number of decimal places[Note 1] or regional data can be provided as an alternative to degraded geocodes.

Note

  1. ↑ Rather than quoting locations to 4 decimal places of as degree (giving an accuracy of about 10 metres), using two decimal places will give an accuracy of about one kilometre while omitting the decimal part altogether will give an accuracy of about 100 km.
”
My concerns would remain. Suppose you took a photo in the inner courtyard of Buckingham Palace at 51.5011,-0.1424, but decided to "down grade" the precision of the coordinates to 51.5,-0.1. That would put the photo in the same place as 51.5000,-0.1000, in the back garden of a property in Southwark Bridge Road. That could embarrass the residents there. I would have to oppose that idea. DeFacto (talk) 19:23, 19 September 2015 (UTC)

┌─────────────────────────────────┘
 I added a second change to the wording before DeFactop's last comment, but that second change appewars to have been lost. I am trying again:

“Where an exact location needs to be avoided, some coarse location data (e.g., regional) should normally be provided in the description field. Location data can be degraded by reducing the number of decimal places[Note 1] and adding {{Location withheld}} - for example see here or regional data can be provided as an alternative to degraded geocodes.

Note

  1. ↑ Rather than quoting locations to 4 decimal places of as degree (giving an accuracy of about 10 metres), using two decimal places will give an accuracy of about one kilometre while omitting the decimal part altogether will give an accuracy of about 100 km.
”


  • Pictogram voting comment.svg Comment In the past few days I have been exploring many aspects of Wikimedia Commons. I now believe that this is an inappropriate place for my suggested additions. I believe that we should do the following:
  • Replace the wikilink to "geocoding" in the sub-heading with
    System-search.svgSee also: Geocoding page.
    under the section title. This will bring the page into line with the English and probably most other language Wikipedias and will also emphasize that the VI criteria are a superset of standard good practice. This is also consistent with advice given on the English Wikipedia of Principle of least astonishment.
  • The changes that I proposed should be moved to the Geocoding page as they apply both to VI images and to non-VI images.
  • The phrase "Where an exact location needs to be avoided, some coarse location data (e.g., regional) should normally be provided in the description field." should be replaced with "The Geocoding page describes how location can be identified when it is inappropriate to give an exact grid location".
Martinvl (talk) 13:55, 30 November 2015 (UTC)
Symbol oppose vote.svg Oppose I find the current rule well suited for geocoding. No need to change it. Anyway, I prefer to see this discussion rather than the arbitrary change that you made. --Medium69 You wanted talk to me? 23:28, 30 November 2015 (UTC)

Review critieria - thumbnails[edit]

There has been some debate about the rules. The guidelines say 'The image must look good on-screen at the review size (e.g. 480x360 pixels for a standard 4:3 landscape image).' They do not say that the thumbnail should be used to judge an image. So I don't think we need to change the rules. If there are two or more images which pass the hurdle of 'look good on-screen at the review size', then we look at the full size image and choose the one that is most valuable to illustrate the scope. @Ikan Kekek, Archaeodontosaurus, Peulle, Martinvl: Charles (talk) 22:29, 3 July 2017 (UTC)

  • As I use the word "thumbnail", I mean the review size. The whole point of using this relatively small size is as I understand it to find images that are useful in web articles. So I tend to check how useful the candidates will be in a Wikipedia article. In any case, it's clear from the wording in these criteria that the quality at full size is not to be considered, I believe because we have QIC for that purpose. If we look at the VI Criteria, the reasoning behind it mentions that mobile phone cameras should be good enough - something they rarely are at full size. I feel that if we start looking at the full size images in VI, mobile phone shots will lose out to photos of higher quality despite looking better at the lower resolutions. The emphasis should be on value before quality.--Peulle (talk) 21:00, 6 August 2017 (UTC)
There has been some strenuous opposition to continuing to use the thumbnail review size as the basis for judging which VIC candidate is most useful. It would be extremely helpful if those who object to this would propose a new rule for us to consider, instead of pretending that those of us upholding the letter of the rule are making up some silly new rule out of thin air. I've been considering COM:VIC#How to review an image as clearly binding. At the beginning of "Review procedure", we have this text:

On the review page the image <!!--or image set--> is presented in the review size. You are welcome to view the image in full resolution by following the image links, but bear in mind that it is the appearance of the image at review size which matters.

Correct me if I'm somehow misunderstanding things, but it's quite clear that since how a picture looks in full resolution doesn't matter for the purposes of VIC, it is not to be considered as a factor in judging which picture is best in scope for this project. I am not a hardline supporter of the existing rule, but I do care about not engaging in jury nullification of rules purely by custom or ad hoc actions by one or a few individuals, especially as, if current review size is no longer going to be used as the basis for judging which picture is best in scope, I feel strongly that everyone who would judge photos on VIC needs some other clear basis for making a decision. -- Ikan Kekek (talk) 21:22, 6 August 2017 (UTC)
This point had never before been a problem. The current wording is perfect and does not need to be changed. Everyone must be free to be able to form an opinion on the value of the image which must remain our only goal. --Archaeodontosaurus (talk) 04:41, 7 August 2017 (UTC)
This is your consistent point: Ignore the rule, and everyone to him-/herself, but when we seek to go by the rule, attack. Anyone want to approach this in a way that helps move things forward? -- Ikan Kekek (talk) 11:01, 7 August 2017 (UTC)
As I said in my introduction to this debate, nowhere does it say that we should only judge on thumbnail size. This would be crazy. Most people looking at a VI of wildlife would wish to enlarge the image (you can do this on a smart phone) to see more detail and in which case good resolution must be important as long as the nominated image shows the important features defined by the scope. Charles (talk) 10:54, 9 August 2017 (UTC)
I don't think I agree with you, Charles. On the VIC criteria page, it says: "The image must look good on-screen at the review size (e.g. 480x360 pixels for a standard 4:3 landscape image). Its usability in printed format is not considered." The last sentence tells me that we should look at the review size and not pay attention to the full size. There is still a quality criterion, but it is to be used on the review size only.--Peulle (talk) 14:28, 9 August 2017 (UTC)
I agree fully with Peulle. In addition may I draw to attention the last sentence in the Valued Image home page which reads "... are less about technical quality and more about your ingenuity in finding good and valuable subjects which matter, and about the usability of the information on the image page", something that I feel is too often neglected. I for one, would like see as least one wikilink to an appropriate Wikipedia in every description, or where this is not possible or not apprpriate, a full description with proper citations. Martinvl (talk) 15:35, 9 August 2017 (UTC)
I'm afraid there is some confusion here. Review size on screen should be measured in pixels. It depends what monitor/laptop/smart phone you have and what magnification you choose. In other words it is a variable. If the rules defined a SIZE, like say 60mm width then that might make sense. Charles (talk) 17:13, 9 August 2017 (UTC)
But the guideline gives a number as an example: 480x360 pixels for a standard 4:3 landscape image. You can't just choose a completely different size, it's not relative. Maybe it's OK to deviate a little bit when viewing, sure, but not a lot.--Peulle (talk) 17:33, 9 August 2017 (UTC)
I agree with you Charles. As per my comments in the related discussion in Commons talk:Valued image candidates/candidate list#Valued images and thumbnail size_.5BUnarchived_thread.5D, I think we are free when reaching a subjective judgement to view the image as we choose. Disagreements can be resolved by discussion, consensus or balance of supports/opposes. DeFacto (talk). 16:52, 9 August 2017 (UTC)
And my feeling is, if Charles is right, it should be made very explicit in the rules. Otherwise, we will continue to have disagreements of this kind, to no good effect, with hurt feelings on both sides. Ikan Kekek (talk) 02:15, 10 August 2017 (UTC)

Proud of PREDATOR[edit]

What did you say? Zamranhusain2006 (talk) 10:47, 25 January 2018 (UTC)

Retrieved from "https://commons.wikimedia.org/w/index.php?title=Commons_talk:Valued_image_criteria&oldid=282681472"