Commons talk:Quality images candidates/Archive 9

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 5

←

Archive 7

→

Categorizing Quality Images

A discussion about categorization of Featured Pictures is being held here]. Because of the generality of the subject, it is expected (and desirable) that it will spread to other foruns: Quality Images, Valued Images and beyond. Please participate. Alvesgaspar (talk) 10:23, 22 December 2014 (UTC)[reply]

When to close a CR

I see that some images that enter the CR 6-7 days after nomination are closed quickly, after one day, or even less of CR. Ask user:Livioandronico2013 why he is closing them so quickly and he replied that had passed 8 days since nomination and I told him that the he rules of CR say that "After a minimum period of 48 hours since the last entry, the decision will be registered at the end of the text using the template {{QICresult}}", this means that a CR should remain open for at least 48 hours (2 days).

So when should a CR be closed for an image that enters CR after being in the nomination page for 6 or more days? --C messier (talk) 11:05, 1 January 2015 (UTC)[reply]

The CR are closed after eight days not in 48 hours and neither after 6 or 7 days, put the examples so maybe someone understands, thanks--LivioAndronico talk 11:45, 1 January 2015 (UTC)[reply]

I just close them when a reasonable time has passed since last comment or where there's a clear consensus. -mattbuck (Talk) 11:52, 1 January 2015 (UTC)[reply]

How I do --LivioAndronico talk 18:32, 1 January 2015 (UTC)[reply]

Example:

A file is nominated at 10 December and moves to CR at 17 December. Should the CR close (if there isn't clear consensus) at 18 December (8 days after nomination) or 25 December (8 days after CR)? --C messier (talk) 14:51, 1 January 2015 (UTC)[reply]

No in this way what file,live an example visible. Which file has been processed in this way ? --LivioAndronico talk 16:06, 1 January 2015 (UTC)[reply]

Now File:Scenery of Pine Township, Columbia County, Pennsylvania.JPG nominated at 22 December 2014, review given at 27 December 2014, then a negative at 03:39 30 December 2014 (before the bot, so let's count it as 29 December), so CR started at 30 December and the next day, with 1 for and 1 against is posted the result with QICresult template writing "Inconclusive result after 8 consensual review days", although it was only 1 day in CR (without obvious result).--C messier (talk) 16:25, 1 January 2015 (UTC)[reply]

Anyway don't happened never (or is very rare) the case you mentioned,but in this way is clearly written "8 days after nomination" --LivioAndronico talk 16:08, 1 January 2015 (UTC)[reply]

The QICresult template writes "Inconclusive result after 8 consensual review days" not 8 nomination days. 8 days after nomination applies for files in the Nominations section.--C messier (talk) 16:25, 1 January 2015 (UTC)[reply]

Is write "8 days after nomination"--LivioAndronico talk 18:30, 1 January 2015 (UTC)[reply]

Agree to C messier. Some days ago an image in CR disappeared less than 12h after my vote. This has formerly not been common. Why this change? -- Smial (talk) 17:23, 1 January 2015 (UTC)[reply]

Smial what case? We want to put the examples or we talk about air? It then depends on the vote if it was already long enough and your vote was not decisive that we have to keep doing the photos 20 days? --LivioAndronico talk 18:30, 1 January 2015 (UTC)[reply]

The Wikimedia projects have long held that "votes" should be determined by consensus. If there is any question, we do whatever is required to achieve consensus.

(1) While it is written "8 days after nomination", it seems the intention was "8 days after nomination to CR" since the QICresult message is "Inconclusive result after 8 consensual review days". Under normal circumstances, we should have 8 days to review a CR. Time is required to achieve consensus.

(2) A review should never be closed less than 48 hours after the last comment, whether or not 8 days has passed. It's an active discussion at that point and should remain open. Side comments such as "Thanks for the edit!" or "Happy New Year" that do not have anything to do with the review should not count towards the 48 hour timer.

(3) There are acceptable reasons to close early:

Nothing to Discuss: A CR has one oppose vote. Changes are made and the original reviewer changes the vote. This should be treated as if it never made it to CR.
Overwhelming: A CR has overwhelming support or opposition, say, 1 vs. 6. See (this example).
Unanimous: A CR has only votes in support or opposition because a vote was withdrawn or changed. (See this example).

Otherwise we must allow time. Not every reviewer is active every day of the week.

(4) These were closed too early:

Sensible judgment is required. If an image is closed early, it should be done so because consensus is obvious. The one principal that matters isn't in the written rules but instead can I close this review with consensus?. -- Ram-Man 18:51, 1 January 2015 (UTC)[reply]

So instead of writing should be like or should be read we write "8 days after nomination to CR and at least 48 hours after the last vote".Pro or Con? --LivioAndronico talk 20:07, 1 January 2015 (UTC)[reply]

Yes. -- Ram-Man 21:07, 1 January 2015 (UTC)[reply]

That is exactly how it was handled in the past. We had several times some deviations from this interpretation, but it happened almost when someone applied the QICresult-Tag for his first time. It also happened to me, when I was newbie in QIC. Updating the manual would be a good idea. --CEphoto, Uwe Aranas (talk) 08:50, 6 January 2015 (UTC)[reply]

Encourage or require original versions to uploaded for QI consideration

We used to encourage contributors to upload original versions and then overwrite their preferred versions. Especially with JPEG images, information and metadata are lost with every edit. It benefits the project to have the original available in the edit history. By seeing the original version, QI reviewers would be better able to judge the merit of an image apart from its editing. As tastes, tools and monitors improve, information lost in downsampling could be recovered by reediting the original. The downsampling rule can be dropped. --Walter Siegmund (talk) 17:23, 18 January 2015 (UTC)[reply]

Nonsense! We have what we get. We can voting only for that! It is better to get a low resolution then nothing. And of course it is much better to get the highest resolution, but we cannot force our donators to upload the highest resolution or the RAW files!!! --Alchemist-hp (talk) 17:44, 18 January 2015 (UTC)[reply]

Oppose No, thanks. There is already too much unedited Flickr-Garbage in Commons; no need to waste space with raw files. --CEphoto, Uwe Aranas (talk) 17:45, 18 January 2015 (UTC)[reply]

Alchemist-hp; the no-downsample rule is equivalent to forcing "our donators to upload the highest resolution" to participate. RAW is not supported, so an exemption would be necessary if the original is RAW (COM:FT). Cccefalon; storage costs very, very little and no one is forcing you to look at earlier versions, if you don't want to. --Walter Siegmund (talk) 18:01, 18 January 2015 (UTC)[reply]

There istn't an argument. A donator can manipulate the EXIF up of a 2-3MP camera like a KODAK DC4800. And you can't see a difference. It is this what you like for our commons image library? --Alchemist-hp (talk) 18:40, 18 January 2015 (UTC)[reply]

We don't need more rules, if enforced the ones we have are enough. We're trying to prevent systematic abuse, not catch every possible EXIF manipulation or occasional (mostly minor) downsample. -- Ram-Man 01:23, 19 January 2015 (UTC)[reply]

Exchanging files during review

Is this allowed? --Code (talk) 11:44, 21 January 2015 (UTC)[reply]

Only when the original file was renamed for the usual reasons like "typo in filename" or "Misleading title". An image cannot exchanged any longer when someone already started a review. In my opinion, this is justyfying a decline with the advise to withdraw the image and renominate again. --CEphoto, Uwe Aranas (talk) 18:26, 21 January 2015 (UTC)[reply]

editor/rewiewer? rules precision

In the guidelines there is the sentence : Any registered user, other than the author and the nominator, can review a nomination. Sometimes we edit images from other authors and usually we abstain from voting. But it seems that it is no always the case, and I don't like very much that a users can vote for it's own version of an image. I propose that we right a more precise sentence :
Any registered user, other than the author, the nominator and the last editor, can review a nomination. -- ChristianFerrer 12:36, 27 January 2015 (UTC)[reply]

+1 --Smial (talk) 12:46, 27 January 2015 (UTC)[reply]

+1 –Be..anyone (talk) 13:20, 27 January 2015 (UTC)[reply]

-1 We don't allow the nominator and author to prevent collusion and sockpuppets. The proposed rule is fine for normal reviews, but not for consensual review. In a consensual review, the editor's vote should still count, otherwise we discourage helpful editing because it might lose a vote and a 5/4 split vote is not uncommon. The rule should state this:

The last editor's support vote cannot be the only support vote used to promote.

This covers both scenarios. Remember the purpose of this rule is to ensure that enough eyes are on the image, not to disqualify people arbitrarily. The nominator/author's vote is treated as one vote and implied. We essentially want a +2 vote differential to promote. My rule treats the editor as a half vote (can break ties, but not meet the +2 threshold alone). -- Ram-Man 15:42, 27 January 2015 (UTC)[reply]

By definition sockpuppets have different names especially for bypass this kind of rules, and to avoid collusion is especially the purpose of what I want, and this prevents anything, there is enough good rewiewers, not need for the editor's votes IMO. I apply this rule in 99% of the cases when I edit an image. Here it's the only exemple of an image I edited and promoted, and to be precise I promoted it before to edit it : after to have try something on my pc, I was so happy that I uploaded it with the agreement of mattbuck, the author/nominator. And further evidences that it will not prevent usefull edits, I already apply this rules and many of my edits are promoted here and even in FPC : 1, 2, 3... and I can find many others QI or FP. And usefull edits are (always) recognized by the good rewiewers, not need for editor's votes. I repeat myself, forgive my poor English language. -- ChristianFerrer 19:34, 27 January 2015 (UTC)[reply]

Your explanation and position are clear. I stand by my position though. I don't see any benefit to the rule adjustment as proposed, only impediment. My suggestion is less restrictive while still accomplishing the same goal. -- Ram-Man 20:35, 27 January 2015 (UTC)[reply]

Agree with the initial proposal by Christian Ferrer.--Jebulon (talk) 23:12, 27 January 2015 (UTC)[reply]

Of more it is also the work of the editor is judged here. -- ChristianFerrer 06:51, 28 January 2015 (UTC)[reply]

+1 Any more precise definition would be fine with me. I understand, that Christian is pointing out the usual behaviour that we followed up the last years. For the sake of clarification to all reviewers, we should modify the said sentence. --CEphoto, Uwe Aranas (talk) 07:48, 28 January 2015 (UTC)[reply]

+1 --P e z i (talk) 18:49, 28 January 2015 (UTC)[reply]

+1 --Berthold Werner (talk) 09:20, 29 January 2015 (UTC)[reply]

+1 Common sense for me, given that we have decided to keep creators from self-supporting, later editors should be subject to same rules. --DXR (talk) 08:53, 31 January 2015 (UTC)[reply]

See this Commons:Quality images candidates/candidate list#File:Poertschach Landspitz Passagierschiffe Thalia und Klagenfurt 27092014 883.jpg, to understand why I started this discussion. -- ChristianFerrer 08:32, 31 January 2015 (UTC)[reply]
Although there it seems to have consensus yet I leave this discussion open a week or ten days and we will proceed to a vote. -- ChristianFerrer 11:21, 1 February 2015 (UTC)[reply]

Remove Wikimedian Requirement

Creator

Pictures must have been created by a Wikimedian in order to be eligible for QI status. This means that pictures from, for example, Flickr are ineligible. (Note that Featured Pictures do not have this requirement.) Photographical reproductions of two-dimensional works of art, made by Wikimedians, are eligible (and should be licensed PD-old according to the Commons guidelines). If an image is promoted despite not being the creation of a Wikimedian, the QI status should be removed as soon as the mistake is detected.

I feel that this rule has always been fairly arbitrary and not really in line with our goal to help users to identify good images. With the introduction of the five image restriction, I do see even less justification for this rule. I think it makes no sense to keep contributors from nominating such images as long as they stick to the normal rules of QI (balanced nomination/review ratio, no more than five images per day). I would therefore suggest to:
(a) remove this rule or
(b) at least make an exception for "genuine" photographs (I can understand that nobody is overly keen to review many paintings as QI)
It is not a big deal most of the time, but I am unhappy each time a image is declined for this reason even though it would otherwise be a QI. Of course I am open to more precise reformulation suggestions. --DXR (talk) 08:41, 30 January 2015 (UTC)[reply]

Support I'm completely of the same opinion. --Code (talk) 09:12, 30 January 2015 (UTC)[reply]

Oppose This would rise the probability, that any issues will be never resolved by the author. And even when a wikimedian is nominating it and willing to fix issues, he can only work on a jpg but not on the original source. --CEphoto, Uwe Aranas (talk) 09:15, 30 January 2015 (UTC)[reply]

Sorry Uwe, but I really don't think that this is a strong argument. Basically we would exclude all potential images because some might be faulty. I agree that not all noms will be QI, but then we can either fix them (if it is just a dustspot) or just plain decline them. In fact, I think that declining might even give us less drama in such cases because there is no creator who is personally hurt in their feelings leading to unreasonable CRs. Also, a good quality JPG can be a very acceptable source for straightforward post-processing that doesn't alter highlights or shadows. --DXR (talk) 09:26, 30 January 2015 (UTC)[reply]

Oppose "Quality images sets out to identify and encourage users' efforts in providing quality images to Commons." would be shortened to "Quality images sets out to identify quality images." This would destroy an important aspect of this event here. -- Smial (talk) 12:01, 30 January 2015 (UTC)[reply]

So your contributions would be valued and encouraged less because more images could become QI? Isn't uploading an image with a free licence to flickr etc. decent enough? I don't know if I am in a minority here that actually considers QI a service to the Commons community first and a reward to the photographer second (after all the bragging rights associated with many QIs are very, very scarce). I can imagine several cases where our current restriction actually creates a disadvantage for users. For example, many aircraft cats have hundreds of images, not few of them in miserable quality, but we are unable to promote the good ones to QI because of the aforementioned ban. VI fails here because e.g. the Scope "F-16, front view" allows only one image and crazy scopes would be needed to cover a sufficient amount of images. Would promoting some of the images uploaded to flickr or taken by public servants really harm any user here? Or in a more simple way: Does somebody who already contributes good quality images (because it's their job) really need encouragement or should we simply recognize that a given image is strong and of good quality?--DXR (talk) 12:21, 30 January 2015 (UTC)[reply]

There is no connection between my images and my work here at QIC. When late MBDortmund years ago invited me to join here, it was a matter that he was looking for someone he thought was competent enough to assist in processing the large amount of images. That is why I have proposed quite a few own pictures in all the time, in the early days even nominated predominantly photos of other uploaders. Today, there are in fact mostly self-nominations, partly in an amount such that we now find this five-pictures-limit necessary. If tons of images from flickr imports or US GOV or PR photos of all kinds will be added, I for my part would ignore them or simply leave this event. -- Smial (talk) 17:26, 30 January 2015 (UTC)[reply]

Yeah, that is exactly my point. It would not reduce the quality or quantity of contributions. By the way nobody is talking about "tons of images". I repeat myself, but since most people are mainly focused on their own images, we would see several per day, max. And those nominators would be reviewing other images. --DXR (talk) 22:37, 30 January 2015 (UTC)[reply]

Comment I'd be interested to know what size the pool of externally-generated decent-quality images is compared to user-generated. Some of the upload bots generate thousands of images, many professionally taken. Granted, many of them would be pixel-peeped to death when nominated here, but many others will be fine. Reviewing images for what QI standard is supposed to be takes some time, far more time than a bot slurping images from another website. I'll say again that I think WP needs a quality rating that can be set by individuals rather than by a community process. Then if people want to spend their free time reviewing externally generated content, they can do so without hindrance of 5-per-day limits or wasting the time of those who don't. We'd perhaps need some "trusted reviewer" flag. But with that, people could self-asses their own images or their own bot-uploads. Thus serving the need for users to find good-enough images and avoid the rubbish, but without all the overhead, edit conflicts, discussions, personality politics and games playing that can feature here. This wouldn't replace QI or FP, but be a level below. -- Colin (talk) 12:45, 30 January 2015 (UTC)[reply]

I completely agree with you concerning the trusted reviewer concept. My suggestion for the rule change is only valid assuming that the project described by you remains a theory. I am not sure what would be an appropriate forum to discuss this concept, though. Regarding workload, I think that the net effect would not change if the nominator sticks to our rule of "review as many as you nominate". Of course we could not rule out people just dumping crap into QIC based on the rule change, but right now even a person with 100+ QIs (and therefore arguably a decent idea of the guidelines) can not nominate a third-party image. --DXR (talk) 13:26, 30 January 2015 (UTC)[reply]

Oppose QI is to encourage people to contribute here. Regards, Yann (talk) 17:39, 30 January 2015 (UTC)[reply]
Oppose per Yann. Thibaut120094 (talk) 17:41, 30 January 2015 (UTC)[reply]
Support We can still have the purpose "Quality images sets out to identify and encourage users' efforts in providing quality images to Commons." even if we allow non-wikimedians. Why can't we have more than one purpose? It's not a zero-sum game. -- Ram-Man 18:50, 30 January 2015 (UTC)[reply]
Oppose per opposers. Please extend this rule to FP.--Jebulon (talk) 18:53, 30 January 2015 (UTC)[reply]
Oppose QIs are awarded to encourage Wikimedians, no? If the creator is just on Flickr, then that won't happen. --Jakob (talk) 21:12, 30 January 2015 (UTC)[reply]
Comment Well, I am not going to fight against consensus, but I am somewhat disappointed that few people even try to debate the arguments I have laid out. I have not seen anybody explain to me how the encouraging factor of QIs would be damaged by a few other nominations. I have noted that the nomination limit rules would still have to be honored, leading to no more unreviewed images. The current rules make the quality of an image as seen by less experienced users dependent on its creator even though this is plain arbitrary. Sometimes it wouldn't hurt to have the mindset of a curator rather than see everything from the photographer's viewpoint. --DXR (talk) 22:37, 30 January 2015 (UTC)[reply]
- Well, Commons does have a tendency to vote rather than discuss, and voting is rightly discouraged on those wikis that value social interaction above upload-count. As volunteers, human nature requires some kind of reward for the effort made. The level of standard (apparently) required at QI may require some improvement by the photographer. We see this at FP where someone nominates a Flickr image or a WLM image and the pixel peepers kill it off for having a little CA or noise. When the photographer can't defend or fix these issues, it doesn't work so well. Tastes on Commons FP/QI are very conservative vs the rest of the world, with minimal (apparent) post-processing, and don't even think about nominating contemporary B&W, low-key, high-key, or any other artistic expression. I agree that for our end-users, QI rules make no sense at all. They just want to find great images, no matter if the photographer has a Commons account or a Flickr account. But QI is just too much hard work to waste on non-Commons users: every time I try to vote on QI I get an edit conflict. So I'm just pretty discouraged from voting and WMF have plenty money to spend if they want to fix that. -- Colin (talk) 23:00, 30 January 2015 (UTC)[reply]
- Since I came back the trend has been as follows: Restrict nominations to 5 per day, restrict people who edit photos from voting, and don't open up QI to help identify all good images. At least the over-extreme pixel peeping has lessened. And overall despite the occasional minor controversy or bickering amongst reviewers, things have been going quite smoothly and effectively. A full reform of QI (changing the evaluation criteria and thus the scope of QI) is unlikely. -- Ram-Man 00:37, 31 January 2015 (UTC)[reply]
Oppose, when I nominated something from absent photographers recently the review ended almost immediately, because nobody is in the position to improve the obsolete JPEG image format without access to the raw photo. In essence I wasted the time of the reviewers. FPC is open for exceptional external photos, QIC is for tweaking. –Be..anyone (talk) 02:18, 31 January 2015 (UTC)[reply]
Oppose QI was set up to recognise and encourage people to contribute a broad range of good quality photographs to Commons, it came about because despite being an active project for about 2years it was impossible to identify just 12 unique images of one subject no matter how broad that were the work of our community. Having contributed QI assessed works is one criteria used by Wikimedia Australia in assessing the impact of photographers work when applying for the various assistance programs available. Gnan garra 06:41, 31 January 2015 (UTC)[reply]
Oppose Although it is the wrong place to discuss this here. Dnnoch I allow myself to state that on commons absolutely no import flickr photos or similar platforms should be approved. Photographers who know flickr, certainly also know commons. When photographers choose to upload their pictures on flickr, then they do this well aware of how those photographers on the commons upload their images. I therefore consider such more than doubtful imports because they even snubbed members of commons in some cases. Some commons Photographer certainly no joy when he finds that his photos, which he did for commons and for which he ultimately invested time and money in a wikipedia article ultimately a photo of a flickr photographer is used. Therefore, the images from Flickr to stay where they are. On QI or FP these photos do not belong; also because the photographer of the opinions to his image can not benefit. --Steindy (talk) 23:42, 1 February 2015 (UTC)[reply]
Oppose for the reasons mentioned above. --Leyo 23:57, 1 February 2015 (UTC)[reply]

Wichtigtuerei

Das grenzt eher an Wichtigtuerei : comment made by Hubert. It becomes very hard these days to do our rewiewer job to calmly. I am not the first nor the only one experiencing this kind of comment following one of my exams, others before me have probably been abused for simply trying to do their job. I think we should consider tightening the rules, and to write its in the guideline, for those who do not respect others. For to have the opportunity to banish those who do not make any effort either in image correction nor politeness. I'm a bit busy now but I will myself think about to a series of proposals so that we can work calmly and without fear of criticizing the images of others. Feel free to do the same. -- ChristianFerrer 12:08, 4 February 2015 (UTC)[reply]

I've listed my suggestions in the order of priority:

Assume Good Faith: A good number of "disrespectful" comments are just imprecise language usage, often by non-native english speakers. There is no reason to overreact to such comments.
Most of the time you just have to ignore the comments (i.e. don't feed the trolls)
On rare occasions, you just need to ask a person to stop nicely on their talk page.
If all else fails, someone can take it to the administrator's noticeboard, but if you didn't try the first three steps, then you're going to have a lot of trouble getting sympathy from the administrators.

-- Ram-Man 17:30, 4 February 2015 (UTC)[reply]

Laughable! Flowers of cherries are even white and not black.

Another mark of respect for the rewiewer. -- ChristianFerrer 17:03, 4 February 2015 (UTC)[reply]

Do you mean disrespect? If so, this is being overly sensitive. Calling specifics of a review laughable is not really disrespectful. It's a disagreement, nothing more. -- Ram-Man 17:22, 4 February 2015 (UTC)[reply]

Ram-Man Disagreement? you mean like we have we both on this topic. :) But me I will not say your comment is laughable, and for your information as for the QI levels I have my own concept of respect and disrespect levels and me I say these levels are not less good than yours. If you accept that they say that our rewiew is pompous or ridiculous, you also accept that I say it's disrespectful to my standards, it works in a way so it works the other. And I hope that the sympathy of administrators for me it does not vitiated because I am friends with many of them. I appreciate truly speaks to you, your attempt discussion and your solicitude, really. But it change nothing. -- ChristianFerrer 18:28, 4 February 2015 (UTC)[reply]

Meine Bilder sind gut, und ihr seid nur zu blöd, um das zu erkennen

...and again another exemple, we should write a book. -- ChristianFerrer 05:48, 5 February 2015 (UTC)[reply]

@Christian Ferrer: Maybe it's a language problem, but nobody really said that. Kreuzschnabel just said that Steindy's attitude was like this. --Code (talk) 05:54, 5 February 2015 (UTC)[reply]

Yes indeed. One way or another that kind of talk not belong here, in the QI page we speak about the image and we do not do the trial of rewiewers. The right for people to push their image to CR is widely enough. We nominator should have the right to respond only with templates

Done or others .... -- ChristianFerrer 06:05, 5 February 2015 (UTC)[reply]

QI Archives

Its pretty difficult to find out an archived image. Like now, I was looking for this one, didnt find it out normally as the archive for this month still does not exists. Suprisingly it is not in the archive Quality images candidates/Archives February 2015 linked from the main page but in category:Quality images candidates/Archives February 09 2015. So all system is confusing.--Juandev (talk) 08:58, 9 February 2015 (UTC)[reply]

Review of Downsampling Rule

I feel this rule is out of date and irrelevant. QI should be based on the final image, and meeting a base resolution threshold. I for one typically downsample. I use a Canon 7D which is notoriously noisy, especially at high ISO or low light - so I despeckle, downsize and sharpen to get result that I think best suits my subject. Additionally, sometimes I crop and sometimes I don't (again, to get the framing right, rule of thirds, or just cut out distractions outside my control to ensure my subject is the clear focal point). Rather than having odd-sized images, I resize to get consistent size and proportions which I feel are important across a gallery or portfolio. I also use standard sizes and compressions that I think best suit my internet uploads. Personally I feel that the downsampling rule in QI interferes with a photographer's right to display images at their best. This is the same control that I had in the old days when printing, or processing an image in the darkroom. At the end of the day (for me) its the result that counts, and the fact the images are over a stated resolution threshold. I do feel the rule is arbitrary, not conducive to how I like to present my images, and is likely to drive some photographers away. This potentially reduces the worth of commons. For me, I'll likely just stop uploading and concentrate on my Flickr account (setting the licence there if I choose to make it CC). If you stay on Commons and want to encourage more photographers then I suggest the rule be reviewed - although its a community choice at the end of the day. --Bald white guy 00:37, 17 January 2015 (UTC)[reply]

Some points:

(1) The rule applies to FPC and QIC, but are only guidelines on the rest of the commons. You can upload whatever files you want to the Commons, but if you want an FP and QI badge, follow the rules. We want the to highlight the best we can, so we restrict artificial limitations.

(2) Commons serves varied audiences and applications and requires the most useful possible images. It's not about the photographer's needs. Contributors may upload images in any format, including uploading two versions of an image, a high resolution one and an edited one.

(3) Years ago in FPC, some users would downsample to 2MP and use the images and image description pages as "advertising" for selling them on personal websites at full resolution. While we allow personal web page links, we don't want users gaming the system. The intention must be to provide the best possible, as the rule states.

(4) With 4k and higher resolutions becoming more common, this rule is even more important than ever. 2MP is really just insufficient and when these images are blown up on my screen, they just don't look very good.

(5) There are mitigating circumstances for downsampling: If an image is taken at f/22 on a 1.5x sensor, then the effective resolution is limited by diffraction (to ~3MP). Downsampling to that threshold is OK, as no real data is lost. EXIF data is required to evaluate this, however, so if there is no EXIF data, we must assume that this is not the case.

(6) Encouraging photographers is not the primary goal of QIC. This was discussed at length when the 5 nomination per day rule was decided upon, and I won't repeat that discussion here. -- Ram-Man 02:13, 17 January 2015 (UTC)[reply]

That's fine. Your choice. I initially started uploading images because the wikipedia images on things that I was interested in (Australian locations, wildlife, culture etc) were lacking or relatively low technical quality. IMHO my images were often superior ... even if "downsampled" and added something to the quality of the wikipedia articles I was illustrating. Lighting, framing, clarity, and portrayal of the subject were the key things for me - and, with this focus, I think I got a better result than some "quality" Images" that I saw with similar subjects. I then uploaded further images even though I wasn't putting them directly into articles. For me it was helping build a library of images, and I saw that QI certification was a useful acknowledgement taking all facets of an image's quality standards into account. Quality isn't limited to those images being at the native resolution of the device that generated them.

Maybe it's selfish, but acknowledgement of image quality is something that I saw as a useful incentive for putting good images into Commons and available under the CC licence rather than posting them online elsewhere and retaining full copyright. This allows for a much broader coverage of subjects and non Euro/US-centric geography. However, if my images become ineligible for QI based purely on the downsampling factor - and are not eligible on overall merits - then I see less point in me participating since there's little in it for me personally. I see the policy as being somewhat officious, and detrimental in the long run. The stated threshold is 2MP, which was met. If you think it insufficient because you've recently upgraded to a 4K monitor then maybe you should really just be pushing for the threshold to be changed. Let's hope for sake of other contributors that you don't upgrade to an 8K monitor too soon. --Bald white guy 03:35, 17 January 2015 (UTC)[reply]

I do empathize with the desire for personal incentive. So long as it doesn't conflict with the Commons' goals, I have no objection. Your images are very good and I do hope you continue to contribute since higher resolution versions would almost certainly receive QI badges. But this is QI where we focus on quality foremost and it is here that resolution requirements play the most significant role (out of QI, VI, and FP).

2MP is the mimimum and I don't feel any need to encourage doing the mimimum. If you had the ability to take a 18MP image and you uploaded a 2MP image, then you are not doing all that you could. If you took a great photo with a 2MP smartphone because that's the only choice you had, then that would (possibly) be fine for a QI. The rule on downsampling has a different purpose than the rule on resolution.

There is not enough consensus to change the resolution threshold. Some people think 2MP is fine while others do not. Some only require images to look good at 100% regardless of resolution. The 2MP threshold is also there so that we don't have to outright reject certain types of images (such as those from older cameras, those from sports photography, etc.). It's possible that the images I objected to will have a majority in support. For me, the 2MP minimum rule, the downsample rule, the proper exposure rule, and the composition rule are all important. -- Ram-Man 03:58, 17 January 2015 (UTC)[reply]

All this discussion and I neglected to state the most important point: Downsampling never improves the image, at worst it loses data and at best it does nothing. It does optimize for a specific viewing scenario (typically 100% on a 100ppi monitor). This is an issue of the merits of the image. -- Ram-Man 04:23, 17 January 2015 (UTC)[reply]

From what you say your images are by definition good for COM:VIC, I don't get your problem. I also don't get why there are no valued media or how Help:MOTD works, but clearly this isn't the talk page where I'd expect to find answers for my V-questions.

–Be..anyone (talk) 09:15, 17 January 2015 (UTC)[reply]

First of all, I agree that downsampling isn't a good idea (for the obvious reason: downsampling never improve the quality, only reducing the size). Secondly, I also agree that encouraging photographers is not the goal of QIC. But I also want (which is more important) to emphasize what QI is about:

And its easy for me. QI is only about good enough quality. Its a good and useful project (it makes it easier for people to find images with a decent quality). For image quality, it is completely irrelevant if a 10MP image was taken with a 10MP camera or with a 24MP camera (and downsampled to 10MP). Unlike FP, QI is not and had never been about highlighting the best.--ArildV (talk) 10:29, 17 January 2015 (UTC)[reply]

There are at least two reasons to downsample images moderately. First: perspective correction is a destructive process, esp. when applied to .jpg instead to uncompressed originals. It allways adds unsharpness in parts of the image, so in my opinion a slight downscaling is an technical improvement, not a severe loss of information. Second: High ISO images at full resolution do not have more information than the same image somewhat downscaled, but usually show heavy unsharpness. Anybody who wants to use such an image needs to downscale it anyway. In both cases I do not support scaling of an 36MPix-d800-picture down to 2 Mpix, but if an image is for technical reasons scaled to about 6 MPix as a minimum, this must not be the only reason for decline. Btw: It's a joke to demand high photo resolutions because of monitor sizes. In theory you can enlarge screens, and camera sensors up to 100000*100000 pixels, but you can not overcome some minor facts such as diffraction. In fact today the resolution of the D800 or of Crop-DSLR above 16 MPixels can only be achieved under optimal conditions - rock stable tripod, best available lenses, never stopped down below f/5.6, maximum ISO 200, etc. -- Smial (talk) 12:04, 17 January 2015 (UTC)[reply]

It's not a joke, but the original intention of the rule. We want to provide images that look good on large physical prints (~20-40MP), high resolution monitors (8MP), and other such uses. On a 36MP D800, diffraction begins at around f/5.6, but is not completely cut off at all wavelengths of light until it passes f/11. Of course other factors (iso, lens quality) further limit resolution. This is understood and the scenarios you've provided would not likely result in any opposes. -- Ram-Man 14:34, 17 January 2015 (UTC)[reply]

Comment A downsampled image would not be FP, as long it is of good quality at the required size, it is OK for me. It doesn't mean downsampling should be done, or create better images, but if you required 20 to 40MP, you automatically prevent most people to participate. Not so good either. Regards, Yann (talk) 14:59, 17 January 2015 (UTC)[reply]

This isn't that complicated. We have one rule for resolution, that's 2MP. We have another rule for downsampling. Whether it be 40MP, 20MP, or 4MP, we just don't want downsampling (read: destruction of quality). There is no implication that a particular camera must be used. So long as no one downsamples, no one is excluded. -- Ram-Man 21:31, 17 January 2015 (UTC)[reply]

Ok, then please remove the template from all of my QI, because I nearly allways use downsampling. Thanks. -- Smial (talk) 23:21, 17 January 2015 (UTC)[reply]

What for an absolute nonsense discussion! The rule for a downsampling can be completely striked. This is complete unnecessary. Nobody can see is it a crop, a true or a downsampled image. All people that say others are only pixel counter and not photographers for me! That remains my opinion for ever!!! --Alchemist-hp (talk) 00:50, 18 January 2015 (UTC)[reply]

Sometimes slight downsampling is even necessary to get the whole information of distorted photos on the edit screen before - by applying a special crop - being able to salvage necessary informations that haven't been accessible before. --CEphoto, Uwe Aranas (talk) 08:44, 18 January 2015 (UTC)[reply]

Yeah, I think downsampling should be handled like anything else... with some common sense. Downsampling just above the 2MP mark is not ideal and a bit annoying, because this restricts many uses. Downsampling to 5 or so MP is absolutely fine and a good quality file of that size is, as said above, usually far superior than most incumbent images on the wikipedias. And one word of advice: Nobody should ever not upload a file just because it is not going to be QI. 95% of users are completely oblivious to the QI sign and its meaning, they just want to see a useful image! --DXR (talk) 09:28, 18 January 2015 (UTC)[reply]
Downsampling isn't strictly enforced -- otherwise none of Diliff's images would be FP. But nobody could claim Diliff is being stingy with the megapixels. We don't have a strong consensus on whether images should always be uploaded at out-of-the-camera resolutions or whether, as artists, we can choose to present our work at a size where it looks great. Both our quality-judging forums have problems with reviews being done at inappropriate magnification, which just encourages nominators to downsample to avoid pixel peeping scrutiny. Sadly the MediaWiki interface doesn't make it easy to upload full-size but indicate a best-size that you intend to be reviewed at. Users downsizing their 24/36MP images to 6 or 8MP should expect the full force of any pixel peeping comments because they are not playing equally with those who retain their MP. Similarly, those who significantly downsize to limit the reuse potential of their images should not expect to win any quality awards on Commons. This isn't "Quality Thumbnail" forum so, no, it is not just about capturing a great image but also about being generous with donating an image that is useful for anyone for any purpose. That's our mission after all. That might include A4 print or filling an iPad screen. There are however valid reasons why particularly wildlife photos tend to be smaller than e.g. photos of buildings, landscapes or studio portraits, and we make allowances for that. Looking at Bald white guy's uploads, I see images that are tiny both in MP and KB. We don't do our re-users any favours by encouraging such tiny images by awarding quality badges. -- Colin (talk) 12:37, 18 January 2015 (UTC)[reply]
-- Ram-Man 01:32, 19 January 2015 (UTC)[reply]
That "no downsampling" rule is a complete nonsense and should be removed. Minimal resolution can be increased if there are too many "small" images and it's completely author's private matter about how he will provide that required resolution. --[Tycho] (talk) 23:41, 12 February 2015 (UTC)[reply]

Review of Downsampling Rule

I feel this rule is out of date and irrelevant. QI should be based on the final image, and meeting a base resolution threshold. I for one typically downsample. I use a Canon 7D which is notoriously noisy, especially at high ISO or low light - so I despeckle, downsize and sharpen to get result that I think best suits my subject. Additionally, sometimes I crop and sometimes I don't (again, to get the framing right, rule of thirds, or just cut out distractions outside my control to ensure my subject is the clear focal point). Rather than having odd-sized images, I resize to get consistent size and proportions which I feel are important across a gallery or portfolio. I also use standard sizes and compressions that I think best suit my internet uploads. Personally I feel that the downsampling rule in QI interferes with a photographer's right to display images at their best. This is the same control that I had in the old days when printing, or processing an image in the darkroom. At the end of the day (for me) its the result that counts, and the fact the images are over a stated resolution threshold. I do feel the rule is arbitrary, not conducive to how I like to present my images, and is likely to drive some photographers away. This potentially reduces the worth of commons. For me, I'll likely just stop uploading and concentrate on my Flickr account (setting the licence there if I choose to make it CC). If you stay on Commons and want to encourage more photographers then I suggest the rule be reviewed - although its a community choice at the end of the day. --Bald white guy 00:37, 17 January 2015 (UTC)[reply]

Some points: