Commons talk:Quality images
Top files of the photo challenge as a source for Quality Images
I think many of the top-voted files of photo challenges could be made quality images (or featured pictures or sth else…not sure which).
I suggest somebody or people here systematically go through them and make many of them quality images so that good-quality content can be easier surfaced. Or at least, if anybody here is looking for an easy way to find more quality images, these photo challenges would be a low-hanging fruit.
- See Category:Photo challenge winners – I think one could browse them like so (maybe also exclude FPs etc): deepcategory:"Photo challenge winners" -deepcategory:"Quality images" (~597 files but deepcat has problems)
- Additionally, I suggested that more of the top-voted files go into a new category, not just the top 3 per challenge – see archived discussion about this here (also ping@Jarekt: )
- Until then, there may be other ways to query for these files, e.g. files in top spots (like 4-15) or files with at least a rating of x (e.g. >6) and one can just browse the Scores pages where they are listed in order of their ratings at the new category Category:Photo challenges voting results
This may also be relevant to the Featured pictures etc. One issue I see with Quality Images etc is that these are mostly just photos and rarely data graphics, modern art or diagrams etc. If you have further ideas relating to any of that, or are working on getting top-voted PC files to become QIs, or have any other input on this, please comment. Prototyperspective (talk) 13:24, 6 March 2026 (UTC)
- This is a good idea, even though many participants of Photo challenge nominate their photos for QI status themselves, and often you can guess the topic of the current challenge from the photos nominated for QI :) But there are also some problems. In particular, the photos that participate in challenges often do not comply with the simplest rules of QI project - they are too small or do not have necessary categories and descriptions, for example. But, of course, one can find real pearls in Photo challenge archives if you look thoughtfully. -- Екатерина Борисова (talk) 00:07, 13 March 2026 (UTC)
File:Radio Televisyen Malaysia.svg
It seems like File:Radio Televisyen Malaysia.svg is mistagged as a quality image, but before removing the tag I wanted to verify i wasn't just missing something. Bawolff (talk) 23:04, 15 March 2026 (UTC)
- Usually under "File usage on Commons" you would see some QIC archive page if the file was ever nominated for QI (independent from whether it finally got promoted or not). This is not the case for the file you provided, so I assume this is not actually a QI. Additionally, I checked the talk pages of both users who have contributed different versions of the logo. None of them have an entry from QICbot, which makes me further doubt that this is a QI. --Aciarium (talk) 13:19, 17 March 2026 (UTC)
Five nominated images per day?
I think that the rule of five nominated images per day should be supplemented with a section that indicates that no more than twenty images of a user can be nominated every four days, whoever nominates them. Lmbuga (talk) 21:13, 16 March 2026 (UTC)
- I am absent from politics. I think that what happens with User:Jacek Halicki is an abuse of the rules. In my opinion it should be corrected- Sorry if what I propose is not appropriate or convenient. I have spoken about the matter: I am silent (I'll shut up if no one says anything.)--Lmbuga (talk) 21:17, 16 March 2026 (UTC)
- I would love a weekly limit rather a daily limit as I stay out of town often for photography and unable to participate during those trips. But it is also true similar type of so many images from a user at a time will be monotonous for evaluation by the reviewers. Tisha Mukherjee (talk) 07:03, 17 March 2026 (UTC)
- I am not going to direct this matter
(@PantheraLeo1359531: @Юрий Д.К.: @Romzig: @Alexander-93: @Jacek Halicki: @Poconaco: @Pangalau: @Berthold Werner: @JackyM59: @MB-one: @Lvova: @Pdanese: @Ermell: @Sebring12Hrs: @Michielverbeek: @Syntaxys: @Tisha Mukherjee: @Poco a poco: @Mike1979 Russia: @Benjism89: @Umarxon III: @Famberhorst: @Agnes Monkelbaan: @XRay: @Plozessor: @Jakubhal: @Igor123121: @Crisco 1492: @Rjcastillo: @Tournasol7: @Екатерина Борисова: @Sebring12Hrs: @Perituss: @Jebulon: @Aciarium: @FlocciNivis: @Smial: @George Chernilevsky: @Boston9: )--Lmbuga (talk) 21:44, 16 March 2026 (UTC)
- I think you are not wrong in this specific matter. I have noticed that Jacek Halicki and Poconaco take pictures in the same area of Poland around Klodzcko (see geocoordinates). In my opinion they know each others. Poconaco does not have any user page. I am not sure that Poconaco exists for real, but I cannot prove nothing. Your proposal is appropriate, but I don’t think it will succeed. Commoners are conservative, you know… Jebulon (talk) 00:52, 17 March 2026 (UTC)
- Poconaco has never nominated his own photos, he always nominates Jacek's photos. So does Boston9. It looks like Jacek has formed a kind of gang, which has the task of nominating his photos. They do it in batches. Let's assume that if Jacek himself nominates photos 1-5 from some location, then Poconaco nominates photos 6-10, and Boston 9 - photos 11-15. About a month ago, there was a situation when they got something mixed up and all nominated the same photos together. I hope that these people really exist after all and are not meat puppets of Jacek, but in general I personally don't like this configuration. It looks like a pipeline and has little to do with the goals and objectives of QI project IMO. -- Екатерина Борисова (talk) 02:52, 17 March 2026 (UTC)
- I am generally against introducing additional bureaucratic restrictions. In this case the situation does not seem particularly harmful. I understand that there is currently some motivation to address it, but people tend to take breaks from time to time. If such a rule were introduced, who would realistically keep track of these limits once some of the supporters lose interest? Even if a user encouraged others to nominate their photos and this resulted in something that someone might describe as a "gang" (a rather strong word), it is still not a problem as long as these are independent users making their own decisions, rather than sockpuppets. -- Jakubhal 05:03, 17 March 2026 (UTC)
- I noticed that too, but I don't see a problem with it. You have to give those three users credit for providing enough ratings. Currently, all three of them together have nominated about 160 images and rated 200 images. Incidentally, the same nomination/rating ratio currently applies to you as well, Poco a poco, Юрий Д.К., Екатерина Борисова, Sebring12Hrs, and even I have occasionally nominated more than 40 images in 7 days.
- The material from these three may be somewhat trivial in terms of subject matter, and you can’t go far wrong in terms of quality, but it also requires no effort in terms of CR. There are several other users who flood QIC with similarly trivial material without conducting any evaluations themselves. In such cases, a more proactive approach will be needed. Achim Lammerts • Syntaxys (talk) 05:07, 17 March 2026 (UTC)
- I'm agreeing with the majority here, that the situation around Jacek Halicki's photos does not constitute an emergency for change. However, I could agree to change the daily limit to a weekly limit, but only if there is a technical way to enforce it. MB-one (talk) 11:14, 17 March 2026 (UTC)
- Dear participants, because I was addressed, I would like to comment briefly. Personally, I think: 1. A nominated image is not yet a rated image. So there is no risk that too many bad images will be rated well, but only the risk that too many of a user's images will be rated illegally. This means that each user still has the decision to rate the images or not. 2. I understand the aim of the competition not to be to create exclusion, but rather to motivate people to make pictures public and receive appreciation. The goal is not called into question by the existing situation. 3. If the competition is aimed at everyone in the world to make their photos public, I see the problem as being that far too few people take part rather than that there would be too much participation from individuals.
- Overall, the existing rule seems sensible to me, but sufficient. If there are individuals who want to circumvent the rule - that still seems to be a manageable number - they should do so. There is no more money or anything else of value in return, but simply more praise and appreciation. That's my opinion. Maybe she is wrong on this point. But I was asked about it. Romzig (talk) 18:57, 17 March 2026 (UTC)
- Hm, it took me like 2 minutes to check that the trio Poconaco-Boston-Jacek are three different people, who contribute independently from each other here on Commons, and not sock or meat puppets.
- I don't see anything wrong with people nominating a lot of pictures of someone else, as long this is not their only activity here on Commons. I mean in the end, they're chosing to nominate Jacek's pictures instead of their own. I'm against this new supplementary rule. FlocciNivis (talk) 18:16, 19 March 2026 (UTC)
I’ve given a lot of thought to how to approach this matter. I wanted to make it clear that I don’t wish to criticise Jackek, but I do want to know what the community thinks about the fact that I’m asking my son to create a user account and, if he likes any of my photos, to nominate them. Why not me, but Jackek? In truth, I’m seeking permission to be free and have a clear conscience. I hope we can conclude that we can all have a friend or a family member (or two) who nominate images of us, without the restriction of having to use the same computer (my son lives with me). I remember the time when Poco a poco nominate more than ten images a day. I have no doubt that this might interest him. I have absolutely no animosity towards Jacek Halicki. On the contrary: he is a user who takes me into account and has nothing but kind words for me. Things must be made clear.--Lmbuga (talk) 06:16, 17 March 2026 (UTC)
- Actually, I’m asking whether it’s really that easy to get round the rule. And if that sounds alright --Lmbuga (talk) 06:19, 17 March 2026 (UTC)
- At the moment, it doesn't seem to be too much of a problem to do it.--Lmbuga (talk) 06:24, 17 March 2026 (UTC)
- I don't see any drastic violations here.
- Let's look at the situation step by step:
- 1. Jacek is no longer a young man, and he has disabilities. As far as I understand, Jacek uses a wheelchair and a special vehicle. Under these conditions, each of his photo shoots is a feat.
- 2. Jacek is a hard-working photographer; processing and uploading such a large number of photos takes time and effort.
- 3. Jacek's photos are undoubtedly high-quality and well-made; they useful for the Commons.
- 4. Jacek's support group evaluates more photos than it nominates.
- To sum it up, I don't think the situation really requires radical solutions.
- In the nominations, i'm more concerned about the flood of very poor nominees from newcomers than about 15 good photos of Jacek. -- George Chernilevsky talk 09:30, 17 March 2026 (UTC)
Various pros and cons regarding a possible regulation of the matter have already been elaborated very well by other users in this thread. My personal opinion on the topic: Jacek's work is certainly valuable (and should be honored with QI promotion where applicable), and the fact that Poconaco and Boston9 evaluate more images than they nominate clearly benefits the QI project. However, one thing bothers me: The constellation Jacek–Poconaco–Boston9 doesn't feel organic at all, but rather arranged/coordinated (as already mentioned by others). If a photographer constantly generates high-quality work, there is nothing with multiple users nominating >>20 of his images in a 4-day period, if they have freely decided to do so. But the undertone drastically shifts if it is always the same users who are nominating always the same photographer. I don't think a strict 20-images-per-4-days rule is the way to go; maybe something like "less than 40% of your nominations from the last 14 days must be photos of the same user" or similar would be a better rule? --Aciarium (talk) 13:11, 17 March 2026 (UTC)
- I was told this back when I was exclusively nominating Tournasol7's photos, and I can understand that it was excessive. It's perfectly acceptable to vary the nominees and not nominate the same user five times a day, but requiring the person making the nominations to keep track of them for an entire week if we haven't exceeded 20 photos would be tedious. However, I would like to continue proposing photos from Tournasol7, Moroder, Dillif, Julesvernex2, etc., with the sole rule being "a maximum of five images per day per user." As for the trio Poconaco-Boston-Jacek, it's true they only nominate Jacek's photos and never anyone else's. But personally, that doesn't bother me too much, as long as they are three different individuals. Sebring12Hrs (talk) 13:51, 17 March 2026 (UTC)
- I’ve helped others nominate their photos before. If I don’t have any candidates myself, I don’t see that as a problem. The number of reviews is also appropriate—maybe not every day, but certainly on a long-term average. I’m skeptical about additional rules. It should remain as simple as possible. -- XRay 💬 16:43, 17 March 2026 (UTC)
- In Jacek Halicki's case the technical problem is that he is a very active photographer and takes several thousand good quality photos per year. Complying with the rule that one person can nominate five files for QI he would be unable to nominate his best photos as there are much more of them. So some other users from Poland are helping with nominating his images. Poconaco and Boston9 are absolutely different people than Jacek Halicki and Jacek is not operating their accounts, they nominate Jacek's photos by themselves, so there is no breach of a 1 person - 5 nominations rule. ~Cybularny Speak? 22:45, 17 March 2026 (UTC)
- I'm sorry for making you think. We've already created a user account for my son for the future. I'll start teaching him a bit. I'm tired.
I can't leave this thread without saying a few words about Jacked. I don't think anyone has mentioned it. The three users behave ethically, as if they were a single user, since they don't vote for themselves.--Lmbuga (talk) 00:28, 18 March 2026 (UTC)
- I wasn't pinged on this discussion so late to it. Why not ask Jacek Halicki to ask his friends to hold back as they are upsetting some in the community. Charlesjsharp (talk) 16:24, 28 March 2026 (UTC)
Multiple similar nominations
The QIC guidelines do not ask nominators to choose their best images like I think they might have done in the past. We have quite a number of users nominating very similar shots. Should we do anything? Charlesjsharp (talk) 16:24, 28 March 2026 (UTC)
- Í do not think so. However, I personally think that it would be way more interesting if everyone chose their best images both for their uploads to Commons and for QI nomination because viewing lots of similar images over and over again might become somewhat boring. In addition, this might be a waste of storage space and I don't understand why so many similar images are required. --Robert Flogaus-Faust (talk) 17:34, 28 March 2026 (UTC)
- I don't think this is a problem - let people nominate what they want, and reviewers review what they want. Thanks. Mike Peel (talk) 17:56, 28 March 2026 (UTC)
- Similar photos can have various reasons, such as being part of a photo series. Different perspectives can also serve different purposes. I don’t see this as a major problem and consider additional rules to be unhelpful, if not counterproductive, especially in this case. Who is supposed to decide what constitutes a significant enough difference? If the goal is truly to select the best photo from a series, then it should be nominated on VI. -- XRay 💬 07:36, 29 March 2026 (UTC)
- +10000 per Mike Peel ! Sebring12Hrs (talk) 00:06, 30 March 2026 (UTC)
- +1 Mike Peel -- Jakubhal 08:55, 3 April 2026 (UTC)
How to avoid QICbot failures
There have been three such failures in less than a month, and I think it's very inconvenient, despite the fact that Robert always corrects it promptly. He explained to me here that the usual reason for this kind of bot failure is an edit conflict. If somebody edits the candidates list while the bot is attempting to remove the images from the candidates list, the bot will just fail to remove them. So I have a suggestion. Maybe we should write an ad somewhere in a prominent place at the top of the QI candidates page? "Please try not to nominate or review images between 05:00 and 05:30 UTC, because at that time the page is updated by the QICbot" or something like that. I am not sure too whether it is sufficient to avoid these edit conflicts, but maybe this way there will be fewer failures. What do you think? -- Екатерина Борисова (talk) 00:52, 6 April 2026 (UTC)
- That might help - you could be more specific with timing, since the edits are normally at 5:08-5:13 UTC, so you could say 05:00-05:15 UTC. Or, I've been wondering about changing the bot so the edit conflict stops it completely, but I haven't figured out if I could get it to auto-restart (should be possible somehow...), or if I'd need to manually run it each time that happens. Thanks. Mike Peel (talk) 09:56, 6 April 2026 (UTC)
- I suppose that stopping the bot entirely might make things worse. As you know, I can clean up the gallery of recently promoted images and the archives quite easily after an edit conflict. Therefore, I support the suggestion by Екатерина Борисова, but with a more specific time interval, i.e. 05:00-05:15 UTC. --Robert Flogaus-Faust (talk) 19:58, 6 April 2026 (UTC)
Automatic Image Quality Assessment
Hello, are there any effort to try and automate the discovery of high quality images in Commons through AI "Image Quality Assessment" ? I tried it to sort events photographs and it works like a charm ! WebSemantique (talk) 17:51, 20 April 2026 (UTC)
- That’s certainly an interesting approach. I tried out a tool like this once and was pleasantly surprised. The inclusion of categories, descriptions, and SDC can also be helpful. However, I would use a different seal/logo for images identified this way, since the QI here is also meant to help users take better photos. -- XRay 💬 18:36, 20 April 2026 (UTC)
- taking the community out of the loop would damage the original purpose, the value the process gives to Commons, and the current benefits of QI. People are the key element to both assessing and producing Quality Images, automating the decision process will kill the artistic human elements its QI not AI. Gnangarra 04:41, 21 April 2026 (UTC)
- While AI can, to a degree, distinct sharp photos from blurry ones, IMO it cannot judge the actual quality and value of a picture. At least not today. It is good if you have hundreds of similar photos, say from the same event, and want to quickly sort out the failed ones. But not to evaluate quality of random images without context. Plozessor (talk) 10:56, 21 April 2026 (UTC)
- What I think AI could do is create a pool of images that humans could check and judge if those are indeed good candidates. One of the shortcomings of the current process is that we see almost exclusively self nominations. Such a process could widen the diversity of QIs Kritzolina (talk) 18:50, 21 April 2026 (UTC)
- Admittedly, I’ve often tried to nominate photos taken by others. It’s very (!) time-consuming to find suitable photos. Even in photo contests like Wiki Loves Monuments, there were relatively few candidates among the top 100. As a result, I rarely make the effort. A shortlist would be helpful. -- XRay 💬 06:52, 22 April 2026 (UTC)
- What I think AI could do is create a pool of images that humans could check and judge if those are indeed good candidates. One of the shortcomings of the current process is that we see almost exclusively self nominations. Such a process could widen the diversity of QIs Kritzolina (talk) 18:50, 21 April 2026 (UTC)
- While AI can, to a degree, distinct sharp photos from blurry ones, IMO it cannot judge the actual quality and value of a picture. At least not today. It is good if you have hundreds of similar photos, say from the same event, and want to quickly sort out the failed ones. But not to evaluate quality of random images without context. Plozessor (talk) 10:56, 21 April 2026 (UTC)