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.
Over the last 7 weeks or so, I'm still relatively new. Have met the "regulars" and the "newcomers". In this time, I have noticed a few things, which may have become a "normal process" for the "regulars" who help in assessing each image that gets added to the Quality Image Assessment page. In most cases, the "regulars", yes in most cases ..... get an "automatic pass" to have their image to QI. Even though the image may be of any importance. It could be best done at "Valuable Image"? Yes, I admit I don’t go around "happy snapping" everything I see around me. Just to get my image to be promoted to QI. For me, it takes days for me to think up what would be a good image for me? Where could I go? What could it be of?
As I said "most cases" regulars, get an "automatic pass". Today, I see about 4 done by the same person and they have somewhere 3,500 to 4,500 QI images. I'll either be in an Insane Asylum permanently or dead before I reach that many ..... At the rate, I am going! A good sign for me to not feel so bad, is that a "regular" and his 2 of his images has been rejected. Which has made me feel relieved. Because it’s finally and about time a regular feels what it is like when no matter how much work (thought goes into it and weeks of patience, like me is) as I don’t submit many. This person should realise they should get their head out of the clouds and come back to earth.
Another thing, I don't understand is when an image gets assessed. No two people can agree on certain outcomes. Person 1, may say blurry and CA. While Person 2, may say not straight and the verticals needs to be done. Then when the person who submitted it, can't fix what Person 2 said, then they are stuffed, needing help. I can keep going on with this. Adamdaley 05:26, 13 March 2017 (UTC)
Sometimes it is not an easy task, right. I was "new" for a lot of months. I've learned to fix all the issues, for example with Adobe Lightroom. Each reviewer has another focus to the issues. Please have a look to the reasons of the issues and then try to fix it. There are always a lot of hints in the World Wide Web. So your photographs will become QI. But please don't stop if your images are technical without issues. Today I'll try some special kind of photography but sometimes they did not become QI. It is a kind of progression when QI becomes boring. It is not my goal to take always technical perfect photographs. But be sure that most of the people will help you to take good photographs. Good luck and always good light. --XRaytalk14:05, 13 March 2017 (UTC)[reply]
Regulars know what will pass, and what won't. So they don't submit images which have very little change to get through. At least that's what I do. That's probably why they have a higher success rate. Regards, Yann (talk) 18:48, 20 March 2017 (UTC)[reply]
On the other hand some former regulars have left because of sometimes absurdly high requirements and also because of surprisingly strange justifications. --Smial (talk) 18:56, 20 March 2017 (UTC)[reply]
Making things easier
Time and time again, I read questions from new QIC contributors about things, photography terms, that are used in reviews. A long time ago, I was one of those who had no idea what "CA" meant. The old regulars are so used to this vocabulary they have probably forgot how confusing it could be to not know that "photo language". This glitch in communication is frustrating, time-consuming and unnecessary. We have a page that was started long ago but it is still very, very incomplete: Commons:Photography terms. I propose that we make an effort - a drive, to fix this page and stock it up with all the terms used at QIC and then some. When the page is up to par, I propose that we put a link to it on top of the Commons:Quality images candidates/candidate list somewhere just above the "Please nominate only a maximum of 5 images per day." orange bar. That way reviewers can simply refer to this list for explanations instead of having long tutorials in the reviews or on talk pages, The page should also have a good short-cut link that is easy to use.
Some of the explanations that are sorely missing are CA, WB, crop, HDR, stacking, fringing, tilt, perspective correction, raw, histogram, CCW, CW, jpeg artifacts, compression artifacts, noise, bokeh, blown, burned out, post processing, and many more.
The explanations should be short, concise and in a simple language, like the guideline on the page says: Please try and write in basic English: most users are not Native English speakers, nor are they photographers. The explanation should also link to Wikipedia articles for further explanation. Thoughts? --cart-Talk17:37, 20 March 2017 (UTC)[reply]
For me, I knew nothing of photography language, example "CA". I'd say finish that page with image examples and articles. Adamdaley 00:42, 21 March 2017 (UTC) — Preceding unsigned comment added by Adamdaley (talk • contribs) 00:42, 21 March 2017 (UTC)[reply]
A very good idea. I proposed a dictionary of abbreviations a long time ago too. (I had the same problems like Adamdaley in the beginning too.) But the page Commons:Photography terms is better. I'll try to help, but my English is not good enough. May be a translation in other languages may be a solution too. --XRaytalk06:19, 21 March 2017 (UTC)[reply]
XRay, any help is good help! :) At English Wikipedia I often work with fixing and sorting out texts by German, French or Spanish editors, so if you just provide what facts you can even in not so good English, I will fix the language for you. The same goes for contributions from other users. After we have a full, working page, we can look at translating it too. --cart-Talk08:24, 21 March 2017 (UTC)[reply]
Update
We are off to a good start with the list as some of our photographers has turned out to be brilliant at explaining things as well. There is now also the start of a list of If you know some abbr that should be in that list, please just add it. It may result in a new section below, which is great. We are also in need of links to good online external sites with photography terms and tutorials. If you know any, please add them to External links. Don't worry about knowing how to spell or formatting, we'll fix that for you. --cart-Talk10:17, 22 March 2017 (UTC)[reply]
Same or very similar pictures
First of all, I am not specifically about my own photos but about a general clarification. How do we deal with such cases at QIC? What are the cases?
Trimming another photo without any other modification
Same photo, different development / color / HDR ...
A crop or different development of the same picture is not IMO something we should be encouraging. Very similar photos is OK. I'd probably say that so long as they're not all nominated at the same time though... -mattbuck (Talk) 20:10, 23 March 2017 (UTC)[reply]
I've been following the Quality Images assessment closely. What I have noticed is that anything under 2 megapixels are not generally accepted. While searching a collection of images I've come across this image and it's below 2 megapixels and I am wondering how it passed? [[1]] --Adamdaley 00:47, 26 March 2017 (UTC) — Preceding unsigned comment added by Adamdaley (talk • contribs) 00:47, 26 March 2017 (UTC)[reply]
The image has more than 9 megapixels (3.390 × 2.807), so it is within requirements. File size is less than 2 MByte. This may point to high JPG compression settings, but this is not necessarily the case, because the file size also depends very much on the subject being photographed. Generally images with very high detail generate larger file size, also oversharpened images, also very high iso images (noise means "much detail" for the jpg compression engine). Images with big blurred areas or monochrome areas lead to smaller file sizes. Heavy denoising often also makes smaller files. --Smial (talk) 03:48, 26 March 2017 (UTC)[reply]
Printed .....
I have printed the following images out at my local imaging place at the size of 12 inches x 18 inches. I must say they look really good at that size. --Adamdaley 07:38, 29 March 2017 (UTC)
— Preceding unsigned comment added by Adamdaley (talk • contribs) 07:38, 29 March 2017 (UTC)[reply]
Please open for pictures
QICbot ot of order
The QICbot is broken again. A number of QI promotions the last few days have not been given their QI seal and the users have not been getting any messages. If you suspect that one of your pics has been missed, check out the archives or Commons:Quality images/Recently promoted and add the {{QualityImage}} to your image. If you have the time to help add the QI stamp to other overlooked pics too, please do so. --cart-Talk08:28, 31 March 2017 (UTC)[reply]
Block of Spurzem
For the recent trolling on QIC, I've blocked Spurzem for 14 days:
On his user talk, he is continuing trolling, accusation of unqualified reviewing / revenge votes. He also summoned Reinhard Kraasch from German wikipedia to unblock him out of process (allegedly for a complain on COM:ANU, although on Commons there is no possibility to complain on ANU about a current block, only afterwards if the block is considered unjust; to appeal for unblock, a request for {{Unblock}} on own talk page has to be used instead; besides, there has been obviously no intention to appeal for unblock). I think this is unnecessary drama from Spurzem and I'm tended to revoke talk page access too. More opinions appreciated. --A.Savin23:34, 31 March 2017 (UTC)[reply]
It's not my own case as Spurzem offends (unfortunately only in German, to prevent transparency for non-German speaking users) every user who has different opinion on his noms. You can be sure that I'm not at all sensitive for his insults (because I don't take all that trash particularly seriously), and if Spurzem hadn't insulted various other users and if this wasn't in fact a long-term problem with him on QIC, I of course wouldn't have blocked him --A.Savin21:29, 1 April 2017 (UTC)[reply]
Yes, I had a small issue with him a few weeks ago. He didn't assume good faith and got angry because I commented that one of his pictures wasn't sharp after he had commented a problem (overexposure) in one of my pictures. I was probably unskillful for having criticized after a criticism in the other direction, but as I never had an issue with him I didn't believe that that review would be problematic. In that occasion he overreacted and took a criticism personal. Things good settled though a few weeks later, so I am not sure whether that is very representative here. I would keep the block for the diffs above but would reduce it to something like 2 days. Poco222:46, 1 April 2017 (UTC)[reply]
There is no warnings, there are very likely several disputes in the past between the user and the administrator, 2 weeks a little bit too long for a first block... Well IMO unblock as fast as possible and let the experimented users of the QI page to do justice themselves, and they will complain at the admin noticeboard if Spurzem is really a danger for them or for the community. Christian Ferrer(talk)18:43, 1 April 2017 (UTC)[reply]
I do think he should consider himself warned, and should then subjected to a block in the future if he continues with his unseemly recriminations. I've observed that he has a thin skin, though it's been some time since I myself had my one (if I remember correctly) run-in with him. (For the record, it can be seen at User talk:Ikan Kekek/archive#BMW Isetta, and it was annoying but not the end of the world.) -- Ikan Kekek (talk) 23:07, 1 April 2017 (UTC)[reply]
I agree with others that it's best if a non-involved admin blocks him, as there are enough admins for that to be possible. And isn't there a board where you can discuss potentially controversial blocks before they take place? If there is, admins should use it. If there isn't, you should create one. -- Ikan Kekek (talk) 23:09, 1 April 2017 (UTC)[reply]
I don't see any offense in the German statements of Spurzem.
O.K., Basotxerri, ich weiß Bescheid --> OK, Basotxerri, I'm aware of it
Schön, dass Du Deine (?) Meinung wenigstens noch bedauerst --> Fine that you regret (your?) opinion at least
Von A.Savin hatte ich wirklich nichts anderes erwartet und frage mich tatsächlich, warum ich hier überhaupt noch mitmache.
--> I did not expect anything else from A.Savin and ask myself indeed why I still contribute here
This may be a questionable tone, especially for QIC, but in my opinion no reason for a block. BTW: One should also have a look on the age of the two opponents - I guess Spurzem is missing a some respect that in his opinion is due to him. (I know, age is no excuse for bad behavior, but this may explain some of the problems). --Reinhard Kraasch (talk) 09:58, 2 April 2017 (UTC)[reply]
I think, when you would meet User:Spurzem in real life you would know what it is about instinctively. I don't say that older persons need more respect. I only say: They (and I include myself here) often expect more respect than they get in the internet, and act irritated when they do not get it. --Reinhard Kraasch (talk) 14:34, 2 April 2017 (UTC)[reply]
Anyway, and besides all age issues: From an admin I would expect more self-control than you are showing e.g. in [2], as an experienced user you should know how fast a flame war erupts and how to avoid it. I would recommend a "military night", at least I will take one. --Reinhard Kraasch (talk) 14:34, 2 April 2017 (UTC)[reply]
My actions are made in good faith and I don't insult people on QIC or elsewhere. And if a user is blocked, I don't unblock them out of process without further consultation. --A.Savin14:55, 2 April 2017 (UTC)[reply]
Did I anywhere write that I did not assume good faith with you? But you also should assume good faith with me in the unblocking issue, especially since I have tried to explain the reasons (and asked for an excuse) on your talk page. And maybe you should also assume some good faith (or at least no malevolence) with Spurzem. --Reinhard Kraasch (talk) 14:21, 3 April 2017 (UTC)[reply]
Toll. Lauter deutsch sprechende Benuter streiten sich und einige, die durchaus involviert sind, bleiben außen vor, weil man sich auf englisch streitet. --Ralf Roleček10:20, 2 April 2017 (UTC)[reply]
Don't understand your point. It's an international project and uses English as the Lingua franca. Discussing this just in German would lock out the majority of possibly interested Commons users. Don't forget: it's a community, not just some "German speaking guys arguing with each other". --LC-de (talk) 18:15, 2 April 2017 (UTC)[reply]
In general I agree with you, but in this specific case the main question is whether the statements of Spurzem (in German) were offensive or not, which hardly can be decided by non-German speakers, since obviously even the German speakers disagree about this point. --Reinhard Kraasch (talk) 14:21, 3 April 2017 (UTC)[reply]
I don't know if it was appropriate that an involved admin blocked him but in my opinion Spurzem's passive-aggressive reactions on every comment anyone was making about one of his pictures would have deserved a (shorter) block much earlier. Just my opinion. --Code (talk) 14:01, 2 April 2017 (UTC)[reply]
It's not mandatory, however, how a good practice and highly recommended the sysadmin involved should do a public request for help of another admin(s) preferably German speaker in this particular case. This is just my humble opinion and is what I would do. --The Photographer18:57, 3 April 2017 (UTC)[reply]
Just curious, how many German (or at least de-3) admins do you know, who are any familiar with QIC process. --A.Savin22:37, 3 April 2017 (UTC)[reply]
Since I'm involved in the"case", I agree that the block for 14 days is a bit severe and pledge for unblocking right now.btw I'll turn 70 this year !-) --Wolfgang Moroder (talk) 19:22, 7 April 2017 (UTC)[reply]
It's already QI. I took out the power lines which were in the image. Adamdaley 07:46, 8 April 2017 (UTC)
Good faith edits (and overwrite) to improve a work is not an issue. But if someone revert it contesting it as not an improvement, please upload it separately. Jee08:08, 8 April 2017 (UTC)[reply]
If a photo has been changed substantially, it should allways be uploaded separately. Retouching powerlines, traffic signs etc. could be evaluated as inadmissible manipulation and lead to decline. --Smial (talk) 08:16, 8 April 2017 (UTC)[reply]
Submission to Wikimania with the aim to improve things for the photographer's community here
Hello, if you agree that things need to get better for photographers, especially photographers worried about contributing HQ content, you make support this submission for Wikimania in Montreal where I'd like to address the issues I am concerned about and moved me to stop uploading. I hope you forgive me this spam wedge but as my purpose is to improve the life of HQ Commons photographers, I thought this may be interesting for you. Thank you. Poco211:20, 9 April 2017 (UTC)[reply]
Out of QIC topic: Panoramio bot uploads
For those, who are also dissatified with the recent uploads of image garbage by the panoramio bot: I opened a complaint about User:Shizhao and his Panoramio uploads at the Administrators' noticeboard. You are invited to share your personal experience and opinion about this user and his abuse of his bot operator rights. Thanks, --CEphoto, Uwe Aranas (talk) 10:20, 17 April 2017 (UTC)[reply]
Additional comment: It might be oot, but as it affects a lot of the experienced photographers, I decided to put it here. Sorry, mattbuck, to rollback your deletion. --CEphoto, Uwe Aranas (talk) 10:46, 17 April 2017 (UTC)[reply]
Just one question. There are some modifications of this user. It looks like a kind of vandalism. How to restore the comments? --XRaytalk13:23, 18 April 2017 (UTC)[reply]
Let's just ask him: David Adam Kess, hi! It looks like you have tried to review some of the photos at QIC. Unfortunately you are not familiar with how the code works there so it all turned out a bit of a mess. Could you please respond here so that we can sort all this out. Best, --cart-Talk13:41, 18 April 2017 (UTC)[reply]
As we have seen too many times before, it all started out as an honest attempt by a new QIC-participant to get some photos reviewed at QIC. The first edits were, as XRay says, just inexperience and were sorted out. Unfortunately, the user did not like the critique he got about his photos and instead turned his frustration on the reviewers in a way that can be perceived as vandalism. --cart-Talk10:06, 19 April 2017 (UTC)[reply]
Thank you for notifying, A.Savin. He was my age and it is a sad news. We lived in the same town but I never met him personally. Rolf was one of the silent but constantly delivering workers here. His legacy of ship photos is a real asset for Commons. --CEphoto, Uwe Aranas (talk) 05:51, 1 May 2017 (UTC)[reply]
QICBot failure on May 13
It appears that QICBot misfunctioned on May 13, stopping very early in the routine. If I am not mistaken, the bot cleared images to be promoted from the candidate list (Diff), but pretty much stopped after that. Obviously none of the images that should have been promoted that day have been tagged as QI.
So I wonder how we should proceed with that batch of images? Could QICBot be run again on the candidate list from May 13 or would it be easiest to add the QI tag ourselves? --DXR (talk) 06:11, 14 May 2017 (UTC)[reply]
I have informed Dschwen about the issue, but I don't know if I have reached him. He does not answer emails and his last edit is a month ago. Previously I have manually tagged all images, but this time I have reverted QICBot edition on QIC page. Still there are probably categorized pictures missing from this and this edits. It's a lot more to correct since part of them have been correctly moved. I think we need to have a set course of action for such cases. Do we manually tag all pictures? What about those from recently promoted page? -- Jakubhal07:34, 17 May 2017 (UTC)[reply]
Unassessed images
Can unassessed images (one that gets no comments at all or no supports or opposes, however it's defined) be re-nominated after a period of time? Perhaps whatever the rule is on this can be added to the Unassessed Images section near the top of the page. Personally, I'd rather get an oppose with a reason (so I can learn from it) rather than no reviews at all. PumpkinSkytalk16:26, 17 May 2017 (UTC)[reply]
@PumpkinSky: Yes, I've seen several unassessed photos getting re-nominated and no one has objected. Right now there are many new nominaters who do not/can not review photos and not enough users who do more than just the required number. I try to go through the backlog a bit, but I also have a full-time job to do. Hopefully some more users can do a bit of extra work. I agree with you that a 'decline' is better than no review at all. --cart-Talk16:47, 17 May 2017 (UTC)[reply]
Thank you cart. I think a short wait period of say 3-5 days after the first nom run ended would be okay. I don't review more than I do because I don't feel more qualified. So I only review obvious pass/fail. I still often don't see what others say is wrong with a photo. PumpkinSkytalk17:09, 17 May 2017 (UTC)[reply]
Since Peulle promoted two photos at QIC CR (1 and 2) after several users had voted in favor for promotion even though the issue of downsampling had not been resolved, are we now to conclude that downsampling is allowed at QIC? The nominator, left a message on my talk page. It all seems to be a bit of a mess. Thoughts? --cart-Talk12:54, 22 May 2017 (UTC)[reply]
Hello W.carter, the pictures that I proposed for QI are from WLE Brasil 2017. In general 92% of the participants are new users, not involved on the Wikimedia Commons or Wikimedia projects.So, they dont know the recomendations for QI and etc. We are using the WLE as a project to engage new contributors and trying to explain more details about the wikimedia dinamics during the contests. I'm requesting the top contributors to upload better pictures with more informations and in large sizes asap. Rodrigo Padula (talk) 13:14, 22 May 2017 (UTC)[reply]
I'm sure it will all be better from now on with the photos you nominate, Rodrigo. The new problem here is not so much with your photos, as with that several users seemed willing to accept and promote downsampled photos against the current policy. If a policy is no longer deemed important enough to follow, should we keep it or alter it? --cart-Talk13:40, 22 May 2017 (UTC)[reply]
I think this topic had discussed several times here and it is very difficult to arrive to a simple answer as we need different standards for different types of works. In this particular case, it is a macro from a 12MP camera which may be downsampled to 50% width and height if not cropped. Another photofrom Charles is also only 4MP (taken from a 18MP camera). I'm one of the macro photographer who do not downsample that much. I upload full size pictures upto 250 ISO and keep 80% in ISO 320/400. But I downsampled this to 50% due to the poor light conditions. Technically that picture can be taken from a distance and crop or approach closely and later downsample as did here. The advantage in downaample is same pixel as in crop but with less noise.
So my understanding on the guidelines is, "Images should not be downsampled (sized down) in order to appear of better quality. But images can be downsampled if the high resolution images contains nothing more." I think this rule is to discourage people who prefer to keep their high resolution files for sales and using Commons as a platform for their publicity. We had such issues earlier but not seeing nowadays. Further, CC had added in their FAQ that maintaining different licenses for different resolution files may not possible. Jee13:57, 22 May 2017 (UTC)[reply]
I must misunderstand what you mean by downsampled. Jee implies above that I downsample. I crop most of my images, but I ALWAYS upload the maximum size I have. With macro photography (Jee will know this) you sometimes have to back off from the subject to get greater depth of field - then of course you have to crop. So I can see no reason for what I understand is downsampling - saving a 6000 x 4000 image at 3000 x 2000 to make it look better. Charles (talk) 14:51, 22 May 2017 (UTC)[reply]
No Charles, I didn't mean it. What I meant is the end-result will be same, that is, same number of pixels in a particular DOF irrespective of whether down-sampled or cropped. And a downsampled image has an advantage in noise level. This is somewhat like using multiple frames combined together to reduce noise. And I see no reason to discourage the photographer from approaching closer to the subject to fill the frame and later downsample to achieve the same result of a crop with less noise. Jee15:54, 22 May 2017 (UTC)[reply]
I don't think "no downsampling" is a hard and fast rule. But I absolutely will judge low-res photos more harshly than high-res photos, at both QIC and FPC. -- King of♥♦♣ ♠ 04:08, 24 May 2017 (UTC)[reply]
I always upload my images at full resolution, that could be as much as 50 MP. It is sad, but a fact, that some reviewers don't care whether they look at a 50 MP or at a 10 MP image and will review them following the same pattern. It is indeed not a piece of cake to get a sharp image with a 50 MP than it was earlier with 21 MP and that is why I do sometimes (especially when I get here feedback about it) downsample them to make them sharp. I just think that it is fair and don't see much of an issue for those consuming those pictures. Poco207:07, 24 May 2017 (UTC)[reply]
This should not be a criterium for a mere quality image (this is not the featured picture assessment!). The requirement ist 2 megapixels, this should be enough (otherwise change it). If a camera does 20 bad megapixels which make a perfect image downsampled to 10 this should not make a difference. --Ailura (talk) 09:36, 24 May 2017 (UTC)[reply]
I wish we had a way to express the "2 megapixels" issue without people thinking it is a sufficient requirement to pass QI/FP on resolution grounds. Is is a barely acceptable size permitted for a few special-case images, and is long-overdue being increased. I also think we should reword the downsampling rule because it sounds too absolute and hasn't ever been enforced that strictly (if it were, Diliff wouldn't have many FPs, and nobody can accuse him of uploading low-resolution images). I think Jee's second clause is along the right lines, though there will always be some who argue (rightly) that information is always lost. But Jee you are downsampling is "to appear of better quality". We do it because in some circumstances the 100% size is too soft or too noisy and not really benefiting anyone. What behaviours we want to discourage are
downsampling to retain better images for personal sale and give low-resolution images a free licence
downsampling because you think Wikipedia only needs a thumbnail and small images are still useful for many web pages
downsampling a perfectly fine image in order to avoid pixel peeping comments about noise, CA or sharpness, which are only apparent when you enlarge the image to be several metres wide.
As KoH suggests, if an image is only 5MP I'd expect it to be razor sharp and noise-free, which I wouldn't expect at 24MP say. As Charles notes, cropping can quite rapidly decrease the megapixels, so small images aren't always downsized. If you apply much vertical perspective correct (common with stitched images) then parts of the image get magnified (always the top left corner that the browser takes you to) so some downsampling is reasonable then.
QI is unusual in that the photographer is always on Commons, so we may know their camera, and can ask them if the image could be uploaded in higher resolution. FP doesn't have that option for images transferred from other sites. There, we just have to accept what we get, and judge it for what it is. In some ways, I think that is a bit unfair on the Commons folk. While on FP we can judge the frog photo against its peers and determine if a 3MP image is among the finest here (no), but QI can't do that. It comes back to what QI is for, which I'm not clear about really. I'd say we should be aiming more at 8 megapixels for many subjects, and perhaps 5 for wildlife, but these are minimums and we need to find a way to express that without making it a target people aim for. A 4K screen is 8MP in 16:9, and a 5K screen is nearly 15MP in 16:9. -- Colin (talk) 10:07, 24 May 2017 (UTC)[reply]
Colin's comment sounds logical to me, the rules should be flexible. Could be carefully phrased in the guidelines somehow, since reviewing photos always is a judgement call and such leeway should not be abused. --cart-Talk10:33, 24 May 2017 (UTC)[reply]
A QI candidate should always be judged based on its quality in the uploaded size, i.e. when you shoot a 50MP photo and upload only a 15MP version of it because you are not satisfied with its quality at full res, it is perfectly OK with me and when the picture indeed looks good at 15MP it deserves a promotion like any other; but when you decide to upload the full 50MP version then and to nominate it, don't wonder about eventual opposes, and refrain from comments like "It will still look fine at 30%". With that said, downsampling is not per se evil and should not be generally discouraged or even disallowed; but I would strongly discourage downsampling when it is done in order to reserve the high-res version for non-free licensing. --A.Savin11:10, 24 May 2017 (UTC)[reply]
I disagree with A.Savin. We are a "repository" of educational media files, not a publisher of JPGs at a given size. The media wiki software lets people publish an image at any (reasonable) resolution they like. The fact that it lets you download the original JPG and view it at 100dpi is not relevant to QI imo. 100dpi is a stupid resolution to judge a 50MP image negatively and anyone who does so is IMO harmful to the project and discourages photographers from contributing here. We are hitting a problem that is solely due to the very limited ways Commons lets you view an image easily. It's either a tiny thumb (QI listings), predefined size (such as the nominations at FP), screen-filling, or 100%, whatever that might be depending on the source resolution and your monitor dpi. It would be wonderful if we had a "QI review" tool/page that made it simple to examine images at reasonable resolutions. Easy buttons to let you say "How does this look at 15MP"? Really, if a 50MP image looks great at 15MP, then view it at 15MP when judging QI and FP. It's embarrassing that this still needs saying, and that we keep discouraging folk from uploading full size images this way. -- Colin (talk) 11:39, 24 May 2017 (UTC)[reply]
I don't think the issue of people reserving the best resolution für non-free licensing can be solved here. If a 50MP image doesn't look good at 50MP, why upload it like that? --Ailura (talk) 12:24, 24 May 2017 (UTC)[reply]
Ailura, can I ask if you own a 50MP monitor? If not, how can you say whether the image looks good at 50MP 100% size? All you are seeing is a tiny portion, much like you would if you took a magnifying glass to a print. At 100DPI, a 50MP portrait image is 2.2 metres tall, taller than most humans. If that was a poster, you'd stand well back from it, but most people sit about 50cm from their monitors. When photos are printed, the resolution is higher and the printing process downplays some issues like noise, so really, most complaints about a large image at 100%, is looking at a ridiculously magnified image, and making us all look rather foolish. Modern high-dpi monitors are better and I'm sure that as more people get them, the complaints about noise/ca/etc at 100% will reduce. Why upload a 50MP image at all then, if people can't see the whole thing on a monitor -- because then people can print it at 300DPI or can crop it and still have a high resolution copy. High quality printing has always demanded high quality high resolution photos. -- Colin (talk) 21:10, 24 May 2017 (UTC)[reply]
I think uploading only good pictures at 100% would improve quality. If you need a house sized poster, there is no big difference between blowing up a downsized image or using a picture developed bigger than it should. --Ailura (talk) 20:12, 25 May 2017 (UTC)[reply]
Ralf Roleček, I don't understand your comment. "educational" is a requirement for Commons scope, even if very loosely interpreted. The image doesn't need to be used educationally, though. -- Colin (talk) 21:10, 24 May 2017 (UTC)[reply]
This "educational" is a relict of old time. It's from the time, who commons only was the source of Wikiepedias. Today we are a source of free images for all, via instantcommons all Mediawikis can use the files of Commons. All can use our files for all purposes. "Educational" is today no longer current. --Ralf Roleček21:21, 24 May 2017 (UTC)[reply]
-Ralf Roleček, "Must be realistically useful for an educational purpose" is still policy. I think you are confusing "encyclopaedic" with "educational", or have your own opinion that isn't supported by consensus/policy. -- Colin (talk) 22:03, 24 May 2017 (UTC)[reply]
The reason why we have one QI section is to improve the photographers quality on theirs pictures and in consequence, to improve the quality of the photos uploaded to Wikimedia Commons anything else is just a deviation from the main objective. However, Rodrigo Padula are using downsampling to get a hight QI images counter for their photo events and in consequence apply for an additional grant. --The Photographer13:16, 24 May 2017 (UTC)[reply]
User:The Photographer One thing has nothing to do with the other. We do not rely on QI numbers to get another grant, while WLE, WLM and other contests generate positive results for commons and Wikipédia, we will continue to carry out these contests here in Brasil like all the other countries. Your allegations do not contribute at all to this discussion. The photographer responsible for the images has been notified to send new versions at higher resolution, he informed that he have all versions in RAW and that the files upload was the JPGS automatically generated. Images promotion and inclusion in wikipedia articles are part of our job as contests organizers. We will try to improve the contest's recomendations to take all the QI in consideration to improve the results during the next editions of our contests. Rodrigo Padula (talk) 19:49, 25 May 2017 (UTC)[reply]
We will regret in the future this practice because many things that seemingly can not be corrected (like extreme noise) in the future could be corrected with a simple filter, however, as the image was downsized, the correction will be limited. Downsizing is allways less image information, the "more quality" how result of a downsizing practice is only a illusion. The user can upload the image size whatever he want, however, on QI section we should not support this practice because we should encourage images correction with adequate practice, downsizing is a destructive practice. --The Photographer13:31, 24 May 2017 (UTC)[reply]
I kind of disagree because IMO many options for future improvement are already lost when RAWs are converted to JPGs. Also it's certainly reasonable to expect a certain learning curve from photographers who participate in QI, but sooner or later you'll get to the point where further improvements mean spending a lot of money. A compact camera or a crop DSLR simply cannot deliver the same quality as a full-frame DSLR, especially with difficult light situations. Denoising and downscaling can fix - oder better hide - some of these limitations. --Magnus (talk) 13:43, 24 May 2017 (UTC)[reply]
I underestand what you are telling, I remember how It was very frustrating for me to get an FP with my humble compact camera, however, this was possible. I say that it was frustrating because my photos were small and of poor quality due to the censor, I felt enormous impotence because it is more difficult to compete with images downsized where any mistake is hidden. --The Photographer14:21, 24 May 2017 (UTC)[reply]
I agree with Colin that the wording ""Images should not be downsampled (sized down) in order to appear of better quality." is confusing. For example, it is not good if an image is downsampled to hide the focus error (front or back focused). But I don't think downsampling is bad if we're only removing the junk pixels that are recorded due to the poor lights or any similar reason. Jee14:30, 24 May 2017 (UTC)[reply]
Is downsampling at QIC ok now? - Part 2
I think the one thing we all will agree about is that the practice of downsampling is mainly a result of the extreme pixel-peeping we're all doing here at QIC. That's something we should talk about. --Code (talk) 08:31, 25 May 2017 (UTC)[reply]
Hmm. On the basic laptop I'm viewing just now, I don't see any CA. Even if it was pixel-level magenta, in a 36MP image that's not a deal-breaker imo. Since CA can usually be fixed without harming quality (unlike over-strong NR say) then I haven't a problem with people making suggestions-for-improvement to other Commons users' images. But to make it an issue for QI promotion, I think the CA has to be pretty bad. -- Colin (talk) 14:32, 25 May 2017 (UTC)[reply]
Code, agree again with you and will follow suit. Sorry for being opportunistic but I cannot see anything on my laptop, sorry, the thread was just too tempting. Poco220:07, 25 May 2017 (UTC)[reply]
I agree too. For a long time the filesize was a problem to. When I started uploading an downsampled image takes 10 up to 15 minutes for upload. Today I replaced a lot of images by the original resolution. --XRaytalk16:37, 25 May 2017 (UTC)[reply]
Comment Regarding the link provided by Diego, I agree with Code, removing these purple fringes will be a plus and an improvement for the image. And more generally, IMO, purple fringes must be removed unless that an even worse contour is created when removed (remedy worse than evil), there is an easy tool for that in Lightroom. Purple fringing is a defect. Regarding downsampling, I fully agree with the use of reasonable downsampling in order to improve an image, exemple of a promoted image that was in CR on the way to be declined. Christian Ferrer(talk)14:47, 25 May 2017 (UTC)[reply]
Christian, I can view an image at any downsized resolution I like. For example 2000px wide. So your source JPG at 4587px wide can be displayed at 3000px wide using MediaWiki software. How have you improved the project simply by overwriting the 4587px wide image with a 3000px image that is now permanently downsampled? This appears to being done so that reviewers can easily display an image at a resolution they are happy with and can promote accoring to our pixel-peeping image guidelines, rather than being done so that Commons respository of images is improved. I don't have a problem with a photographer choosing to publish their image at a resolution where it looks great, provided the motivation is just that they are personally unhappy with the "flaws" visible at 100%. Where it becomes harmful is if a photographer feels obliged to downsize (or gets their image downsized by someone else) in order just to pass QI. Once again I wonder what the purpose of QI is. Sure, we can all find pixel-peeping "room for improvement" in each other images and make out that purple fringing is a crime against one's eyesight, but that doesn't seem to be a reason to decline an image at QI. Plenty images at QI that are a crime against one's eyesight for reasons other than pixel level faults. -- Colin (talk) 16:41, 25 May 2017 (UTC)[reply]
My version is far not "only a downsampling", this is as I already did for several other users, an edition, a work, of minimum 1 hour, maybe more, and likely with several softwares. So yes, if I uploaded it in 2014, with the agreement of the author, it was because I thought it was really an improvement, and the votes and comments there gave me and give me right. Secondly your link is a downsampled version of my "edited" version, that has been already improved after more than 1 hour of my work. And my downsampling is far less big than yours. Christian Ferrer(talk)16:57, 25 May 2017 (UTC)[reply]
Furthermore I never use MediaWiki software to display an image at a different resolution than the resolution of the choice of the author. This is also why I naturally do not think about this kind of practice, it simply never comes to my mind, and it is highly possible that the vast majority of users and visitors do as I do. Christian Ferrer(talk)17:03, 25 May 2017 (UTC)[reply]
Christian the link I gave was just an example of how to specify a size. I can't give a link to the original image because you overwrote it with your edit. So we can't compare the two, plus there is the issue of MediaWiki applying compression along with its thumbnailer. You say the "vast majority of users and visitors". But Commons purpose is not a publisher of images, and its display software for images is rudimentary compared to other photo sites. It is a repository. If you treat it as such, and QI is meant to be determining the quality images in that repository, then "user display resolution" is not a factor. Really, QI is being driven by the very limited mediawiki software that positively encourages reviewers to consider 100% size. Photographers should be permitted to upload at "out of camera" resolutions if they choose, and not be expected to downsample just to make the folk here happy. That you never consider to view it at another resolution is, I'm sorry to say, your choice to limit how you review and assess digital photography. It's a very poor choice imo, and leads to harmful practices at QI/FP. Elsewhere on the internet, people take great pictures and enjoy them. Only at QI/FP do folk worry about great pixels. -- Colin (talk) 17:33, 25 May 2017 (UTC)[reply]
I can follow you and I judge also, of course, the image as a whole composition, with the image open and fully visible in a window : when the image is open at full resolution I click on the "-" and the image take the size of the window. One does not prevent the other, for me, to be a QI, an image has also to have a visual quality visible to virtually all resolutions. An image that does not pass QI is not necessarily a bad image. I myself don't understand a lot of decisions in QI, exemple I still don't understand why this image was not successfull. Christian Ferrer(talk)05:04, 26 May 2017 (UTC)[reply]
Christian why do you think it important/essential that a QI must be visibly perfect at "virtual all resolutions". If it is good at 4MP, 8Mp, 12MP but you can spot a little CA at 24MP and a little noise at 36MP, how does it help Commons to simply remove the last two options by uploading a 12MP image for example. If you regard Commons as a repository, then any image is available at any resolution one may choose. Removing choices does not make the image better. And most uses for our image will be for smaller-than-screen-size. We still have the woefully small 2MP minimum size requirement. If instead we judged QI at 8MP say (and 4 or 5MP for wildlife), then all images would be judged equally, and those with higher resolutions than 8MP would simply be a bonus for those who wish to view at the higher resolution. It would help prevent some of the nit-picking that goes on here. I continue to ask what QI is for. In my opinion it should be for helping to select the images that are high enough quality to be widely useful to our users, and so help separate them from the rubbish and the weak images. Instead we have this over-fussiness with technical achievement wrt noise/CA/sharpness. I don't really think QI should be assessing images at 24, 36 or 50 MP at all. It is fair to expect better at FP because we are looking for "finest" and many times the standard is high, but I think QI has gone wrong for trying to set its technical standard to equal FP without the wow. -- Colin (talk) 08:56, 26 May 2017 (UTC)[reply]
in this situation is it difficult to take sharp photos
Je nach Vibrator, äh, Fluggerät kann u.U. sogar 1/2000s noch verwischt sein. Und ein Stativ fällt naturgemäß halt als Verbesserer aus ;-) -- Smial (talk) 13:25, 29 May 2017 (UTC)[reply]
Da beißt sich, wie so oft, die Katze in den Schwanz. Licht gibt es nicht unbegrenzt, und man hat die Wahl zwischen ISO1600 (Detailverlust, Rauschen, Dynamikverlust) oder eben Verwackelgefahr. Ich persönlich würde ebenfalls eher 1/1000 oder kürzer erzielen wollen, sch*** was auf Rauschen, andererseits hat Ralf eindeutig ein Fisheye verwendet, da sagt die Erfahrung, damit kann man 1/400s eigentlich gar nicht verwackeln. Wie man sieht: Geht offensichtlich doch! ;-) --Smial (talk) 13:59, 29 May 2017 (UTC)[reply]
Zwischen iso200 und Rauschen im Walde ist aber auch bei der d300 noch etwas Spielraum. Aber bei Weitwinkel und Landschaft fehlt mir wohl auch etwas der Erfahrungswert.--Ailura (talk) 14:18, 29 May 2017 (UTC)[reply]
Ich war halt zum ersten Mal mit diesem Maschinchen unterwegs. Statt Objektivwechsel habe ich 3 Kameras genommen, das hat sich bewährt. Bei dem Sturm an Bord hätte ich kein Objektiv gewechselt. Man kann sich aber kaum wie in der Do-27 etwas zurückziehen und dem Wind entweichen. Lehrgeld, beim nächsten Mal werde ich auf ISO400 und 800 gehen. Wenn der Motor gedrosselt wird, dann wird es auch bedeutend ruhiger, das werde ich beim nächsten Mal auch anders machen. Angenehm ist die Bewegungsfreiheit und die sehr schnellen Reaktionen des kleinen Vogels. Man lernt nie aus. Worum es hier aber eigentlich geht: Ich habe herunterskaliert, um scharfe Bilder zu veröffentlichen, zumindest nahezu scharfe. Nicht, um die großen zu verkaufen oder um hier jemanden zu ärgern. Runterskalieren kann einfach auch mal Gründe haben. Und die Bilder aus der eigentlich sehr viel moderneren D5300 mit 24 Mega sind deutlich schlechter als die aus D300 und D300s mit 12 Mega. --Ralf Roleček20:00, 29 May 2017 (UTC)[reply]
Wenn beim "Projekt Fotoflüge" eh alles über 3-4MP hinaus verwackelt ist, könnte Wikimedia Deutschland/Österreich stattdessen darüber nachdenken, auf Drohnen zu setzen. --A.Savin20:30, 29 May 2017 (UTC)[reply]
AFAIK ist das rechtlich und versicherungstechnisch eine ziemliche Katastrophe. Und meine Fotoflug-Bilder waren damals IIRC nur vernebelt, nicht verwackelt, aber lange her. Abgesehen davon, 4MP ought to be enough for everybody. --Ailura (talk) 05:47, 30 May 2017 (UTC)[reply]
Das war ein rein privater Fotoflug von mir selbst bezahlt, ohne jegliche fremde Finanzierung. Kostet 150 Euro die Stunde und ist damit viel billiger als mit einem Aufklärungsflugzeug. WM-AT bezahlt mein Photoshop, deshalb sind sie genannt. Ich wollte dich sowieso mal ansprechen, weil ich deine Drohnenflüge bzw. die Bilder davon beeindruckend finde. Wo ist ein geeigneter Ort dafür? --Ralf Roleček20:57, 29 May 2017 (UTC)[reply]
Seit den Gesetzesänderungen bezüglich Fotodrohnen in Deutschland ist praktisch überall dort das Fliegen verboten, wo man es für Wikipediazwecke wirklich gut gebrauchen könnte. --Smial (talk) 09:12, 30 May 2017 (UTC)[reply]
In Deutschland sind trotz der Restriktionen aus der Drohnenverordnung durchaus noch Einsatzmöglichkeiten mit Drohnen vorhanden, selbst in den Städten. Zudem könnte theoretisch für z.B. Flüge über 100m eine Aufstiegserlaubnis beantragt werden. Für solche Sachen sollte doch eigentlich der Verein da sein. In Österreich ist alles wesentlich komplizierter und braucht teure Genehmigungen, aber siehe vorheriger Satz. Sorry, it's offtopic about community support of aerial photography and what the chapters could do. --A.Savin18:55, 30 May 2017 (UTC)[reply]
This text summarizes the main arguments that I have propagated since the release of the D800 and corresponding high-resolution APS-C cameras, and for which I have been attacked quite violently in the past. Thanks for the link. --Smial (talk) 12:12, 29 May 2017 (UTC)[reply]
XRay, thanks. I read it with Google translate. How familiar is this: "A FullHD monitor can represent a resolution of 2 megapixels. So if you display a 24 megapixel image as a full image on this monitor, this is reduced to 1/12 of its area. If you now zoomed to 1: 1, you see 2 megapixels from this picture, that is 1/12 of the overall picture. A pixel of the captured image is now projected onto exactly one pixel of the monitor, in the 1: 1 view. A monitor with 60cm width and 35cm height now represents 2 megapixels. In order to display the complete 24 megapixels in 1: 1, your monitor would have to be nearly 2 meters wide with equal pixels. And you would still sit at a distance of 50cm.". His maths don't account for the 3:2 camera sensor vs 16:9 monitor screen, but still, same argument. -- Colin (talk) 15:04, 29 May 2017 (UTC)[reply]
If everybody can do it for their own images, that's enough. After two days I'll check any images remain and will take care of them. ;) Jee07:16, 28 May 2017 (UTC)[reply]
No worries. It was an experiment with COM:VFC and it worked. I didn't think you've 11 images in that set! I oversighted two unassessed images too that A.Savin corrected. Jee10:28, 28 May 2017 (UTC)[reply]
Failure again on 2nd of June. Luckily the archive page has been already build and I have tagged all promoted images with template. I do not think it is enough to mark just own pictures. There may be people who are new on QIC page, and they do not follow discussions here. I would check archive pages from previous failures to see if all promoted files have been tagged. Unfortunately I cannot do that right now - no time, maybe in the evening. -- Jakubhal06:24, 2 June 2017 (UTC)[reply]
There is now also the start of a list of If you know some abbr that should be in that list, please just add it. It may result in a new section below, which is great. We are also in need of links to good online external sites with photography terms and tutorials. If you know any, please add them to External links. Don't worry about knowing how to spell or formatting, we'll fix that for you. --cart-Talk 10:17, 22 March 2017 (UTC)
![[1]](https://commons.wikimedia.org/wiki/File:Bare-faced_curassow_(Crax_fasciolata)_female_head.JPG){kind=link}
Done. --A.Savin 13:42, 2 April 2017 (UTC)_035.JPG){kind=link}
_female_underside.jpg){kind=link}
Comment Regarding the link provided by Diego, I agree with Code, removing these purple fringes will be a plus and an improvement for the image. And more generally, IMO, purple fringes must be removed unless that an even worse contour is created when removed (remedy worse than evil), there is an easy tool for that in Lightroom. Purple fringing is a defect. 
