Commons talk:Quality images candidates/Archive/2007-02

Urgent need for concensus on Map illustrations

We urgently need some consensus on what are the requirements of maps to be accepted as Quality Images. We have one reviewer failing every map as not being an SVG image and another eloquently telling us why a scalable format is wrong for a map.

1. Is there an SVG option suitable for specifying what the 'proper size' is for the image so that it can be printed/displayed appropriately
2. Or is there info that can be printed on the SVG image to ensure scaled versions are interpreted appropriately
3. Can non SVG maps not just be evaluated as the images that they are, there is actually no QI guideline demanding they be SVG

I am tempted to put the whole lot together in CR until this is sorted :-) --Tony Wills 10:58, 24 June 2007 (UTC)

• Thank you for bringing this issue to discussion. The problem is much harder to solve than it appears. First of all, because we don't have here the means, or the knowledge, to evaluate the intrinsic quality of any map (that is a rather complex process, believe me). All we can do is look at the form and make a decision based on some easy to evaluate parameters. Secondly, it makes little sense to promote or not promote a map only based on aesthetycal considerations, disregarding its main purpose: to depict in a clear and accurate way useful geographic information. Maybe the wisest thing to do is not to accept maps in FPC or QIC (except historical ones, which are evaluated differently). - Alvesgaspar 11:23, 24 June 2007 (UTC)

• Valid point. Besides the factual accuracy which one ought but cannot check by viewing only, I tend to think of a map as a style or design or template, that also could be applied to a different location. But then there are good and bad designs, and good and bad implementations for a given locality, and I think we should not bar ourselves to give some guidance.
• Now regarding bitmap versus vector, I'd say that features like relief shading can be justified to be bitmap. However, I strongly feel that features that derive from a vector description, be it contour lines, roads, rivers, areas and also lettering should not be subsequently rasterized. Regarding the proper printout size of a SVG image, there are subtleties in that a 24pt font scaled 50% should look somewhat different from a 12 pt font, but in the context here I largely consider that a red herring. So from my side SVG please (or any other open vector graphics format). -- Klaus with K 11:54, 24 June 2007 (UTC)

• Let me paste here what I have written in CR about svg and scalable maps:
  
• SVG shouldn't generally be used in cartography because a map has a determined scale, planimetric accuracy and detail associated with it. When we enlarge a map beyond its nominal scale we are implicitly giving a wrong information about its spatial accuracy. Also, the density of the information, as well as the size of the symbols and the letering, will be inadequate to the new size. When we reduce it, we will have a cluttered representation. This problem is solved in the web maps (like Google Maps) by a mechanism we might call "automatic generalization", which consists in adapting the size of the symbols and the information density to the actual scale. In my opinion, every map should have associated with it, not only a graphical scale, but also a numerical scale (for example 1:10000), so that its normal size can be deduced from it. I have been calling the attention of the users for this problem for some time, but the fashionable "Saint Svg" still seems to prevail over reason. A map is not just a beautiful drawing that we can reduce or enlarge at will. All maps always have a nominal scale associated with it, which is closely related to the spatial and thematic accuracy of the information depicted. In paper maps, the length of 0,25mm (which is the typical thickness of a thin line) is normally taken as a reference for planimetric accuracy. In a map with a scale of 1:50 000, this means that the horizontal accuracy of the survey was, at least, 0.25 x 50 000 = 12 500mm = 12.5 m. In other words, it is guaranteed that the error in the position of all objects depicted in the map is less than 12,5m. When we enlarge that map by a factor of 10, keeping the line width of 0,25mm, we are implicitly assuming that the horizontal accuracy is 1,25m, which is wrong. And when the nominal scale of the map is not even stated, the map becomes useless for anything other than trivial applications. That is precisely the problem with the svg format. - Alvesgaspar 12:11, 24 June 2007 (UTC)

Alvegaspar, point on map accuracy taken. With a bitmap format one does not know either, how many (sub-)pixels accuracy one gets. Yes I know some bitmap files include meta information, typical values 72dpi or 180dpi or 300dpi, but I hesitate to use this information to convert the value of 0,25mm into a pixel accuracy value. I think the main problem is not SVG or bitmap, it is a contributed map's accuracy can be anywhere in between completely dreamt up and accurate to the micron, and people may not know about the usual accuracy expectancy . Sorry I read your comment at the original location only after writing my reply above. -- Klaus with K 13:17, 24 June 2007 (UTC)

P.S. You are talking about proper maps. But look around how much artistic licence one can find in, let's call it map sketches, in the real world. And I wonder how much deviation one may find in overlaying two road maps of the same scale... -- Klaus with K 13:17, 24 June 2007 (UTC)

Sorry, but the often repeated "SVG is bad because it gives a wrong sense of accuracy" statement is essentialy a non-argument. There are better ways to state the map accuracy than jagged edges. An accuracy should always be mentioned either on the map or the description. Many bitmap maps are less accurate (and more schematic) than their pixel resolution. --Dschwen 19:01, 24 June 2007 (UTC)

• For your information: I will not be able to participate in this discussion or QI in general for the next seven days as I have important exams on Thursday (June 28) and Friday (June 29). I plan on getting back here on Sunday, July 1. Sorry for the absence. After that, I will gladly continue to participate in the discussions here and on QI/QIC. --Florian Prischl 12:28, 24 June 2007 (UTC)

• • i already state before that all digital maps should have a graphic scale, in oposite to a word statement of scale or a ratio statement of scale. since this two have originaly designed for paper maps and are useless in a computer world, weather bitmap or vector. each screen resizes an image when it is shown up, also when it is suposely 100% zoom, this comes from the diferent density of the screen and not from the file, or program in wich the image is displayed. want to have fun with it? then open an image in 10 diferent screens at 100% zoom and measure it. in some cases there are milimeters of diference (or more in non falt screens). and the bigger the object the more complicated it becomes becouse zooming in makes the whole thing even more dificult. Listen i am not cartograf but i am grafic designer and my work is with images, screens and printing. you can trust me in this one. SVG is an ideal format for maps and diagrams ( proven that it is done right in first place) what needs to be worked on is that the scale that is used works together with the computer. and the only solution to hand is the grafic scale.-LadyofHats 15:17, 24 June 2007 (UTC)

...and due to the fallacies of real life hardware one should consider giving both a horizontal and a vertical scale. -- Klaus with K 20:01, 24 June 2007 (UTC)

• • •  Comment - No need, in ALL maps scale varies with location. Also, and except on conformal projections, scale also varies with direction. The so-called "scale of a map" is only valid for certain locations or lines (the "standard lines or points). However, in large scale maps covering small regions the scale can be considered as constant for common, non-geodetic, purposes ;-) - Alvesgaspar 11:06, 25 June 2007 (UTC)
• I recognise User:Alvesgaspar is much more an expert in this area than me and his arguments are persausive, but then I recognise that in other areas we are not as stringent. We don't decline to review anatomical diagrams because we are not in a position to evaluate whether they are strictly correct and whether a surgeon could use them to do surgery. We are not in the business of providing certified maps, but only evaluating the workmanship and whether they are 'correct' as far as our cursory inspection will allow. Just as I am not an expert on classifying plants and animals, I will at least check that the animal looks like the rest of its declared species. All wikipedia and wikimedia content has to be accepted as as-is-where-is, and QI is not a certification of correctness.
• I propose that all maps, apart from images that are illustrations of maps in other media (eg printed historical ones), are required to have a graphical scale (whether they're bitmap or vector), and the description should include the date when the map was applicable, and the source from which the map detail was gathered, and the use for which the map was made/or a description of what it proports to show.

I expect that list probably means all maps here will fail ;-) --Tony Wills 11:52, 25 June 2007 (UTC)

I second the propousal-LadyofHats 12:39, 25 June 2007 (UTC)

Language versions and orphans

i just noticed that Image:Joyopis.svg is already a quality image, and another version of it without numbers is trying to become selected (even when i think both have some quality problems). the wuastion is, is it really necesary to see each diferent version of each diagram? i mean some of my diagrams have like 6 or 8 translations and i do not see any sence in trying to place all and each one of them here. one thing is to have diferent sunsets and all. but i do not think this should be permited. i would also like to sugest that one criteria to be acepted is that the image have at least one category ( wich is not refering to copyright status)-LadyofHats 12:53, 25 June 2007 (UTC)

Unsharp image corners

It is not the first time that I see photos that are technically ok except for the corners where they become somewhat unsharp and possibly show some colour fringing. Short of using this camera and not another one, the photographer has done everything else right and produced a lovely and useful photo. As Quality Images are about encouraging photographers, I feel while pointing these problems out we should still promote such images if these camera-technical shortcomaings are minor. How to define minor? -- Klaus with K 19:29, 25 June 2007 (UTC)

It seems to me that taking photographs that would qualify as QI is impossible with my Canon S3 IS. At least a rough guide stating the line with some idea as to the minimum equipment needed for making it to QI could be useful. As it stands I'll refrain from nominating my images until I get a new camera. –Dilaudid 11:55, 28 June 2007 (UTC)

Yes, better cameras and better lenses can help, but you might be surprised what's possible. Regards, Ben Aveling 02:55, 29 June 2007 (UTC)

May I qualify my remark. Lense design is an optimisation process, and I understand that rather often manufacturers accept a lesser image quality in the corners because it can reduce costs - or at same cost, bring an improvement in a different aspect they consider more important. -- Klaus with K 08:42, 29 June 2007 (UTC)

I love the Canon S3 IS. It's one of my alternate cameras and is a pleasure to use. I even have a featured picture from that camera. There is no reason you couldn't have a ton of QIs from that camera. -- Ram-Man 12:19, 29 June 2007 (UTC)

Can we combine "nominations" and "consensual review"?

I just took a long look at this page, trying to find ways to make it a more approachable size. It occurred to me: why do we separate the nominations section from the consensual review section? They're there for the same reason. If more than a couple of people comment on a nomination then it has to be moved to the other section anyway; why not make the entire page "consensual review" and remove the shaky dichotomy? I would personally find the page much nicer-looking that way, and as a bonus we'd make it a lot shorter (with less text needed in the intro). Thoughts? --bdesham 00:17, 28 June 2007 (UTC)

The intent is that a majority of images should be promoted or rejected quickly, without the need for moving to discussion. I know it doesn't look that way at a glance, but that's because clearly good and clearly not good enough pictures get processed fairly quickly, 2 or 3 days usually. The ones that hang around are the borderline cases. Moving everything to a consensual review model would move away from the "1 vote is normally enough" model. I agree that it isn't a clear system for the newcomer, but, for me, I think it works. Regards, Ben Aveling 11:12, 28 June 2007 (UTC)

For me it works fine, too --Simonizer 12:11, 28 June 2007 (UTC)

i do like the actual system but i would like to coment something, on the template for the image thumbnails, i would rather like them to be wider becouse when one writes a lot of text then it becomes sooo long... -LadyofHats 20:31, 28 June 2007 (UTC)

• I, too, think that that the current systems works satisfactorily. I think LadyofHats's proposal to widen the preview boxes should be examined, because she is right about the text. I am afraid this would require some change in the formatting of galleries in general, because that is what is used for the previews, right? --Florian Prischl 21:24, 1 July 2007 (UTC)

Strange problem with template display

I'm seeing lots of image descriptions (the bit done by template), just showing up as small coloured square boxes. As far as I can see there's nothing wrong with the descriptions - maybe a server problem? Anyone any ideas? --Tony Wills 12:13, 10 July 2007 (UTC)

• dunno, i sometimes see only a normal gallery and sometimes the descriptions, dont know weather my changes are taken in count or not-LadyofHats 21:45, 12 July 2007 (UTC)

Gallery layout (scripting issue) was changed recently ([1]) to fill the whole screen. Maybe that's causing it? I experience similar problems as Tony. Lycaon 21:53, 12 July 2007 (UTC)

Thanks for that, yes it looks fine with javascript turned off. I've left a query on the VP page. --Tony Wills 13:26, 13 July 2007 (UTC)

Don't turn off Javascript, use Special:Preferences to disable the galley resizing! I'll have a look at why it screws up QI.... --Dschwen 07:56, 14 July 2007 (UTC)

My problems are gone. Thanks to whoever solved them :) Lycaon 00:15, 15 July 2007 (UTC)

What means Value

From Guidelines, Value: "Our main goal is to encourage quality images being contributed to Wikicommons, valuable for Wikimedia projects."

Grapes in a bowl

I had discussion with picture Grapes in a bowl. It has good technical comfort, but not clear where can be valuable for Wiki projects. What are opinions? --Beyond silence 10:20, 14 July 2007 (UTC)

I have no doubt it has potential usefulness, it is an illustration of a food item, there is even a 'food and drink' category under Quality Images. --Tony Wills 11:32, 14 July 2007 (UTC)

A picture should have value if it is going to be a QI, and we should always remember that it is our primary goal, but it doesn't take much to be useful. If it can conceivably be used to illustrate a topic on any Wikimedia project, present or future, then it is useful. It only needs to be potentially useful. QI is not like FP, in that our focus is on technical quality. Usefulness is less important for QI, although still required. Just my opinion. As for the picture of the grapes, I was not pleased with the quality, even though I disagreed about its value. -- Ram-Man 13:09, 14 July 2007 (UTC)

• Ok, thanks for opinions. --Beyond silence 13:19, 14 July 2007 (UTC)
• But then there is any meaningful of this guideline??--Beyond silence 13:33, 14 July 2007 (UTC)
  • The real technical quality guidelines are Commons:Image_guidelines, the 'value' clause is basically a let out clause so we don't end up promoting absolutely useless (but otherwise technically good) images. But as a general criteria for QI, 'value' is too subjective. Because it is likely that no one has in-depth knowledge of all the wikimedia projects and what they might use, I think that images should rarely be declined on a 'value' basis. This is not FP, we don't need 'high value', we don't mind duplication, we are just trying to promote and improve technical quality. --Tony Wills 23:59, 14 July 2007 (UTC)
  • Ok. But the value can be promoting effect at the QI too, right? --Beyond silence 01:55, 15 July 2007 (UTC)
• technically value isnt a promoting effect, value is subject opinion take this image is technically outrageously bad(my image I can be that cruel) and shouldnt have even been uploaded but it is valuable as its a demonstration image for the guidelines. The same would apply to abstract artistic photographs that are common on most photographic sites where they have a technical quality but aren't useful to Wikimedia projects. Gnangarra 03:23, 15 July 2007 (UTC)
• there is something about this subject, in my case i do take value in count, i mean it may be a subjective thing as you say but i do think we should place ourselves a limit. so what i do is a try to be much more strict and less tolerant to tecnicall mistakes in images that are with no real value. (lets face it an image of a brick wall was really close to the limit, 3 versions from it was out of mind) .what worries me is that the more of this pictures we upload the less we could reject other pictures. i mean i could as well make a diagram of a circle. just a simple svg of a perfect circle. and you would have no tecnicall claims to send it back.-LadyofHats 09:24, 15 July 2007 (UTC)

• Have you looked at Category:Plain circles there are lots of SVG circles, some with different colours! I think we set a limit for QI evaluations when it came to SVG and other renditions of individual letters (in nice fonts perhaps) - we just declined to evaluate, or promote trivial images however well executed they are. So people are welcome to upload them (they do have conceivable uses) but they're not going to get QI promotion. --Tony Wills 11:51, 15 July 2007 (UTC)

• A tremendous variety of images can be used to illustrate Wikimedia projects. Encyclopedias and dictionaries are only the beginning. WikiBooks has the potential for a book on anything. Fg2 11:20, 15 July 2007 (UTC)

SVG displaying problems

ok as you may know (or not) since i started to work with svg files in wikipedia i had the problem that sometimes they would not show. either in the main page of the image or in the thmbnail. this would be solved by changing the file (or thumbnail) size a bit. i made a lot of inetnts geting an answert from the tecnical groups in wikipedia. and finally i got one :). now since this may a quality issue and i know how desperate i was to know an answer i will copy here the text i got.

If you have problems with Commons thumbnails for your SVG, you should know that it is not caused by the size or the graphics features it uses or by a non-standard XML feature. This is caused by the fact that the Commons renderer for SVG does not support the local definition of an internal DTD within SVG images.

To solve this problem, you must remove the part in the first tag of the header element (before the root "<svg ...>...</svg>" element, in the "<!DOCUMENT ...>" declaration, after the referenced external DTD and before the ">", i.e. all the "[...]" part) that contains these named "<!ENTITY ...>" definitions.

But as a consequence, the named entities must be substituted in the rest of the SVG file (for example if there's a definition for the entity named "ns_svg", you must substitute "&ns_svg;" by the value defined in the definition of the ENTITY (here the URL of the SVG specs). As this is just used by Adove illustrator to define the URLs to the XML namespaces used in the opening tag "<svg>" of the root element, within the value after "xmlns...=", these definitions are completely unnecessary.

You can do that using a simple plain-text editor (notepad, Vi, Emacs...). Make sure you keep the quotes around the substituted URLs.

I hope this will help you understand what is happening: private named entities are standard in XML, but their support is not required in SVG specs (that don't require a validating XML parser).

This is a current limitation of Commons' SVG renderer (which does not occur with the Adove SVG renderer plugin for IE or Mozilla) used to generate bitmap thumbnails, and will not forbid you to edit again the image with Illustrator... You actually don't need to tweak with various sizes just to avoid thumbnails (all images on Commons should be resizable with thumbnails).

Make sure that your Adobe Illustrator settings allow removing the XML internal named entities, if it has settings for that (may be it has some settings for compatibility). Verdy p 15:48, 15 July 2007 (UTC)

Retrieved from "http://commons.wikimedia.org/wiki/User_talk:LadyofHats"

i hope this helps you as much as it did me. -LadyofHats 21:39, 15 July 2007 (UTC)

Only self-made?

"Quality images have to be uploaded to Commons by copyright holder under suitable license. " Only self-made pictures can be Quality images? That PD where the copyright is expired can't? And self-made photo about a painting like this? --Beyond silence 18:26, 16 July 2007 (UTC)

There are typically two copyrights in a photo of a piece of artwork. The copyright on the subject, which may of course have expired, and the copyright on the photo itself. Craft is exercised in the making of an image of an image, just as in making an image of an object, or a creature or anything else. It's true that there is more work than just the photographers on display, but we just pass judgment on the part attributable to the photographer. I don't think there can be copyright issues on the painting (Malczewski died 8 October 1929) so I don't have a problem here. Regards, Ben Aveling 19:50, 16 July 2007 (UTC)

And what about Quality images?--Beyond silence 20:24, 16 July 2007 (UTC)

I don't think that a photograph of a painting is eligible to be a Quality Image, after all the goal is to reproduce someone else's work. Same thing with a scanned document. It may be a FP, but not a QI. Cacophony 23:36, 16 July 2007 (UTC)

QI is an acknowledgement of the photographers skill in making high quality images. If they take a photo of a stained glass window or a painting the quality that we are evaluating is their reproduction of the original work - this is not an entirely mechanical thing (that scanning an image might be), they must pay attention to proper lighting, focus, framing, angle, geometric distortion etc etc. It is clear that the requirement to be 'self made' is because a purpose of QI is to promote and improve contributors skills in photography (or creating SVG etc illustrations), not their skills in finding good images on the web etc --Tony Wills 00:25, 17 July 2007 (UTC)

QI is a categorization scheme to identify pictures of high technical quality to make commons a more valuable source for its users. Secondly it is back patting for users who draw their motivation to contribute from putting little green symbols next to the pictures on their user pages. --Dschwen 09:52, 17 July 2007 (UTC)

And third, and maybe most importantly, QI is a forum in which we educate each other and ourselves. When I put up a QI candidate, I expect to get feedback that helps me make my future photos better. Doesn't always happen, but often enough. Regards, Ben Aveling 19:17, 18 July 2007 (UTC)

Wow that was quite an acerbic comment I made half a month ago. Tsk tsk tsk. --Dschwen 15:59, 8 August 2007 (UTC)

I see the focus of QI on content created by Commons users - no matter what the content is. For example, a photograph of a PD image/painting is wellcome at QI, as long as it is created by a Commons user. On the other hand, a photograph of the same painting made by the Museum exhibiting the painting might not be eligible, even if the Museum puts it under a free license. --Florian Prischl 13:09, 30 July 2007 (UTC)

Yep. Thats my view. --Gmaxwell 15:52, 8 August 2007 (UTC)

Wider galleries

There now appears to be parameters for galleries (if they existed previously, then the documentation was well hidden, I couldn't find it!). Anyway there now seems to be parameters: <gallery caption="title" widths="120px" heights="120px" perrow="4">. So I have editted one gallery (20-June-07) to use slightly wider images so that the image description templates are slightly easier to read but still allowing 3 images across at 800x600 or 4 images at 1024x768. What do people think? --Tony Wills 00:06, 21 July 2007 (UTC)

The new wide galleries are nice, except that I think it broke the QICBot, which was probably looking for a <gallery> tag, which no longer exists. -- Ram-Man 13:45, 26 July 2007 (UTC)

Looks like it ;-) Try to undo the changes and see what happens tomorrow around 13:12 UTC :-)) Lycaon 13:53, 26 July 2007 (UTC)

Perhaps the bug will be fixed by then. It'll only take a minute to change, so we can put it off for now. -- Ram-Man 14:50, 26 July 2007 (UTC)

I commented out some of the 'widened' galleries for now. Lycaon 15:38, 26 July 2007 (UTC)

Fixed it yesterday. Please try the wide galleries again. --Dschwen 10:28, 29 July 2007 (UTC)

Great, will do. Thanks. Lycaon 10:50, 29 July 2007 (UTC)

Great, I really like the new, wider galleries! --Florian Prischl 13:04, 30 July 2007 (UTC)

We missed the big number!

Currently 1163 QI have been tagged, the actual number of images assessed is obviously much higher. Great work guys! We missed the opportunity to celebrate number 1000 though. To keep a closer eye on the number I created a little counter. If I'm really bored I might make the QICbot auto update the count on COM:QIC... --Dschwen 18:26, 7 August 2007 (UTC)

Good idea thanks --Ikiwaner 20:30, 7 August 2007 (UTC)

Yes, and now we get a big pack of new. Only problem, the nominators looks like not too good review writers... There is some work! ;) --Beyond silence 22:30, 7 August 2007 (UTC)

Well done guys. QI is a big success. I love that it is such an inclusive and positive process. Keep up the great work (your next challenges will be how to scale the process appropriately). pfctdayelise (说什么?) 13:29, 8 August 2007 (UTC)

I posted to foundation-l about this: [2] --pfctdayelise (说什么?) 13:54, 8 August 2007 (UTC)

Great!--Beyond silence 18:46, 8 August 2007 (UTC)

Thanks for the mention Pfctdayelise in the email but really QI has only been a success because so many editors spent time making it work. You all deserve a barnstar for making QI what it is from Pfctdayelise for the early encouragement, Wikimol & CyrilB for efforts in putting this together Ladyofhats for the QI seal. Then theres all the editors who spent hours reviewing nominations, doing the housekeeping until QICbot stepped up to the plate. IMHO I glad the 1000th picture passed unnoticed because as this is a wiki and everybodys efforts are equal the 999 that past before or the many that will pass after are equally deserving of recognition. Gnangarra 09:49, 9 August 2007 (UTC)

Yes it's a great gallery of quality images, thanks first to the photographers, thanks to the reviewers, thanks to QICbot, and thanks to everyone who contributes in any way to QI :-) --Tony Wills 11:44, 9 August 2007 (UTC)

POTY2007

This might be of interest to you guys too. User:Pfctdayelise started the brainstorming for the Picture of the Year 2007 preparations: Commons:Picture_of_the_Year/2007/Preparation. --Dschwen 07:16, 8 August 2007 (UTC)

Statistics for QIC

I just parsed the old archives and created a graphical representation of the past verdicts. Keep in mind that month 13 (this August) has only just started. Last month (12 = July'07) is a dip probably due to the summer climate and holidays. Month 5 was December'06, and again it looks like people had something better to do during the holidays :-) (and they seemed to have tried to catch up on the noms in the following months). All in all it shows a pretty stable upwards trend with a ratio of promotions to declines of about 1:1. The blue and yellow curves are just improperly closed nominations without the use of a Promotion or Decline verdict template. --Dschwen 08:47, 8 August 2007 (UTC)

Improperly closed isn't quite the right word for the yellow and blue graphs. Nominations are archived if not voted on in 15 days (so stay blue). Similarly two nominations that were discussed but had no determination were archived in May. PS what summer climate?, it's winter around here ;-) --Tony Wills 10:50, 8 August 2007 (UTC)

No offense, I know you shoulder most of the load in closing consensus noms and sorting the executed promotions into the right categories. Your work is most appreciated! --Dschwen 10:57, 8 August 2007 (UTC)

Hear, hear!! -- Lycaon 14:00, 8 August 2007 (UTC)

No offence taken :-), thanks for the support, we'll have to look a getting QICbot munch on the CRs too :-) --Tony Wills 11:47, 9 August 2007 (UTC)

All is not what it seems grasshopper ...

A prize to the person who names the common factor that links all these images:

• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 
• 

--Tony Wills 11:51, 9 August 2007 (UTC)

• They are all taken by User:Fabelfroh? --Digon3 ^talk 13:24, 9 August 2007 (UTC)
• They are all in the quality image category without being promoted --Simonizer 14:05, 9 August 2007 (UTC)
  • We are going to need a bot to untag them all. --Digon3 ^talk 14:28, 9 August 2007 (UTC)
    • Yes I think all except Image:NMOS E OFF.png are by the same person, and yes none have gone through QIC but are categorized as QualityImages. A significant distortion of our QI numbers! As they are all of reasonable quality, I will suggest to the authors that they submit them to QI and perhaps participate in reviewing others.

I'm doing a bit of an audit of the QI images as there is a considerable mismatch of the numbers in the category and the images displayed on the QI pages (I expect there may be a few accientally not added to galleries, but there may be other 'mistaken' additions :-) --Tony Wills 04:37, 10 August 2007 (UTC)

I could do an automatic comparison of the image set in the archives with the images on the QI pages. And it should be easy to find all images without the QI template which are including the category directly. --Dschwen 06:54, 10 August 2007 (UTC)

Not quite straight forward because some image names in the galleries have underscores where they should have spaces which means a simple text comparison doesn't work. I have come up with a list of 94 images in the category that aren't on the QI pages (this includes the 70 odd images above). So far I've found 3 which have accidentally had the {{QualityImage}} tag removed, and two that were added to QI pages without going through QIC.

A nice automated check would look at:

1. Images in category:Quality Images but with no {{QualityImage}} template
2. Images with {{QualityImage}} template but not listed on commons:Quality Images pages
3. Images listed on commons:Quality Images pages but without {{QualityImage}} templates
4. Images with {{QualityImage}} template but not linked to QIC or its archives

(and probably other circumstances I haven't thought of). I did these checks when I started here back in March/April, obviously I should recheck more often :-). I've also been meaning to do similar checks over on FP. Of course none of that checks that the images were actually promoted, nor whether that promotion was legitimate :-). --Tony Wills 08:09, 10 August 2007 (UTC)

Ok, finished my "audit", found a few images that were promoted but not listed in QI galleries, a few self promoted or images with mistaken use of QI template or category and a number of images which had "inherited" their QI tag from a previous version - I make no judgement about the quality of the edits but think it is unreasonable to have images inheriting QI. They need to go through QIC so that at least one other person verifies that the "improved" version is indeed QI. (User:Fabelfroh images still need de-categorisation). --Tony Wills 12:03, 10 August 2007 (UTC)

I have put in a bot request for de-categorisation. --Digon3 ^talk 21:48, 13 August 2007 (UTC)

expansion

While discussing the POTY proposal I suggested that maybe QI should be expanded to include sound/video thats meets the self made criteria. Thought I'd bring it here and let it get thrown around for a while. Gnangarra 14:37, 9 August 2007 (UTC)

Agreed. We'd need to ignore most of the existing guidelines and work from common sense for a while until new guidelines emerge. I don't see that as a problem. Anything you'd like to nominate? Ben Aveling 08:02, 13 August 2007 (UTC)

• i myself would step out in what quality of sound or video is.. it goes out of my knoledge area -LadyofHats 09:27, 15 August 2007 (UTC)

QICbot - tag

I know its a dangerous night here two different thoughts inside a couple of hours.

Anyhow what I'm thinking is for the promotion template to have an extra variable where when reviewing we could just add |a or b, and so on after signing and the QICbot could read this variable as a sort key and place the images directly into QI categories and galleries. with un-categorised images still being dropped on to the sort page for human processing.

The other thing question is it possible for each on 4 image galleries on the main page to be changed to an equation that selects the first four images off the main subject gallery removing the necessity to update this gallery as well. Gnangarra 14:44, 9 August 2007 (UTC)

No, but if images get category tagged teh QICbot could update those lists as well. Incidently I'm currently working on a slightly different solution.... --Dschwen 14:57, 9 August 2007 (UTC)

• A simple solution would be to ask people to submit images into galleries by topic rather than having a page full of galleries grouped by date submitted. The images would still be sorted by submission date if people add images to the top of the gallery. (Of course QICbot would have to be changed slightly). Mis-placed images could be moved by reviewers (or anyone else) to keep things tidy. --Tony Wills 04:44, 10 August 2007 (UTC)

• This is actually a pretty good Idea. Then I'd propose to keep the consensual reviews in the topic-sections too. It wouldn't be too hard to adapt the bot. My idea was a javascript helper, which in edit mode shows previews of the images with dropdown lists to select a QIcategory and which inserts the correct tags into the edit box with the push of a button. --Dschwen 06:49, 10 August 2007 (UTC)

My idea might not be quite that simple after all, as there are a lot of sub-galleries on each QI page so we would need many equivalent galleries on the QIC page. I also see some of the FP pages have had a face-lift and likewise we probably need to do something more than just bunging images in at the top of each gallery. --Tony Wills 12:10, 10 August 2007 (UTC)

Thinking aloud atm, what if a) we work a full redesign of the gallery pages, the current format isnt go to scale especially if we continue/increase to maintain the current 100 images per month average. b)the images are/should already be in normal categories when nominated can these be used to to put them into the categories/galleries within QI. Gnangarra 12:24, 10 August 2007 (UTC)

Megapixels

I noticed that recently objection has been raised as to whether the resolution of an image should affect whether it is a QI. I understand that guidelines are guidelines, not rules, but I disagree that image resoulution should be completely removed from the decision-making process. Pictures that look soft and blurry at 6MP are often very sharp at 2 - but the sharp downsampled 2 MP image can only be as (and almost certainly less) detailed than the 6MP one. If indeed the number of pixels is to be ignored, there should be some sort of note regarding the detail. I am all in support of removing the quantity measurement, given some sort of quality standard for detail. I think that's it is outrageous opposing for the difference between 1600 × 1200 and 1600 × 1250 or even between 1600 × 1200 and 1800 × 1250 - the difference in ignorable at most. Some other opinions on the importance of megapixels? Thegreenj 18:47, 12 August 2007 (UTC)

I think too a bit smaller than 2 MP can be a quality picture. But in fullsize view have problems between resized versions... --Beyond silence 23:20, 12 August 2007 (UTC)

Originally the 1600px wasnt a guideline but a requirement, this is about qaulity images size is as much a quality issue as focus or noise. The purpose was that all we had here on commons was vast amounts of images under 1000px and that was all that was still being uploaded, images over 1600 px were non existant. to get a good quality print at A4 the image needs to be 3000 px (long axis) which will then also give an ok A3 and even some will be able to be taken to A2 but at 1000px an image is nothing more then 6x4 snap shot print after that the quality is degraded to quickly. Its fine if people want to upload small images but they arent quality images, I actually think the limit should be above 2000px remember that any reasonable quality point and shoot digital camera(even some phones now) are creating images files with this dimension anyway, DLSR even the cheap ones(nikon, pentax) now are 6mp(3000x2000px) or above. Gnangarra 02:01, 13 August 2007 (UTC)

Print to a whole A4, is it requiement - you are serious...? --Beyond silence 02:31, 13 August 2007 (UTC)

Please read again 3000px for A4, the requirement is 1600px, its not beyond the ability of the average mobile(cell) phones on the market atm neither is it beyond the ability of most digital cameras. QI was something I came up with after seeing the end results of a Wikimedia investigation into attempt to provide 12 images of different subjects that were create by commons users for a calendar to raise funds and the profile of Commons, something that didnt happen because there wasnt the images avaiable. Its also beyond me why after 1000+ QI have been promoted we'd degrade the requirements to suit those that are unwilling(able) to provide images at resolutions of 75% of the average camera produces. Gnangarra 03:37, 13 August 2007 (UTC)

I think the definition of QI is: has the user done a good job. In general, that means that they managed to extract a reasonable number of bytes: if you can't get 2M out of a modern camera, unless we are talking something exceptional, you didn't achieve good technical quality. I believe that there are some exceptions, perhaps microscopic shots for example, and there are some images that are great images despite a lack of technical quality, but as a general rule, 2M is no big deal, and for easy to take photos it probably falls short of what we should expect. IMHO of course. Regards, Ben Aveling 08:00, 13 August 2007 (UTC)

Fully agree with garra and Ben here. I don't see a reason to lower the bar. If anything we should raise it to accommodate the technological advancements on the camera market... --Dschwen 08:18, 13 August 2007 (UTC)

• I understand that QI would be overwhelmed with submissions if we do not have some arbitrary limits. (for instance we don't evaluate 'trivial' images). I think an image should be evaluated for what are truly 'quality' characteristics (things like focus/blur, composition, exposure, noise, detail etc) and if it hits all those bases then resolution is less relevant. So I have a problem with evaluations that reject images out of hand without reviewing the image - if it is a 'quality' image in all other respects then excluding it because of a resolution requirement that is subject to current technology, does not seem reasonable.
• Resolution relates to what you want to use the image for rather than inherent image 'quality'. If QI is indeed to find images for calendars or posters or screen 'wallpaper' a resolution requirement makes more sense, but then 2MP is too small (4MP for A5, 8MP for A4?). Most wikimedia projects are web based and have rather smaller resolution requirements.
• In general I think people should be encouraged to supply images with as high a resolution as possible, and a 2MP guideline is certainly reasonable. For many things like shots of static objects (buildings, scenery), studio shots, and captive or tame animal shots, it might be reasonable to expect higher resolution and greater perfection. For un-repeatable shots in the field - wild animals, moving vehicles, sports shots - a 'quality image' evaluation has to make allowances, it is much more likely the shot will need to be cropped or have other minor technical deficiencies (I am not yet convinced either way about microscopic images). --Tony Wills 12:05, 13 August 2007 (UTC)

This is a question of what we accept as a quality image, I was given a keyring camera with fixed focus that takes a 1mp image its ghastly, my kids have phones all of which have cameras with 3-5 mpx the image size they provide meet this physical 1600 px requirement. The first digital I brought 3 years ago was a 4mp (hyndai HYDV4) fixed focal, it gave a reasonable result the image sizes (this image was taken with it) it exceeds a 1600px requirement that camera only cost $75.00AUD, this image is a QI taken with a Kodak Z7590 that camera cost $300 to buy them now 2nd hand for $100 it size meets the requirement. Equipment costs arent prohibitive to get good photo at sizes comfortably exceeding the requirement, all point and shoot camera that are in Harvey Norman catalogue are 3mp or more which all exceed the 1600px requirement. So the question is if an image taken with a 4mp camera has to be cropped/downsample by more than 40%(1300/2300 or 0.56 reduced 44%) how can it be a technically good image, even at 1600px its had 31% of the originally captured image removed. Whether image is repeatable or not isnt a technical consideration, FP can and does evaluate on the uniqueness of an image, hence the dislike/higher requirements for sunsets there. Allowing at least a third(with my current camera its 2/3) of the image to be removed either by cropping or down sampling is more than ample allowance for minor technical deficiencies.

Maybe we should be assessing the image based on a camera capabilities ie if it captures at 3000px then the image should be above 2000px or 2/3 of the capture. If the camera captures at 2400 then the minimum is 1600px, so to get a 1000 px minimum the camera has to be capturing at 1500px. Gnangarra 15:25, 13 August 2007 (UTC)

What!? This is an upside down approach. QI should be for the end-user, creating a library of technically good images. The end-user couldn't care less which camera was used to capture the image, it's the result that counts. --Dschwen 16:13, 13 August 2007 (UTC)

I don't have 2 megapixel photo because my camera can't capture this size. I haven't because on a long holyday I haven't got enough memory card to capture lot of photos at big size... But the pictures don't look whorse because they only made at 1,6 megpixel. --Beyond silence 18:18, 13 August 2007 (UTC)

Not to get off topic, but the DX7630 is 6MP. While 2MP is not out of reach even from cell phone cameras now, it's worth noting that many of the pictures that come from these cameras, and other extremely cheap P&S is usually not worth its face value. And, while I agree that the limit is not to encourage downsampling, the difference between a sharp 1.5 MP picture that doesn't meet QI and a straight-from-the-cheap-camera 4 MP really is not that much. 2 MP/1600 px is very easy to reach by today's standards, and, if anything, as Dschwen said, should be raised. However, setting any arbitrary limit ignores the pictures under the limit that are sometimes superior to those above it. Currently, I don't think that this is a problem - the de facto community standard seems to be working well enough at seperating those pictures that are worth less than their MP. Thegreenj 20:20, 13 August 2007 (UTC)

Right, and if we realay get spectacular slightly below the limit pics we don't turn them away either, cos' we're not rule-execuing robots, but human being with common sense! --Dschwen 21:29, 13 August 2007 (UTC)

I think we do need to be a liitle pedantic and turn images under 1600px away, but at the same time we can exercise some descression for exceptional images as well. It's like going fishing the small ones dont taste any different to the bigger ones but there are good numbers (100+ per month) being caught so theres no need to take the little ones and if we keep taking the little one theres no incentive for larger images. QI is about encouraging photographers to improve what they are providing to Commons and 1600px isnt a difficult line to get as all cameras are providing above that its just comes down the the photographers ability to get the best result out of the equipment they have. QI isnt meant to be a trophy for the photographer because an image isnt QI doesnt mean its not valuable to Commons nor does it mean the photographers efforts arent appreciated. If a photographer wants to get trophies then they should be targeting FP as QI is about building a collection of very good images that are suitable for various usages including paper reproductions. Gnangarra 15:29, 15 August 2007 (UTC)

So, what is an exceptional image, then? Is it one like the microscope candidate, which would have been hard to take at a high resolution because of cropping? Is it a spectacular photograph whose strengths make up for resolution (like a few FPs)? Just to grap a current nom, Image:Darter August 2007-25.jpg has strong migitating circustances, which, according to the guidelines, can make up for low resolution. However, for all its illustrative quailities, it's still a 1.4MP photo that is blurry - looks like it has been run through a median filter. Is that QI (I'll leave the judging to the reviewer)? And what about the 1500 * 1000 picture that's tack sharp as compared to a typical 3MP? Although, 1600px is very reachable in near all cases. Should we just turn away all under 1600px images because it's so easy to get higher resolutions? I'm not disagreeing with you, but I'd like to know just where to draw the line. Thegreenj 22:39, 15 August 2007 (UTC)

I think it has to be a case by case decision. In a given situation, were the right tools used, and were they used well? For an easy to take picture, such as a common or captive animal, that normally means at least 2MP, or more. For something completely unpredictable, like a bridge collapse, I might support a well composed shot from a camera phone. For microscopic work, I'm not completely convinced, but I'm sympathetic to a lower number if such cameras are still widely used in professional level work? Regards, Ben Aveling 10:54, 16 August 2007 (UTC)

Microscopic pictures look difficult and out of reach for most of us because we don't have access to microscopes + cameras. For people working with a microscope on a daily basis (I could if I wanted), those pictures are run-of-the-mill, and 800x600 is also in microscopy images, a low resolution. Technology didn't stop with digital cameras, it didn't for microphotography!! -- Lycaon 14:36, 18 August 2007 (UTC)

Changes to subpage templates

I have changed /Decline, /Promotion, and /Nomination so that they make more sense. This way, it is completely obvious which piece of text is the nomination and which is the declining/promoting review. Before, it seemed as though the nominator was the one declining or promoting the photos, which didn't make sense at all. --bdesham ★ 20:55, 16 August 2007 (UTC)

Nice, looks much more structured and self explanatory. --Dschwen 22:49, 16 August 2007 (UTC)

QI CR closing

(post from user's talk)

• Hi, as per rules consensual reviews can be closed after a period of 48 hours during which there have been no further votes or comments where there is a clear result (majority of one), or after 15 days if no decision has been made.

Quality Images was set up to be a fast process to promote images that are clearly quality so that we can process lots of nominations without wasting a lot of time - that is why images are promoted/declined on a single review. If people disagree with the review they take images to consensual review, but if no one has added further votes after 48hours it is clear that no one (apart from the nominator perhaps :-) strongly disagrees with the review, so it can be closed as per that review. Often because of time constraints I'm not able to close CR after 48hours, anyone can close reviews as per the rules if they think I'm too slow :-). Sometimes I try to clear out a backlog and close everything that can be technically closed as soon as possible. If you really want to stop something being closed, just keep adding comments every couple of days! But! that will probably annoy people, so they won't support it anyway :-).

The consensual review process is really a sort of quality control of the voting process, to ensure that reviews are not out of touch with what most other people think. --Tony Wills 20:56, 18 August 2007 (UTC)

• I think there is strong need to change. 48 hours is realy to short, you can see when you don't close the voting as soon the picture get some votes after 48 hour too! But I think with only one vote the voting may can't be closed, need to use 5 day rule I think. What are other's opinion? --Beyond silence 21:09, 18 August 2007 (UTC)

• I agree with the 48 hours period but won't oppose a slight increase if consensus is reached. As for the 15 days' period it is way too long. I would reduce it to 8 or 9, like we have already done in FPC. Another point is the excessive number of pics submitted by some users, which is making the whole thing quite boring to the reviewers and is affecting the quality of the evaluations. I've raised this point before (Commons talk:Quality images candidates#Dumping pictures on QIC - a bad idea) and there was then a suggestion for limiting the number of daily submissions. Maybe we should discuss the issue again. Alvesgaspar 22:11, 18 August 2007 (UTC)

I see as don't need to limit the possible nomination's number. Users are seeing if there is a lot of image without review and I experinced after a great wave didn't arrive unhandable nomination. And technicaly hard to working this rule, I think. --Beyond silence 00:12, 19 August 2007 (UTC)

I agree we don't really need to limit numbers, if people flood submissions they will probably not be reviewed very fast and may just drop off the end without review - self defeating for the nominator. Reviewers pick and choose what to review and are under no obligation to review everything, so submitting lots of similar images also is likely to be self defeating.

I agree we could change the CR rules, and require at least one vote in addition to the reviewers and shorten the timeout period to a week. There is also another problem - lots of CRs sit for a long time with no decision as votes are equal, or even with no votes because it has been moved to CR because the reviewer can't decide - I'm not sure whther this is because people don't notice the stalemate or are just not interested. Perhaps we can also make it policy to only accept things for CR that have actually been promoted/declined and for others, where the reviewer can't decide, just leave them as nominations for someone else to review. --Tony Wills 00:05, 21 August 2007 (UTC)

We could also ask for the bot to move images after two days once a decision has been made, that might make the main page less cluttered. --Tony Wills 00:05, 21 August 2007 (UTC)

I guess I agree with all but one of your points (how wonderful!). The extra vote from the reviewer, would again make the issue drawn out in time as it would turn the procedure into half a CR. As for shortening timeout and bot timings, that might indeed help.
Speaking for myself, I often see pictures where I can't decide and may only react (with slapping a CR on it!!) when i see the arguments of the voting party. I know I'm rather picky for QI (less so for FP), but I like to think that I use the same rules for my own images and won't submit any I wouldn't let pass myself. Lycaon 06:00, 21 August 2007 (UTC)

After two day is a bit too short period for the bot. Need more days to keep up the control, if there is lot of pictures. --Beyond silence 06:31, 21 August 2007 (UTC)

Proposal for QIC closing

I also agree with Tony Wills ideas, except the one regarding the need for two votes. I suggest we vote on the following proposal for altering the rules:

1. Reduce 15 day's period to 8 days;
2. Images can only be moved to Consensual review after a positive (promotion) or negative (decline) review has been made;
3. Images should be removed 48 hours after a decision has been made.

•  Support - Alvesgaspar 22:47, 21 August 2007 (UTC)
•  Support Except for #3. Could you expand on that a bit? Ok, as long as #3 only means the main QIC reviews and not CR. --Digon3 ^talk 14:37, 22 August 2007 (UTC)
•  Support - Lycaon 14:45, 22 August 2007 (UTC)

 Info Just to clarify, I was talking about CR closing, not general QIC closing. But I don't have a problem with the QIC and CR timeout being shortened to 8 days (is that what voters are meaning?). #2 is clear. #3 I presume relates to my suggestion for the BOT to be run after 48hours so relates to the main QIC reviews and not CR - so reword as "QICbot should execute promotion/decline decisions 48hours after review". --Tony Wills 21:31, 23 August 2007 (UTC)

•  Support all if my interpretation is correct. --Tony Wills 21:31, 23 August 2007 (UTC)
•  Oppose I don't know why need this. If somebody take a picture to CR presumably he/she has an other vote. Or he think need other opinions because he/she is neutral. --Beyond silence 22:46, 23 August 2007 (UTC)
•  Support1, 2 and 3 in general. On item 1, I support the time for both CR and QIC to be shortened. 2 is clear. On 3, I concur with Tony Wills that the mentioned 48 hours pertain only to decided QICs and not to CR. --Florian Prischl 17:34, 24 August 2007 (UTC)
•  Question I suppose a week will be long enough for this vote, so run it to 22:47, 28 Aug (UTC)? --Tony Wills 03:10, 25 August 2007 (UTC)

Proposal for CR closing

1. That CR needs at least one vote other than the reviewers, before being closed. (Which basically honours the request for CR, otherwise no CR has really happened). (If no CR votes are forthcoming then the original review would be accepted after 8 days). --Tony Wills 21:31, 23 August 2007 (UTC)

•  Support --Tony Wills 21:31, 23 August 2007 (UTC)
•  Support Good idea. We really should not have any unassessed QI candidates. --Digon3 ^talk 21:38, 23 August 2007 (UTC)
•  Oppose Unnecessary I think. If somebody take a picture to CR presumably he/she has an other vote. Or he think need other opinions because he/she is neutral. I think if don't opposse anybody can be agree one support vote too. If there isn't anybody who think that picture can be QI, then I can be a QI. --Beyond silence 22:52, 23 August 2007 (UTC)

 Comment!? strange, this proposal was in response to your queries as to why some CR were closed with only one vote (the reviewers). Most often it is the nominator who takes things to CR, their vote is not counted, so to actually get some idea of a consensus should we not have at least one more vote to confirm or reverse the initial review? --Tony Wills 04:23, 24 August 2007 (UTC)

•  Support It strengthens the idea of CR. --Florian Prischl 17:35, 24 August 2007 (UTC)
•  Support Lycaon 17:43, 24 August 2007 (UTC)

 Info Much to my embarrassment I have just noticed the the CR rules state "In case of draw, or if no additional opinions are given other than the first reviewer's, the image will stay in Consensual Review for a maximum period of 15 days, counted from its entry." (emphasis added). Which really covers the situation already. So Beyond silence was right to query a couple of CRs that I closed without further votes after only 48hours. I will go back and reinstate CRs that I closed too early :-( --Tony Wills 01:02, 25 August 2007 (UTC)

I have added 6 images back to CR, which were mostly closed by me as soon as possible to get things flowing through CR faster. If the new rules are agreed to it will help, but the best way is for people to add their opinions to stalemated reviews to get them moving. --Tony Wills 02:57, 25 August 2007 (UTC)

Resolution requiement

I think there is need to clear: size can be decline reason or not. Tony Willis says not, but Lycaon using it (as on Image:Warzenbeißer (Decticus verrucivorus) m 01 (HS).jpg's nomination. So I think voting: 2 megapixel is minimal requiement or not?

There are not rules regarding promote/decline, people vote, someone can decline an image of 4Mpixel being 'too small' if they want, people can decline if they don't like green things etc. Presumably such cases would be taken to CR and a majority would decide otherwise. But this is a community thing, not some rule bound measurement of photos to see if they fit in some exact formula. There is a guideline suggesting 2Mpixel is the usual 'lower limit', basically to save people wasting too much time on things which are unlikely to be promoted. So once again the answer is "no" but people are free to vote as they see fit (but they shouldn't invoke some 'rule' or 'limit' to back up their personal preferences). --Tony Wills 04:17, 24 August 2007 (UTC)

Then if any reason is acceptable may I won't write reason ("people can decline if they don't like green things")... --Beyond silence 05:15, 24 August 2007 (UTC)

QI is about raising the standards of self made images, but if the lowering or removal of size standards occurs then its only reasonable that all images since QI inception where the size requirement was applied should be list at CR. Gnangarra 06:09, 24 August 2007 (UTC)

(answer to Beyond silence), If reviewers make poor decisions and promote/decline for bad reasons their reviews will be over turned at CR (if they are making silly decisions they may be taken to task for disrupting commons). But my point is as always, guidelines are not RULES, we accept or decline images on a case by case basis. Guidelines give nominators and reviewers clues as to what others consider quality images. It is hard to get things smaller than 2Mpx promoted, it is hard to get things with lots of noise promoted, it is hard to get things with poor depth of field promoted but none of these things bans an image, it is case by case. --Tony Wills 06:31, 24 August 2007 (UTC)

I would support a suggestion to require a minimum of 2 megapixels (and the possibility to increase that minimum in the future). However, I do see problems with a minimum, namely that we would have to create an exact list of exceptions (for example, the earlier-mentioned microscopic images) lest we do not want just guidelines on exceptions, which would soften the whole minimum requirement --> reductio ad absurdum. Maybe a wording of something along the lines of "2 megapixels is the strongly encouraged guideline for the minimal size of pictures" is better? --Florian Prischl 17:42, 24 August 2007 (UTC)

I am happy with a 'strongly encouraged' guideline, but not a 'rule' or 'limit' or 'minimum'. Your suggestion about increasing the 'minimum' in the future implies you think it is a technology question. This is what I have always believed, it is not an image 'quality' question, but an arbitrary limit defined by current technology. Some people seem to want an absolute rule that can be used to eliminate images for this one criteria, why? We have no such 'rules' for any other criteria. What is wrong with evaluating images by giving reviews like 'poor focus', 'insufficient DOF', 'too noisy', 'lacking detail' etc, is that so hard? --Tony Wills 02:46, 25 August 2007 (UTC)

Is image quality not relative to avaliable technology? What would have been considered astoundingly good image quality in a digital camera five or six years ago is generally cheap junk today. I think that it is safe to say that the level of detail expected for a consumer camera is rising, and with expectations come standards. I think that it is fair to have changing guidelines to reflect current technology. Thegreenj 03:33, 25 August 2007 (UTC)

Agreed. The image standard has increased in the past, and will probably increase in the future. The reason that the standard hasn't increased as fast as the technology is 1. the real amount of information has not increased as fast as the number of pixels would indicate and 2. the 'need' for more pixels has not increased as fast as the number of pixels. 2M is enough for most purposes, and as you can see, is also enough that many images fall short. When either of those things stops being true, the minimum requirement will go up. Regards, Ben Aveling 12:11, 25 August 2007 (UTC)

Stupid question, but what is the maximum acceptable? I do not remembre well, I have had a message once or twice saying that the file was more 4 Mb and this is recommended.Romary 11:11, 25 August 2007 (UTC)

If there are more pixels than there is information, it's too big. It's not hard and fast. Regards, Ben Aveling 12:11, 25 August 2007 (UTC)

That sound good, but for a non-specialist, what does that mean? How do I know there is more pixel than information?Romary 10:39, 26 August 2007 (UTC)

When you don't lose any detail by downsampling. Thegreenj 02:57, 3 September 2007 (UTC)

• I also don't see the need for a strict rule on resolution. I like the text adopted in FP guidelines: Photographs of lower resolution than 2 million pixels (e.g. 1600 x 1200 = 1.92 million) are typically rejected unless there are 'strong mitigating reasons' . - Alvesgaspar 16:06, 25 August 2007 (UTC)

But here can't use such guidelines because it not FP, it's QI. --Beyond silence 21:16, 25 August 2007 (UTC)

I think we can, and I'd even remove the word strong. Quality is about doing a good job given the subject matter at hand. Regards, Ben Aveling 00:01, 26 August 2007 (UTC)

As nominater I've no problem with this guideline, in this case I nominated it because of the motive. But I would change the bottom line of the guideline to 1600x 1200 as this is a 4:3 ratio and 1600x1250 isn't one. I didn't expect to start a discussion that fills one screen ;-). It wasn't possible for me to get closer without having a shadow of the camera itself on the bottom on the wart bitter. A better technical equipment for macros will help next time and I will be able to stay far enough away without causing shadows. I cropped it because I wanted a picture with minimum surrounding landscape for the article where pictures are normally 250 - 300 px broad. Imho with only ~300px it is important to minimize loss of details caused by landscape. --Hsuepfle 13:12, 26 August 2007 (UTC)

Call

More reviewers needed for QI, please. Lycaon 12:31, 1 September 2007 (UTC)

Proposal for QIC closing - result

I declare our little vote has decided:

  1. Reduce 15 day's period to 8 days (for QIC and CR)
  2. Images can only be moved to Consensual review after a positive (promotion) or negative (decline) review has been made;
  3. Images should be removed 48 hours after a decision has been made (for QIC QICbot)

I will edit accordingly --Tony Wills 12:06, 2 September 2007 (UTC)

Problems with QI promotion

Looks like there's some sort of bug in the system. I personally declined Image:Cicindela hybrida 4 Richard Bartz.jpg and yet when I visit the image page it says it's a QI. I'm kinda new to Commons so I'm not really sure how the system works. Calibas 02:14, 14 September 2007 (UTC)

Looks as if Richard was his own bug. He added the template on uploading (probably copied it together with the rest of the info from an already QI'ed picture). I removed the template. Nothing to worry about, but thanks for noticing. Lycaon 04:05, 14 September 2007 (UTC)

Pictures not showing up

For some reason I can't see the thumbs for about half of the images that have been posted in the last few days. I've tried the button to purge the page cache and I've manually cleared my entire cache to no avail. I've tried switching from Firefox back to IE and I still cant see them. The only way I've figured out is to change the image width in the gallery tag to something other than 145. Anybody have any idea how to solve this? Calibas 03:18, 16 September 2007 (UTC)

I've got this problem all over the 'pedia... it's not just this page. There's a couple of threads about it on the village pump at en.Wikipedia. —JeremyA 03:32, 16 September 2007 (UTC)

Apparently the problem was "The disk was full, the source of the problem has been fixed so just a lot of purging is needed". Oops. Calibas 04:03, 16 September 2007 (UTC)

Maybe we should suspend/postpone QIC - promotions/declines for a few days untill the thumbnails are back. It is both not fair to the declined ones (who could get a dsicussion), or to the voters who might differ in opinion about a promotion. Without thumbnails, this becomes very difficult. Lycaon 16:24, 16 September 2007 (UTC)

Just ?action=purge these pictures. --Dschwen 16:36, 16 September 2007 (UTC)

External links break {{Discuss}}?

I just posted something in a CR discussion and wanted to include a link to a diff I created. However, when I previewed it, the link broke the {Discuss} template. The discussion was not displayed, just a {{2}} in its place. I will look at the template a little later, but I'm not that good with the Wiki markup, so I might not be of much use. If you have any comments or idea, please contribute. —the preceding unsigned comment was added by Florian Prischl (talk • contribs)

• I have noticed that I cannot use my signiture (--Digon3 ^talk) inside discussion, nomination, or promotion, so I guess the template can't do html (or is it javascript?). By the way, this page need archiving bad. --Digon3 ^talk 01:33, 19 September 2007 (UTC)
• Having "=" signs within the template is the problem with the 'diff' (it no doubt tries to substitute parameters), need to hide the "=" from the template, I can't remember the best way but I'll find out :-) --Tony Wills 09:18, 19 September 2007 (UTC)

• I've encountered the same problem and I solved it with a workaround, putting the offending link out of the box. Lycaon 09:41, 19 September 2007 (UTC)
• I noticed another way to allow "=" inside the template, just put "2=" at the beginning of the review section (setting the value of parameter 2 of the template). eg example.jpg|{{/decline|good pic --fred|2= Its size = too small}} --Tony Wills 19:02, 4 November 2007 (UTC)

• how did you find that out ??? Lycaon=

Policy on objecting an assessed image

It has happened more often recently that an already assessed image (promoted or declined) has been "brought back to life" by a second reviewer who replaces the used template with {{Discuss}} and adds his own comment whilst the image is left in the regular gallery and not added to CR. To me, this is confusing, as my understanding of the CR policy is that if there is a voiced objection, the objecting party should move the image to CR and not leave it in the gallery. Based on this, my view is that a third opinion in either direction (promote or decline) may immeadiately promote or decline the image while it is still in the regular gallery. In my eyes the only alternative to that is immediately moving the image to CR as soon as an objection emerges. A discussion within the regular gallery is very distracting. --Florian Prischl 18:32, 22 September 2007 (UTC)

• I agree, that is really confusing. But I admit one exception. If the second opinion goes in the same direction as the first there is no harm in inserting it while the picture is in the gallery. That is normally done when the reviewer wants to reinforce the first evaluation or express his admiration/disgust - Alvesgaspar 18:58, 22 September 2007 (UTC)

• That is right. Of course, a second opinion in the same direction does no harm. However, I would say that it should not get too long - but I do not remember that ever being a problem for me. --Florian Prischl 19:40, 22 September 2007 (UTC)

• If you read the instructions you will see it tells people to just change the header to /discuss if the procedure looks to complicated, this is so that anyone without prior experience with CR can easily contest a review. No need to brow beat people about it, just move any /discuss templates to CR yourself if needed --Tony Wills 13:02, 23 September 2007 (UTC)

• You are, of course, right, and I should have read the instructions more carefully. However, I would like to add that the instructions also contain the sentence "Or just try, you'll get it right if you carefully follow what everyone else has done." ;-) Especially users who have been around QI for a while should be able to do this. --Florian Prischl 23:41, 30 September 2007 (UTC)

Lost thumbnails

I think the only way to have back the lost image thumbnails is to re-upload the pictures. I have just done that with the  Info template and it worked. - Alvesgaspar 10:24, 23 September 2007 (UTC)

Procedure to retroactively deprive QI status of an image

I would like to jump on the train of Beyond silence's discussion at the CR of Pannonian Knapweed regarding already promoted QIs. There are (in general, not only in the gallery Beyond silence drew up) and always will be images that are awarded QI, even though they should not and would probably not have withstood CR. The reasons for this are many. Personally, I think it is that most reviewers tend to skip over already promoted or declined nominations (I know I am doing this) and catch only the most glaring mistakes. This is why I suggest thinking about a procedure to retroactively dis-award QI status to images. For this, I would imagine a re-nomination directly to CR where the image could be debated. However, I think this should only be applied to very strong errors - some margin of error may be acceptable. What do you all think about this? --Florian Prischl 10:25, 30 September 2007 (UTC)

• I strongly agree. Unlike FP status, which is more like a award, QI status is a certificate of quality, and the whole set of QI is like a warehouse of high quality pictures where we can look for a good illustration for our projects. Moreover (I keep repeating myself), quality images should be used as positives examples of sound photographic technique specially for those making the first steps. Because of tecnological evolution quality standards are changing quite rapidly over the years and better photos are being uploaded to Commons, with obvious reflexes on the reviews. In short, I think that a new delisting section should be created in QIC, with similar (but "symmetrical") rules to CR. - Alvesgaspar 11:35, 30 September 2007 (UTC)
• One more word: I don't think that size should be invoked as the main reason for deslisting a picture. In my opinion, this new section (if approved) should be used mainly to correct obvious errors of judgement (which are relatively common in QI) regarding the relevance ("value"), exposure, DOF, composition, sharpness, framing, colouring, etc. of the images - Alvesgaspar 12:02, 30 September 2007 (UTC)

In general, I don't agree. QI is no big deal. All that it means is that someone liked it and no-one disagreed, or at least, that more people liked it than not. In particular, Quality means Quality of Execution - if was a good quality photo at the time it was nominated, that's all we ask. Unlike FP where we are judging the best of the best, QI is asking "did the user do a good job with the tools reasonably available at the time". That's not something that stops being true simply because it becomes possible to do better later. Unlike FP. Regards, Ben Aveling 11:36, 3 November 2007 (UTC)

• I think we could relatively easily find a compromise solution. To get the QI withdrawn could be a "painful" experience for the author, unnecessarily hurting his/her feelings. But we could somehow separate the "active" QI gallery from the "historical" gallery, and the previous one would be the present-day collection of positive examples (Alvesgaspar), the last one showing the history of our judgments. --Szilas 05:27, 4 November 2007 (UTC)

• In a blue sky world I would support the delisting of QIs, but I think our limited resources are better spend other ways. The QI process is a fast track process, which allows for processing significantly more images than at FP per day. The drawback is that there are variances in the reviews. not as many reviewers look at the same photo as at FPC. If we were to introduce some kind of delisting process I can only see we would steal time from the reviews thus introducing further errors in the intial review process. I'd say, keep QIC as it is. -- Slaunger 13:26, 6 November 2007 (UTC)

• QI is/was intended to be a fast process to identify images of quality at/near time of uploading. To retroactively delist QI would have a greater negative effect than having images still tagged with the seal. Maybe the galleries need to be cleaned out and the proposal to included a QI category in the POTY would be a way to do this without the baggage of a delisting process. The main page currently displays a handful of the most recently promoted, with a link to a larger complete gallery and there's no issue with images being removed from the main page. Why not just take this process down one level with the gallery and have a yearly gallery that way over time comparison between the technologies can also be made based on the judgement of editors current at that time. Commons has built itself into a valuable repository of media, for someone in 10/20/50(ok 100 as well) years time researching photography to be able see a complete gallery of images(recognised as being of merit in 2007) and compare them with images taken in 2017,2027,2057,2107 which were also recognised as being of merit. IMHO it'll be this ability that will make Commons and in particular QI unique. The other thing I see is that editors will upload and participate for a few years then move on, over time they'll come back unlike FP we dont need to turf images out and for someone to return in the future or one their decendents it'd be nice to see that the person was recognised then, and that those efforts are still valued. Unlike the words in the encyclopedia that can will change with time an image uploaded here is forever the work of that one person, yet when we upload we relinquish input into its use something we could have chosen to retain for ourselves and pass onto our decendents which would mean that its existance on Commons wouldnt be permitted until copyright expires 70 years after the photographer dies. I just think we need to look further forward than tomorrow or next week, having to reconstruct something is never as effect as having kept it in original condition in the first place. Gnangarra 15:34, 6 November 2007 (UTC)

Undo Discuss

Alvesgaspar undo my discuss template to promotion without any right at second time. I think it's illegitim, and I ask him don't play this bad game! Thanks --Beyond silence 20:28, 30 September 2007 (UTC)

• First: Please show us the diff, then we know exactly what happened. Frankly, I do not want to go through the whole history. Second: I think that Alvesgaspar acted upon his best knowledge, or rather, interpretation of the guidelines on CR. In fact, I myself brought up this exact topic here on the talk page recently, and Alvesgaspar concurred with me that the usage of {{Discuss}} within the gallery is very confusing and should not be done. So, I suggest that you use the suggested CR procedure in the future and move discussion-worthy images to the CR section instead of leaving them in the gallery. --Florian Prischl 23:34, 30 September 2007 (UTC)

I can not see what the problem is, /Discuss works perfectly well in the main review section and may be useful for those who like to make comments without actually reviewing (which is just as 'confusing'). When designing the /Discuss format we actaully had in mind doing CR inline with the other reviews which would actually make the whole procedure simpler. --Tony Wills 10:48, 4 November 2007 (UTC)

Split up candiature subpage?

The candidature list is extremely long and has all the activity on one page. I think it would make more sense to do daily log subpages - the insanely long edit history makes back-tracking stuff very difficult and we can link to QI candiatures more effectively if they have decent subpages.--Nilfanion 22:49, 11 October 2007 (UTC)

I agree, it should be like FPC, but that would mean someone would have to make changes to the bot. Dori - Talk 22:18, 12 October 2007 (UTC)

FPC is a different matter, you have every image on a seperate page. The suggestion above was to just put all the images for one day on one page. I agree about the pain of finding things in the history, but can't quite see how it would ease linking to images (unless User:Nilfanion really did want each one on a seperate page). --Tony Wills 10:53, 4 November 2007 (UTC)

Location

I think we should add the location coordinates as a requirement to images that portray a place, if the location is known (meaning for own work). It doesn't have to be the exact coordinate, just something that you can get through Google Maps and the like. It's a bit more work, but I think it adds a lot of value. If no one complains, I'll make the addition in about a week. Dori - Talk 22:16, 12 October 2007 (UTC)

• While I do not fully object the idea because I think geotagging is important, I am not sure if this is what QI wants to encourage - quality images or quality image pages? I guess both are very important (witness rejected/corrected nominations that lacked information). I support your idea, but I do not think it should be a very strong factor - a picture should not be denied QI status because it is lacking a geotag, it should only be withheld that status until the geotag is added. --Florian Prischl 16:37, 14 October 2007 (UTC)

• My feeling is that if we require species identification why not location? What makes the species of a flower more important than the location of a bridge? Dori - Talk 19:57, 14 October 2007 (UTC)

• I think it is a good idea, but I rather follow Florian's line of thought. Lycaon 20:24, 14 October 2007 (UTC)

I oppose. Geotagging is not part of an image. If you want to label something as a quality image or not a quality image, it's necessary to concentrate on the image. You can add a new quality description, as Florian suggested. Maybe it would be called "quality media" or something similar. But the decision about whether something is a quality image should be made on the basis of the visible image. And status as a quality image should not be withheld because an image lacks the information (or until it is provided).

Furthermore, geotagging might be useful for some subjects, but it's irrelevant to subjects like tabletop and studio shots.

Encourage geotagging, but do not withhold the label "quality image" for something that is not part of (and often is irrelevant to) the image.

I feel the same way about species. Fg2 20:51, 14 October 2007 (UTC)

Tabletops are not discussed here. Dori talked about places, and then geotagging is very relevant. Furthermore a picture of good technical quality of an unidentified organism has no value whatsoever. Lycaon 21:20, 14 October 2007 (UTC)

Agree with Dori and Lycaon. I see no reason for opposing this proposition except lazyness on part of the uploader. The usefulness/quality of an image on commons depends on more than just the bitmap-data/vectors, identification, categories, and location data are important to facilitate reuse. --Dschwen 23:35, 14 October 2007 (UTC)

I'm also a friend of geocoding and want it if possible, but only few cameras have a GPS-device, so for species and other nature pictures it can be difficult to know the coordinates, so I would describes the demand softly. --Kolossos 19:27, 15 October 2007 (UTC)

You don't have to own a GPS to be able to geocode. Google maps or other electronic charts or even a paper map are most of the time sufficient to find out where you where when taking that FP picture :). Lycaon 06:03, 16 October 2007 (UTC)

Exactly, and in many cases if the information is not given by the author it is lost forever. This is nothing just anybody can quickly fix (except in obvoius cases). Believe me I tried to geocode some african cities based on satellite imagery, triangulation of buildings etc. If it is even possible it is a lot of work. But it is just a small task for the author. So I'm all for strictly enforcing it. --Dschwen 13:31, 16 October 2007 (UTC)

• We don't have many requirements, and I don't think we should start. Add it as a guideline for sure, but don't try to enforce requirements without a very wide mandate. --Tony Wills 10:57, 4 November 2007 (UTC)
• I'd say we'd near clear instructions on how to obtain co-ords and then add the appropriate template first. For manmade fixed location objects maybe make it a recommendation and then reconsider it as a requirement after everybody has had time to change. I do have concerns over them being made for protected species as they would then make them easy to locate for illegal collectors. Which is serious problem for native flora in Western Australia something I do take photos of and wouldn't want to see put at risk. Also some photographers will enter restricted/protected areas to get images as such we shouldnt be asking them to identify that they did such, mind you we shouldnt be encouraging them either. Gnangarra 11:31, 4 November 2007 (UTC)
• I support adding geocoding as a guideline requirement for many types of photos, but not all. Not just a guideline, but as a requirement. A quality image is much more than just the image, it is also to associate good meta-data to the image such as a good description, categories, additions in relevant gallery/galleries and geodata. Geocoding can easily be done without having a GPS device. See Commons:Geocoding for some good advise and links to tools etc. There are types of images, where the location is not relevant and where geocoding should not be done. Examples incude:
  • Studio shots of objects and people
  • Diagrams and maps (of course)
  • Endangered species
  • Species which are rare at the location
  • Photos where privacy concerns are important. Users may want to contribute photos, where the user has an interest in staying anonymous or where the exact location should be kept as private knowledge to the contributor to protect, e.g. the subject of the photo. For instance, war scenes, catastrophies, etc.
  • Photos from otherwise restricted areas, where Commons photographers have been granted access. -- Slaunger 13:16, 6 November 2007 (UTC)
  • And there are probably other examples yet to be revealed....

It is important to state that it is the location of the photographer, which shall be geocoded. For landscapes, the heading parameter shall be used. To geocode a subject (like a muntain or a building) the location should be added to the category or gallery page of the subject.

It should be introduced smoothly. That is no declining just due to this, but instead start out with a comment: "Add geodata and I'll support". or something like that. As we get along the list of exceptions will probably grow using Common(s) sense. The guideline shall be updated gradually in accordance with this. Just my opinion. -- Slaunger 13:16, 6 November 2007 (UTC)

Changed my mind about requirements vs guideline. The exceptions should be settled before changing the guideline into requirements, if ever. -- Slaunger 16:07, 6 November 2007 (UTC)

Personal picture contests

Please help choosing my best fly (here) and bee/wasp (here) pictures. I've shoot so many that am no longer sure about their quality... The winners will be nominated to COM:FP. - Alvesgaspar 21:55, 19 October 2007 (UTC)

Glossary

Some shorthands should be clarified. QI means "Quality Images", DOF is "Depth of Field" (a technical term). What does CR mean? Luis Dantas 04:53, 20 October 2007 (UTC)

• CR is Consensual review. Regards, Ben Aveling 08:36, 20 October 2007 (UTC)

POTY - QI involvement

Hi,

We are getting close to being ready for this year's Picture of the Year competition (Commons:Picture of the Year/2007/Preparation). Help and opinions are still and always welcome.

Last year we didn't have QI, and the entries were just 'all images that became FP in 2006'. We have kept with that for this year but I'm unsure what the best thing to do for QI would be, if anything. I think it would be great to have some kind of QI category because I think it really fits in with the Wikimedian spirit and it is great that it is all just Wikimedians.

However there are definitely too many QIs to just show people a list of hundreds (there are like 350 FPs already) and expect them to have any hope of evaluating them. So I was thinking we could use the first round as "write-in" nominations - people can browse existing QIs however they like and explicitly write the name of the one they want to nominate. Then, all nominations become "finalists" in the second round [or possibly all with more than one vote... kind of hard to tell how it will go], and for the second/final round we just present them the same as FPs, show thumbnails and people just tick the boxes.

The idea is to push the final for much wider Wikimedian participation - of course everyone who is eligible is welcome to vote in the first round but for some people it might be a bit tedious. :)

Also by "eligible QI" I mean "QI that is not also FP" since FPs are just part of the regular thing.

So, what do you think of that idea? Are other good ideas for how to manage it?

Another idea might be to ask each person who has a QI to nominate one or two of their own QIs (if you only have one or two then they go in automatically). That could be interesting...

So, thoughts about this please (maybe post here, and then we can post a summary back on the POTY preparation page.) thanks --pfctdayelise (说什么?) 13:00, 30 October 2007 (UTC)

adding an individual image as the current setup is may not draw sufficient commonality to be able to get a final result, maybe the images that have been added to COM:QI#Quality_images_by_technical_merit become a short list as these have already been identified as having some technical feature of note, this may require a few editors to trawl thru the other cats and include any others. If its to be part of this years then all images to the cut off date shouldbe eligable including the early ones which were promoted prior to 1st Jan 2007. Gnangarra 10:40, 31 October 2007 (UTC)

Well... it is up to you guys. If no one shows any enthusiasm, it won't happen this time. And maybe that's fine. But if you do think it's neat then time is of the essence... --pfctdayelise (说什么?) 10:31, 2 November 2007 (UTC)

Seems kind of strange to have a QI POTY category. I think the POTYs should only be the cream of the crop. Calibas 18:11, 4 November 2007 (UTC)

Well, I was thinking as an entirely separate category. But anyway, it's pretty clear you guys aren't interested for this year at least. :) --pfctdayelise (说什么?) 12:02, 5 November 2007 (UTC)

• • Agree with Calibas. Also it makes little sense to elect 1 winner QI among more than a thousand. How do the users review them all? - Alvesgaspar 19:45, 6 November 2007 (UTC)
• Hello everyone, and sorry for my long absence. I think the idea of a "QIOTY" is a good one in general. I understand the concerns Calibas voiced and think he is at least partially right. While I think that a seperate category within POTY would be cool, it would probably look like the littler, unwanted half-brother of POTY. So maybe we should hold some kind of contest totally separate of POTY, just on the QI page itself. By that I mean a small-scale contest/vote, where the procedures, results, etc. are only discussed and published on the QI page. --Florian Prischl 16:11, 14 November 2007 (UTC)
  • The ones that win will probably end up being one of the FP images that happened to be QI as well, unless a rule is put that it must be QI but not FP. Dori - Talk 18:29, 14 November 2007 (UTC)

Scientic naming conventions

From Image:Larus_Argentatus_Juvenille_Close_Up_Eating.jpg QI review 3 Nov 2007:

• Please use correct naming conventions for scientific names of organisms. Thanks. Lycaon 00:44, 4 November 2007 (UTC).
• I dont understand, could you explain the problem please? Is there a naming convention page? Acarpentier 03:31, 4 November 2007 (UTC)
•  Info I haven't found a description of a naming convention page. Usually the format "common came (scientific name)" is sufficient. Since many of the images on Commons are of organisms, it might be good to make naming part of the guidelines. Lycaon, can you elaborate on the naming convention? - Relic38 03:50, 4 November 2007 (UTC)
• This is not a decline (or even reason to). At the moment there is some information here. I will write some appropriate guidelines for QIC and FPC. Lycaon 09:38, 4 November 2007 (UTC)

• In short: name in italics, species not capitaised, ie "Larus argentatus" not "Larus Argentatus" --Tony Wills 10:35, 4 November 2007 (UTC)

• Got it, thanks ;) Acarpentier 12:18, 4 November 2007 (UTC)

Who is allowed to review the photoes?

Who is allowed to review the photoes here and write: promotion\decline ? MathKnight 13:32, 5 November 2007 (UTC)

Every registered user but the nominator. Lycaon 14:04, 5 November 2007 (UTC)

Thank you. MathKnight 14:08, 5 November 2007 (UTC)

Evaluating the oldest images first

We have this note in the guidelines, but almost every reviewer starts from the top. Would it be an idea to reverse the date ordering such that the oldest nominations are shown at the top? This would perhaps give a more natural flow? -- Slaunger 21:42, 8 November 2007 (UTC)

I've caught myself doing that a couple of times, and started from the bottom. However I don't think it would be a good idea as it would become annoying to add new candidates (would have to scroll a lot). Dori - Talk 22:17, 8 November 2007 (UTC)

That does kinda make sense and adding new listings wouldnt be an issue. As the accumulation of older ones would probably be reduce anyway, and a link to add new could be added to avoid concerns. Gnangarra 06:52, 9 November 2007 (UTC)

At the moment old nominations that are marginal in some way naturally 'drop off the bottom' when unassessed after 8 days. The flow feels right to me. I think one reason people start reviewing 'at the top' is that is where the new (and interesting) images are, the ones further down they have already looked at and perhaps decided not to review (marginal nominations?), if we reverse the order people will just scroll down to the new entries. I might be wrong :-) --Tony Wills 08:02, 9 November 2007 (UTC)

I quite agree with Tony on this one. I'm convinced that most reviewers have looked at all submissions several times. Unassessed image are just that. they don't gather enough interest as they are just not good enough for promotion but also not that blatantly bad as to immediately be declined. The flow should IMO stay as is. Lycaon 08:07, 9 November 2007 (UTC)

I have the impression that most of the hangers are those, which are difficult to review because they maybe have more unusual subjects, compositions, etc., images that may often end up in CR. I think there is a good reason to get these images reviewed first to get a more even QI processing time. I agree, however with Dori that it would be annoying to add images at the end of the QIC page, unless the QI system is changed with for instance one subpage per day and a top level page which makes it easy to add new nominations. Yes, maybe some users will just scroll down to the end of the page to see the new ones, but they would be more aware that this is not what they are supposed to do. My impression is that many reviewers (like Dori mentioned doing above) simply forget the guideline stating that the oldest images should be reviewed first. -- Slaunger 19:27, 14 November 2007 (UTC)

A number of potential candidates awaits...

Featurable?

Featurable?

I have a somewhat unusual request. A friend has been uploading images to Commons and Serbian Wikipedia for a while, mostly plants, animals and landscapes (she is a biologist). However, not knowing about the recommendations, she was uploading them in quite a low resolution.

So, could anyone go through her galleries at Commons and at Serbian Wikipedia to identify pictures which have the potential to become QIs or featured? I would then ask her to upload them in the highest resolution. Nikola Smolenski 14:41, 10 November 2007 (UTC)

There is no way to know, actually. At that resolution almost anything can be made to look good (especially focus and noise issues tend to vanish). It is a bit strange though, IMO, that she would only upload low res versions if the hi res are available anyway. Lycaon 14:57, 10 November 2007 (UTC)

Slow Internet. Now, you are right, but people could at least check the crop, composition etc. (For example, do those berries behind snail affect the composition, does that grass at the left side of the moss looks bad?) Nikola Smolenski 18:31, 10 November 2007 (UTC)

Many of the images suffer from lighting and focus (very likely due to limitations of the camera). Some that I liked better than others:

Dori - Talk 19:08, 10 November 2007 (UTC)

• 
  Nice colours but sky probably over exposed
• 
  maybe

Maybe she should upload a couple that she thinks are good, and get some feedback about what to look for. With QI the idea is you nominate images that you think pass the QI criteria and others review your decision. For QI I would say the first criteria she should look at is that the subject of the image should be sharply in focus, if she wants to submit to FP, then some 'wow' factor should be of primary importance (although recently QI criteria have been pushed to the front).

A lot of her images are not only quite low resolution, but also quite compressed. She needs to be looking at at least 1600x1250, 1 to 2 megabytes in size. (hopefully she is taking pictures at the highest resolution and highest 'quality' (least compression) settings) --Tony Wills 22:39, 10 November 2007 (UTC)

Problem?

I realized that it was quite impossible to see the promotion vote of an accepted QI - apart if one has a patient archeologist mind! That implies a risk: anyone should be able to add the template QI on any image! On French Wikipedia, the quality article template includes a link to the vote so that anyone can check it has passed the test! Apparently it's unfortunately undone it on English Wikipedia, nor here. Couldn't we do that on Commons for QI? --TwoWings * to talk or not to talk... 17:31, 13 November 2007 (UTC)

OK no answer but I understood it myself. Thanks for your time ;-) --TwoWings * to talk or not to talk... 15:14, 18 November 2007 (UTC)

QI from non-users

I was wondering if we could create another status of QI by people who are not Wikimedia/Wikipedia users. Actually is there such a rule? --TwoWings * to talk or not to talk... 17:13, 27 November 2007 (UTC)

That rule is the first one of the guidelines. I don't think we need such a label for non-commons users. Those authors would never know (or probably care) anyway. FP's at the other hand get more visibility and can even reach POTY status (last year's actually was non-commons). Lycaon 19:15, 27 November 2007 (UTC)

Yeah I know it's in the guidelines but I just wondered why. But I didn't know pictures from non-Commons users could become FP's. So I guess that's a bit like a QI status for them... Thanks for the answer. --TwoWings * to talk or not to talk... 18:04, 28 November 2007 (UTC)