User talk:Mikhail Ryazanov

From Wikimedia Commons, the free media repository
Jump to: navigation, search
Welcome to Wikimedia Commons, Mikhail Ryazanov!
Afrikaans | Alemannisch | العربية | Asturianu | Azərbaycanca | Беларуская (тарашкевіца)‎ | Български | বাংলা | Català | Čeština | Dansk | Deutsch | Deutsch (Sie-Form)‎ | Zazaki | Ελληνικά | English | Esperanto | Español | Euskara | Estremeñu | فارسی | Suomi | Français | Frysk | Galego | עברית | हिन्दी | Hrvatski | Magyar | Հայերեն | Interlingua | Bahasa Indonesia | Italiano | 日本語 | ქართული | 한국어 | Latina | Lietuvių | Македонски | മലയാളം | मराठी | Bahasa Melayu | Plattdüütsch | नेपाली | Nederlands | Norsk bokmål | Occitan | Polski | Português | Português do Brasil | Română | Русский | Scots | Slovenčina | Slovenščina | Shqip | Српски / srpski | Svenska | Kiswahili | தமிழ் | ไทย | Türkçe | Українська | Vèneto | Tiếng Việt | 中文 | 中文(简体)‎ | 中文(繁體)‎ | 中文(台灣)‎ | +/−

Drawing jpeg[edit]

La mia era solo una modifica di un file esistente di "Andy Dingley", volendo posso tranquillamente caricare una versione in PNG, ma sarebbe meglio prima parlare con "Andy Dingley" e dirgli di caricare direttamente i file di quel tipi in quel formato, anche perché dovrei fare un file nuovo e sinceramente non trovo utile avere diversi file uguali tra loro, se carico una versione PNG sarebbe da cancellare quella in versione JPEG.--A7N8X (talk) 12:54, 8 March 2011 (UTC)

File:Lange_diercke_sachsen_erdkarten_verbreitung_wichtiger_nutzpflanzen.jpg[edit]

You added a request to convert this and some other files from JPEG to PNG. Why? The result from the scanner is a JPEG file. I don’t think that a conversion to PNG with the same quality will save disk space. Or is there any other reason for your request? In my opinion PNG is more useful on graphics with larger areas of the same color opacity/transparency effects. But this is not the case with a scan.

If you want to convert it to PNG anyway, you may do so. I, as the original uploader, do not have any better “original” file as the uploaded JPEG file. Best regards! --Bananenfalter (talk) 12:12, 31 March 2011 (UTC)

D'accord. PNG is good for graphics with a clear defined number of colors and clear-cut edges. On the other hand, scans such those you've mentioned have a lot of shades and a smooth histogram - so if you don't want to remove information from the image by reducing the color space, a JPEG would be always much smaller than a PNG. With a high quality level you won't see any differences between the JPEG and a lossless format on such images. The problem with this scan is, that the source (JP2000) already had some artifacts and I don't have the orignal book to rescan it with a better quality.
PS: please note that Commons has a limit for PNG and GIF images of maximum 12.5 million pixel - which is not that much for a good scan. --Alexrk2 (talk) 16:41, 31 March 2011 (UTC)
Such drawings usually have a very limited number of colors (that one has black, pink, cyan, yellow (2 shades?), green and orange, and white background), the "smooth histogram" comes from noise in printing and scanning, which can be removed or significantly reduced by digital image processing.
12.5 Mpx limit is not a significant problem, since a smaller version for display can always be uploaded with a link to the original (as guidelines suggest). In fact, the smaller version can be uploaded over the original, with an explaining comment (this is often done to avoid Wikimedia's technical issues while keeping the original without creating superfluous filenames and links).
About your scan (BTW, is it a reply to my comment on your talk page? :-) ): from where did you get the JPEG2000 source?
Mikhail Ryazanov (talk) 01:58, 1 April 2011 (UTC)
To Bananenfalter: The scanner itself outputs raw data. If you have JPEG, it means that the scanning program was configured to save the data acquired from scanner in a JPEG file. This is generally not a good idea, since all raw scans require some sort of correction (colors, denoising, and so on), and JPEG is not suitable for editing. I mean, the final result can be saved in JPEG format, when appropriate, but the raw material should never be. (Also here.)
I've uploaded a PNG version made from your JPEG file. (Please change it's description accordingly -- I didn't understand how to put "other versions" in your template). Since the paper original had only few colors, I did "cluster analysis" with ImageMagick's segment and then optimized PNG compression with pngout. The result could be better if started from the raw image and spending little more time on better segmentation (it usually has to be done only once for the whole set of scans from one book).
Bitonal images (like that front page) just need thresholding, which can be done during the scan (usually called "Text/Lineart" mode).
Mikhail Ryazanov (talk) 01:58, 1 April 2011 (UTC)
The result looks good! :-) For an example, how to add “other versions” see here: [1]. The template does not vary that much from the normal Information-template. Only the title, year and publisher tag is static and translated by the template. Thanks for your hints on how to produce a good scan. May be if I have enaugh time, I could try to get access to the book and a scanner again to upload some better versions in PNG format. But this will not be in the next weeks. Last time I had some problems with the scanner software, at the computer where I did scan the maps. I wanted to scan with 600dpi and maximum JPEG-quality, but what I got were warnings and error messages ;-) So far Bananenfalter (talk) 12:18, 2 April 2011 (UTC)

"Bad JPG"[edit]

I see that you've tagged File:Seattle - Windermere map.jpg and some other maps I am uploading with "BadJPEG". I think this tag is wrong for these maps:

  1. On a theoretical level, these come from the Seattle City Clerk's Neighborhood Atlas, so they are significant in and of themselves. Anything we substituted would not be the maps that the city government itself uses.
  2. On a practical level, these JPEGs are available to us for the minimal effort of a quick download from the City Clerk's site and upload to Commons. Any PNG would take enormously more work. If someone wants to put in that work, fine (though it still wouldn't supplant these because of point 1), but I certainly am not going to do it.

In short, having a PNG as well would be useful, but there is no reason not to upload maps from a significant map set.

Please just respond here rather than my talk page, so that any conversation doesn't get split. - Jmabel ! talk 00:25, 26 April 2011 (UTC)

"BadJPEG" does not mean "do not upload", it means "if possible, upload something better in a lossless format". :-)
I checked the originals and see that they are of extremely poor quality. For example, I can't even read the name of the streets that go along the blue borders (like the top one, almost horizontal, in the mentioned file). Thus it would be absolutely impossible to make anything reasonable from this source, even with an effort... :-(
Therefore, the question is: are there any better sources? If not, then it might be worth trying to communicate with the authors. I can't believe that the city government uses these awful files, and the website actually says specifically: "Chris Neman developed the online version of the Neighborhood Map Atlas." Perhaps, if this Chris Neman (or his successor) still works there, we can persuade him to replace these ugly JPEGs with something more relevant (SVG, PNG, GIF -- whatever their system can make)?
Anyway, the situation should be somehow explained in the file descriptions, and the BadJPEG tags left somewhere (below the summary?), so that the replacement eventually can be done in future. Once again, "BadJPEG" is not a accusation or an invective, but a call for help. -- Mikhail Ryazanov (talk) 02:31, 26 April 2011 (UTC)
"BadJpeg", as you use it, is a clueless insult to content uploaders.
JPG & PNG are both competent graphics formats. They work. One may have advantages over the other, but this is very minor. If you had to label photographs as "BadGIF" instead, then that would be a different matter - there's a format that really does have serious limitations.
The situation here, as for all of your "BadJPEG" tags that I've seen has had nothing to do with the file format, but instead has been related to the quality of the source image and its scan. Obviously some scans are better than others, some scans are post-processed to be better than others. There are real quality issues involved, but they are not dependent on the file format and they're certainly not (as your tag claims) going to be fixed by a mechanical conversion from JPEG to PNG! You have done this conversion in the past (often to others uploads, and you've then called to have the original deleted) and your conversion has been of poorer quality than the original because you've upped the contrast so much that other artefacts have started to appear, such as jaggies.
Your continued use of this tag is disruptive. It doesn't address any real quality problem, it doesn't offer any useful solution or improvement, and it's quite rude to uploaders to label their work in this manner. Not that this stops you being even ruder at times - I still remember you describing my scans as "the worst imaginable". Andy Dingley (talk) 09:04, 26 April 2011 (UTC)

Well, we've got what we've got. I agree these are of limited use. They are certainly not effective street maps, and that is not their purpose. Their purpose is to show (approximate) locations/borders of the named neighborhoods. And, as I said, I can grab these easily off of an existing site, and we have permission to take them. I'm in no position to ask the City Clerk's office to put more work into this than they saw fit to do for their own website. If someone wanted to use these as a basis to make much cleaner maps (probably SVGs) that would be a commendable project, but a large one; I certainly am not taking it on. - Jmabel ! talk 14:16, 26 April 2011 (UTC)

Honestly, I don't understand, why you afraid to ask the City Clerk's office. My feeling is that they just did not understand the technology, and not "saw [it] fit for their own website" (I saw many such examples on the government/municipal websites).
Anyway. If you need just neighborhood borders, take a look at something like OpenStreetMap -- they allow export and support many formats (even SVG, but that would be huge unless you remove unnecessary information). If you don't want to do it yourself, then just tag all these maps with {{SVG|map}} -- and somebody else will do that eventually.
Since you agree that there are problems with the current files, and it's obvious that any such tags on the description pages do not affect the use of the files themselves, but can help to improve them, the tags should be better present. I myself do tag my own contributions when I know that they are far from perfect and don't feel any bad about that. :-) Sometimes I also look through the "inappropriate" categories and help when I can (and I'm surely not the only one ;-) ).
I'm not "afraid" to ask the City Clerk's office. I'm just careful with when I ask favors of an organization that I sometimes need to ask.
Starting from OpenStreetMap and marking up these precise borders for a hundred or so maps would be a bunch of work. If you want to do it, fine. It still wouldn't make these irrelevant, as remarked above) but it might be useful for encyclopedia articles. Again, what I'm doing now, uploading these and incorporating them in en-wiki, is a bunch of work. If you want to do even more, go for it. However, the fact that something else would also be useful doesn't make these "bad." - Jmabel ! talk 15:32, 28 April 2011 (UTC)
I think, there is a sad misunderstanding. "BadJPEG" tag does not mean that you are "bad", or that the images are "irrelevant". :-) It only tells that the format is "bad", and (quote) "If possible, please upload a PNG or SVG version of this image..." (as a clarification: it tells this not to you personally, but to everyone who could upload it). If you don't like the "Bad" part in the tag name, use {{SVG|map}} -- it will be even more specific.
And, again, I totally understand that there is a lot of work associated with preparation of better maps, and that you might not want to do it yourself. Exactly this is the reason why I ask you to put the tags -- let other people know what is needed! -- Mikhail Ryazanov 06:54, 29 April 2011 (UTC)

BadJPG in the Popular Science Monthly Illustrations (PSM)[edit]

Kindly remove the (BadJPG) from the PSM categories pages, as it is confusing and not relevant for the following reasons:

  • There may be 24 images which contain text only. Most of the images of the project are of poor original quality, have very limited use, and qualify only for placement in the Wikisource PSM project where they come from. Many "originals" are themselves copies from other journals of the time.
  • Please calculate how much time it would take to upload the images in other formats. I tested this by creating four different formats of the same image with the following results: .JPG = 214 Kb, .SVG = 335 Kb, PNG = 777 Kb and .TIF = 4,288 Kb — without any difference in the quality except .SVG which had the poorest quality.
  • I assume that the Commons cares about wasting disk space, as much as I care about the time donated to clean, categorize and upload the images. Larger files means substantially increased upload times.
  • The PSM project contains an estimated 12,000 images dating from 1872 to 1918, of which about 5,000 have been uploaded so far. The publication began in 1872 with drawings and some of them are quite complex. Photos became prevalent some 8-10 years after, but drawing were used at all times.

I was in the process of uploading an additional 170 images for Volume 51, but stopped the process until this issue is resolved and I can continue as before. Thank you. Ineuw talk page on en.ws 22:04, 26 April 2011 (UTC)

Answers to the list:
  • "Text only" should have never ever been JPEGs. The image at the upper level (File:Category_selection_bug_in_commons_upload_form.jpg) is an extreme example of inappropriate format. As for the poor quality of the original, the page says "were ... properly contrasted and cleaned". If this is true, then saving them back to JPEG wastes at least a part of the cleaning effort.
  • Your numbers are somewhat strange. Usually, for such images, PNG takes less space than JPEG at the same quality level. (TIFF actually incorporates a large variety of formats, so the given number corresponds to some really poor choice.) It seems that you just took the JPEG file and directly converted it to other formats (the note about SVG quality supports this guess). This is indeed a meaningless process. You should have saved the result of "contrasting and cleaning" directly to PNG and then make any comparisons of quality and size.
  • Commons does not care much about the file sizes. Quality is more important. (But use of PNG compression optimizers is recommended, and it will save your upload time.)
  • Photos should be stored in JPEG, because it was designed especially for that purpose. But drawings should be stored in more appropriate formats; PNG is the most reasonable for scanned drawings (unless you want to reconstruct vector versions, which usually takes some effort).
If you explain where to take the "originals", I will take a look at them and try to give some useful recommendations. -- Mikhail Ryazanov (talk) 04:18, 27 April 2011 (UTC)
(Addition) I have uploaded a PNG version of File:PSM V01 D008 Samuel Finley Breese Morse.jpg. The original was taken from the huge tar file (I suppose, these are the only "original" originals available?), run through GraphicsMagic's convert with -colorspace gray -enhance -level 40%,1,80% -colors 16 options, then cropped, rotated and optimized with PNGOUT. The PNG file size is indeed smaller than for JPEG (although not significantly in this particular case). The quality is also kind of better, partially because your JPEG was apparently extracted from a DjVu file, which had a its own lossy compression added (plus its mask part really tortures JPEG compressor).
For File:PSM V01 D061 Range of height measurements of union soldiers 1864.jpg I get about twice smaller file size in PNG (did not upload it), plus it does not suffer from JPEG artifacts when resized. -- Mikhail Ryazanov (talk) 06:53, 27 April 2011 (UTC)
  • First I want to thank you for taking the time out to answer my points clearly and in detail.
  • I am using exclusively this upload form, because it had a very useful script which erased old category selections under certain conditions. There was a software change and the script stopped working. Since my category selections were not visible in the information window, I didn't see the accumulation of the categories from previous uploads. I posted notes regarding this problem, but was never resolved. You can check back on my contributions to see these posts. — I corrected most of the bad categories but missed this File:Category_selection_bug_in_commons_upload_form.jpg, However, this is no longer a problem because I manually clear the categories with each change.
  • The image source is a zipped .JP2 file on Internet Archive HTTP downloads at http://ia600406.us.archive.org/17/items/popularsciencemo51newy/ The above is the volume 51 on which I am working. Each volume is about ~500mb's.
  • I work with Windows XP - Irfanview which batch converts and grayscale the images, but contrasting and cleaning is done manually image by image. Irfanview can convert .JP2 to .PNG as well, but the problem is that some images are photos and some are sketches - the important question is, can I process photos as PNG?
  • The image sizes of my post were the result of converting from .JPG to the other formats, but afterwards I tried direct from .JP2 to .PNG and I got the same result. I now downloaded the RIOT .PNG compression optimizer for Irfanview and studying how to use it. Thanks again for the help. Ineuw talk page on en.ws 13:50, 27 April 2011 (UTC)
(My remark about File:Category_selection_bug_in_commons_upload_form.jpg was not regarding the meaning of the file, but that screenshots should not be saved in JPEG.)
I checked both "orig_jp2" and "_jp2" archives. The "_jp2" files seem to be cropped and aligned versions of the "orig_jp2" files, but there is no noticeable loss of resolution, so they probably are the best starting points. (The digitizing process these people use is rather strange, and the inability to retrieve separate pages is quite inconvenient, but that's what we have...)
Contrasting and cleaning should be done manually indeed, but IrfanView is not the best program for that. As the name suggests, it's mostly viewer rather than editor. ;-) In particular, contrasting should be done by level adjustments looking at the histogram and the resulting image preview in full size and native resolution. I did not find how to make this in IV (if it's possible at all). Additionally, it's color reduction algorithm works in a strange way, shifting the black and white points. :-/ (Color reduction is desired to reduce the file size. Grayscale PNG in full "16M"-color representation is about twice larger than in 256-shades-of-gray, without any difference in the image itself. Reduction to 16 shades does not alter the image visually, but reduces the file size about twice more (don't use dithering!). The default output produces those most inefficient "16M"-color files, which probably explains why you're getting large PNGs.)
Since PNG is a lossless format, it does not harm to save anything in it. :-) Actually, it's even strongly recommended to save all intermediate results in something lossless like PNG. The only problem is that the file size might be quite large. However, the final result can always be converted to, for example, JPEG, when appropriate.
As a suggestion, read the related guidelines:
  • en:Wikipedia:Preparing images for upload about the formats in general. It does not cover scans directly, but clearly suggests "JPEG for photographic images, and PNG for everything else".
  • Help:Scanning. Not everything is applicable to our question (since the "scans" were done already, and not in the best way), but it's quite useful to read. At least, it discusses the illustration types and explains what to do with half-toned photographs.
  • Commons:Pearson Scott Foresman has an example of processing. (They seem to have higher resolution of the originals, thus blurring/sharping suggestion might be not directly applicable.) I can't persuade you to install GIMP, :) but you'll surely need something with the same kind of level adjustment capabilities.
Good luck! :-)
P.S. I'll try to make something out of volume 52, since you are already working on 51. -- Mikhail Ryazanov (talk) 04:25, 28 April 2011 (UTC)

Again many thanks for all this valuable info and I am embarrassed that you spent so much of your time on helping. I read the linked pages, and bookmarked them to revisit.

A word of caution about the IA volume scans. Their image numbers do not usually correspond to the actual .djvu. They make a LOT of mistakes! My usual first step is matching the image numbers to the actual .djvu numbers in Wikisource, and I don't wish you to bother with this process, as I have more time than most (I am retired). I suggest that you upload a few select images and I will rename/move them if necessary, as well as use them as a study guide. In the meantime, I will experiment with the JP2 files of Volume 51 images by converting them directly into PNG.Ineuw talk page on en.ws 06:22, 28 April 2011 (UTC)

I have uploaded all illustrations (except photographs) from Vol. 52. But without descriptions, categorizations and good names (actually, I screwed up with the names -- used 4 digits instead of 3 and even uploaded 6 files without the "PSM V52" prefix...) I hope, you can fix the names and make descriptions during proofreading. Sorry for that.
I did clean the most visible dirt from the pictures, but some still need finer cleanup. Some of the maps also got too dark cross-hatched areas during contrasting, they might need more manual work. Let me know, if there are some significant flaws -- I'll keep the original and the intermediate files for a while, so it shouldn't be difficult to revise.
Mikhail Ryazanov (talk) 11:46, 28 April 2011 (UTC)

Hi. Please don't worry about Volume 52, by now I am pretty good how to re-organize etc. — I've been uploading .PNG images to Volume 51, with the exception of the first 21 .JPG which were ready for upload and they were photos. Otherwise, they are all .PNG. I am using PngCrush for now beause it's simpler for now and studying and studying the use of GraphicsMagic. I also tried a couple of other compression software, but so far PngCrush is a good start. Can you please look at the quality and let me know if the images are OK?Ineuw talk page on en.ws 06:28, 1 May 2011 (UTC)

Sorry for a delayed response. I looked at some of the images and noticed a couple of strange things, therefore it would be very helpful if you give a step-by-step description of your processing technique. The "things" are (with examples):
  • File:PSM_V51_D026_Global_male_stature_distribution.png is too blurry. Perhaps even blurrier than the original.
  • File:PSM_V51_D041_Stature_distribution_of_the_austrian_tyrol.png has something with the background color. The outer part is white, but all enclosed areas are light gray (the map itself -- mostly noticeable in the upper right part with sparse hatching, and the letters -- in the legend).
  • File:PSM_V51_D158_Mons_meg_cannon_at_edinburgh.png is very pale. I would set the black level to a higher value, and perhaps the white also a little higher (the letters have suffered most from whitening).
  • Minor issue -- the compression is not optimal. My pngcrush compressed File:PSM_V51_D041_Stature_distribution_of_the_austrian_tyrol.png from 1224322 bytes to 987984 bytes (at default settings). pngout does not work well with large images with many colors, but after reduction to 16 colors results in 430639 bytes.
  • I'm not sure that photographs in PNG are needed. In principle, JPEG is good enough for them (and is smaller), and the originals are anyway available from IA for those who want to do any modifications. However, it depends on the purpose of all these files, and having both PNGs and JPEGs or just PNGs might be reasonable.
As for GraphicsMagic, it has a limited use here because it lacks interactivity. I mean, the batch processing with trivial operations like conversion of .jp2 to .png with -colorspace gray -enhance as the initial step and color reduction (-colors 16) as the penultimate one is very helpful, but level adjustments without visual feedback are really frustrating. I did it after examination of histograms in IrfanView and noticing that they are very similar for all the images, but nevertheless some of the results required revision... -- Mikhail Ryazanov (talk) 02:13, 3 May 2011 (UTC)

BadJPG in the Popular Science Monthly Illustrations (PSM) Part 2[edit]

Thanks for your reply and before answering in detail tomorrow, (I am exhausted) I uploaded 2 additional copies of File:PSM_V51_D026_Global_male_stature_distribution.png named:

and saved it alongside the first copy. Please check them out and let me know any one of them is OK and which is better? I already know what I did wrong. Thanks. Ineuw talk page on en.ws 05:36, 3 May 2011 (UTC)


20:02, 4 May 2011 (UTC)[edit]

Hi. Essentially, I forgot to convert the images to 16 colors. That's why the sizes were skewed. Also, Irfanview does a very good conversion as you specified for GraphicsMagic, and has a very sophisticated GUI setup for batch processing. Also, after conversion, trimming and cleanup, compression is not necessary. My major concern is that image process management into separate .PNG and .JPG became very complicated and time consuming.

I made a big mess with Volume 51 and ask you to delete all images which you think are of poor quality and I will replace them. Also, please delete the two test images of 16 and 256 color as well because I forgot to straighten them out. Thanks again. Ineuw talk page on en.ws 20:02, 4 May 2011 (UTC)

Re: Volume 52 — Renamed and added some images to Volume 52 where you can see the improved work.Ineuw talk page on en.ws 20:11, 4 May 2011 (UTC)

Removal of categories[edit]

With this edit, you removed the categories from an image you superceded. (Which, btw, the X-29 SVGs are most excellent!) Please don't remove valid categories from images, even if they are replaced by superior ones. Cheers! Huntster (t @ c) 09:20, 8 June 2011 (UTC)

Why? Commons:Categories says: "The category structure is the primary way to organize and find files on the Commons." The Superseded templates automatically put the file in that category, but what is the purpose of listing worse images together with the better versions? If somebody searches an image by browsing categories, inclusion of the superseded files is only cluttering the galleries. Moreover, leaving the superseded files categorized presents a risk that they will be reused again -- the quality is not always obvious from the tiny thumbnails, and, I suspect, not all potential users will go to the file description page to see that the file is superseded, and another file should be used instead.
Are there any guidelines about the issue? -- Mikhail Ryazanov (talk) 09:49, 8 June 2011 (UTC)
While Commons:Superseded images policy doesn't specifically mention this situation, see the talk page for a section where multiple admins state it is unacceptable to remove categories. Remember, the "superceded" status and relative "quality" in general is rather POV: While you and I may agree that an SVG file looks better than the raster equivalent, there are *many* people who vehemently disagree for whatever reason. For purely technical reasons someone may want to download a copy of a JPG rather than a SVG. Thus, all files should remain categorised so that everyone can easily find them. If quality control starts to become an issue, it is always possible to create a gallery page that showcases the best examples of images from that category.
As an addendum, you mention "leaving the superseded files categorized presents a risk that they will be reused again". It is not our job to determine what images the downstream projects use...we are simply a host of those files. For whatever reason, be it intentional or otherwise, if they choose to use a lesser-quality image for their project, that is their choice, not ours. Huntster (t @ c) 18:25, 8 June 2011 (UTC)
Thanks for pointing at the Commons talk:Superseded images policy! Most of the opinions there were kind of supporting the decategorization (and I remember that my decision to remove categories was caused by some discussion or examples...), while some convincing arguments were given against it as well. I wrote there what I think would be the best solution. — Mikhail Ryazanov (talk) 22:31, 8 June 2011 (UTC)
To repeat what I wrote there: I think you may have misunderstood my comments on your talk page. A category is to provide categorisation to *all* related images on Commons, not just the good ones. A gallery page, such as Apollo 11, can be used to showcase the best images and provide additional information.
Also, note that those superceded conversations were from 2010 and died without generating change (Incnis Mrsi was the only person to suggest that categories be removed), thus there is no consensus at this time to remove categories from superceded images. Perhaps this will change in the future, but for now, please don't do it. Huntster (t @ c) 03:09, 9 June 2011 (UTC)
I think, actually you misunderstood what I wrote there and how it is related to the discussion here.
First of all, my comments here and there are about different things, so let's do not mix them.
The opinions I mentioned can be found throughout the whole page, not just in the last section. I referred to them not to support the removal (actually, you may notice that a wrote "removal ... is not justified" there; and please read the whole comment carefully again), but to explain that other people also had such ideas. You may also notice that I did not remove categories from the superseded files after 8 June 2011, what means that I did understand you comments and are already doing your "for now, please don't do it"...
Best regards. — Mikhail Ryazanov (talk) 02:48, 10 June 2011 (UTC)
I was most drawn to your MediaWiki software solution for sorting, or perhaps selectively removing, superceded images from the category listing. It is an interesting idea, one I've not seen before, but I'm not sure I'd want to see it implemented. I still feel that the categories should hold everything, and let galleries sort out the chaff. (Frankly, I wish folks would be more careful about what they upload in the first place, but that's irrelevant to this discussion!) More realistically, I see no real possibility that such a feature will ever be implemented. For better or worse, I've seen very little change in how the categories function in all the time I've been here, despite a myriad of suggestions for improvement being put forward. I have to assume that the software devs simply have no interest in changing their functionality anymore. Huntster (t @ c) 06:19, 10 June 2011 (UTC)
Addendum: I just read the responses at the other talk page. I have no idea what changes the board is talking about, but perhaps they'll kick-start the devs into going places they previously would not. Interesting. Huntster (t @ c) 06:22, 10 June 2011 (UTC)

Обращение Гравицкого...[edit]

Спасибо за категоризацию и т.п.! Я, правда, сейчас обнаружил, что оригинал из ВП уже удалили — теперь придётся как-то объяснить происхождение файла. Как это лучше всего сделать? — Mikhail Ryazanov (talk) 05:44, 6 June 2011 (UTC)

  • Категоризация весьма важна. Без категорий никто файл среди семи миллионов не найдёт. Объяснить происхождение весьма просто: нужно дать первоначальное происхождение файла (это либо скан Главкома, либо фото Главкома из подшивки газеты в отделе редких книг ХГУ). Надо написать удалившему файл с требованием (а не просьбой) перенести полное описание файла на склад, согласно правилам для для администраторов по удалению файлов из руВП. --Vizu (talk) 06:53, 6 June 2011 (UTC)
    • Насчёт категоризации я прекрасно понимаю и всегда стараюсь делать. Просто в случае данного файла категории, поставленные в исходнике, отсутствовали в Commons, а искать подходящие существующие не было сил, поэтому я оставил его вообще без категорий, чтобы робот сделал пометки где надо, и сведущие люди обратили внимание (что, собственно, и произошло ;-) ).
    • А как найти этого удалившего? Когда я иду по бывшей ссылке, там вообще никакой информации об истории нет... — Mikhail Ryazanov (talk) 22:39, 8 June 2011 (UTC)

File:Caran-d-ache-dreyfus-supper.jpg[edit]

The "BadJPEG" template is primarily for images which have a non-photographic source (charts, diagrams, computer-generated maps) or a geometric definition (flags etc.). It is not appropriate for that image. AnonMoos (talk) 02:03, 11 July 2011 (UTC)

P.S. In any case, the image would become less usable in Wikipedia articles if it were to be replaced by a PNG, since Wikimedia software PNG thumbnailing is quite poor, and generates large files... AnonMoos (talk) 02:18, 11 July 2011 (UTC)
JPEG is not suitable for bitonal images and images with sharp edges in general. Wikimedia software can be fixed in future, and even now the guidelines suggest to upload the original in PNG and, if necessary, an additional JPEG version.
Moreover, that particular file seems to be resampled from its original resolution. Thus the upload of the original would be useful.
And please do not undo edits without a preliminary discussion. — Mikhail Ryazanov (talk) 10:50, 11 July 2011 (UTC)
It would be nice if a better scan were to be uploaded at some time in the indefinite future, but there's really nothing to be gained by converting the current image to PNG, and the file doesn't fall under any of the main types of image for which JPEG is recognized as inherently highly-unsuitable. What the "BadJPEG" tag is actually saying in that case is "If you upload a different better version of this underlying picture in future, it would probably be best to upload it in a non-JPEG format" -- but the tag says nothing relevant about the specific image file that it is added to, and is therefore inappropriate. And meanwhile, you've been adding the "BadJPEG" tag to images where it's even less appropriate (and in fact quite ridiculous), such as File:General boulanger.jpg... AnonMoos (talk) 11:17, 11 July 2011 (UTC)
You have a very strange interpretation of the BadJPEG tag and make claims that are not supported by any evidence.
As for File:General boulanger.jpg — it has pronounced JPEG artifacts, so the tag is even more appropriate than it was for some other images (which are more or less fine, but just stored in a bad format). BTW, your second revert there compels me to remind you again: "An edit war occurs when editors who disagree about the content of a page repeatedly override each other's contributions, rather than trying to resolve the disagreement by discussion. Edit warring is unconstructive and creates animosity between editors, making it harder to reach a consensus. Users who engage in edit wars risk being blocked or even banned."Mikhail Ryazanov (talk) 05:24, 14 July 2011 (UTC)
It might be best not to get too sanctimonious about Wikipedia policies (not actually Wikimedia Commons policies) when you're clearly going against consensus here.
Look, I'm actually not a great fan of JPEG files being used for inappropriate types of picture data, and I've added the "BadJPEG" template to several file pages myself (mostly flags) and uploaded File:Color Blind Envelope 4 - JPEG artefacts TEMPORARY.jpg. However, the purpose of the "BadJPEG" template is really not for obsessive image-geeking, or for complaining that a scan is not quite as high-quality as it theoretically could have been, or for indicating that a hypothetical speculative future different scan of the same picture might best be uploaded in a different format. The purpose of the "BadJPEG" template is to indicate that the underlying picture is of a basic type which is inherently not very-well suited for the JPEG file format -- usually because it's a chart, diagram, computer-generated map, flag, etc. That's why adding the tag to File:General boulanger.jpg is ridiculous. AnonMoos (talk) 07:35, 15 July 2011 (UTC)


P.S. It seems from the discussions higher up on the page that you've been warned about this several times before. Too bad you don't appear to have learned anything... AnonMoos (talk) 11:20, 11 July 2011 (UTC)

It seems that you did not read it attentively or have not learned anything from it. Please try again. — Mikhail Ryazanov (talk) 05:24, 14 July 2011 (UTC)
You know what was said about the French Royalists -- "They forget nothing and they learn nothing". Meanwhile you should immediately cease a pattern of behavior which has been found to be unconstructive and unproductive with respect to the purposes and goals of Wikimedia Commons. AnonMoos (talk) 11:10, 14 July 2011 (UTC)

Why JPEG? - reply[edit]

Hello you tagged Category:Engravings from Album du Centenaire with badJPEG. I made those 400 by photographing from an old book, which permitted enlarged images. I could not reach that enlargment by scanning. You may make them PNG or SVG, if you know how to do. Or you may learn me to do so. Greetings, --Havang(nl) (talk) 11:15, 13 July 2011 (UTC)

Raised at Commons:Village_pump#{{BadJPEG}} Andy Dingley (talk) 17:29, 13 July 2011 (UTC)
Havang(nl), it would be great if you can retake the images using a scanner instead of a photo camera. Even the cheapest scanners generally produce better images from flat surfaces, than general purpose photo cameras can (since the cameras are not really intended for the task). If you can't (or are not inclined to), please replace (or supplement) the BadJPEG tag with an explanation that these images were obtained by a photo camera in JPEG format, and that upload of better quality versions is welcome. — Mikhail Ryazanov (talk) 05:00, 14 July 2011 (UTC)
It is already evident that they are photographs, not scans. I have put an additional photo licensing on each image for that. The originals have 7 x 10 cm. I wanted large portraits. I tried scanning first, but that gave bad quality, even going up to 100 MB. After that, I tried a few photographs and that gave much better result. Looking around on internet, there were quite a number of scanned copies which all had lesser quality if enlarged to the same dimension. I will not make a new series of the 400 portraits. You may buy (as I did) the old book (available on internet) and do a better job. Calculate the time investment first. --Havang(nl) (talk) 07:02, 14 July 2011 (UTC)
I think, there was something wrong with your scanning attempts. The current pictures have pixel resolution about 900 dpi, and the real resolution is somewhat lower (there are no single-pixel resolved elements, or they were lost in compression; some blurring is noticeable, especially towards the corners — from the lens aberrations). Scanning in bitonal 1200 dpi (probably, even 600 dpi will be enough here) should have produced better results (full resolution, no aberrations and better contrast). BTW, the scan quality is not measured in Mb, :-) but the 600 dpi bitonal PNGs should be about 0.5 Mb in size, which is even smaller than your JPEGs. Could you show some examples of these unsuccessful tries? Then I might give some suggestions how to make them better. — Mikhail Ryazanov (talk) 04:34, 15 July 2011 (UTC)

I have some experience in making large format scans from small photographs see e.g. File:Flamanville 1978, des explosives détruisent la falise et l'ancienne mine de fer soumarine pour la centrale nucléaire.jpg. But for these AduC portraits the scans had much stains, much difference in background greyness, and texte from the backpage came through, even if I tried to prevent that by using a black sheet over the backside. In adding contrast to suppress those imperfections, parts of the images fainted away too and/or the stains became blacker. Therefore I abandoned the project of scanning, and afterwards, I came to the idea to try photographs. Using full sunlight gave the best results to suppress the unwanted stains. --Havang(nl) (talk) 08:48, 15 July 2011 (UTC)

Шаблон[edit]

Насколько я понимаю, вы столкнулись с миром троечников. Вы считаете, что если нет передачи тонких тонов и цветовых переходов, тогда надо обрабатывать в малом количестве цветов и заливать PNG. Вы совершенно правы. Проблема в другом: мало кто понимает эту идею и ещё меньше кто принимает её к сердцу. Философия "и так сойдёт" очень сильна. Поэтому, если я неверно разобрался, объясните мне суть дела. Кроме того, я подозреваю, что текст в шаблоне BadJPEG действительно может огорчать исходного автора, хотя он — заливатель — не виноват в том, что не силён обрабатывать графику.--PereslavlFoto (talk) 11:58, 24 August 2011 (UTC)

"bad jpeg", follow-up[edit]

Here a follow-up of our "bad JPEG" discussion and about scanning versus photographing. I bought a scanner, and scanned "Voltaire", resolution 2400, uploading in jpg (the scanner supprts jpg, tiff, with compression mode 1 on scale 1-100; using office, I made contrast maximal and I had to reduce size; using paint I converted JPG to PNG and finally I uploaded it over your png-version at File:AduC 002 Voltaire (1694-1778).png. Before I start scanning the 400 images (in the coming winter month), please your comment, advices and wishes for final formats end file extensions? --Havang(nl) (talk) 22:04, 5 November 2011 (UTC)

Looks so much sharper and cleaner! :-) Good job!
I think, the "hardware" part has no problems now, so the question is only in the "software" processing. First of all, I'd recommend to avoid the lossy JPEG altogether and scan to lossless TIFF instead. If there are no color stains, scanning directly in grayscale should be faster, take less disk space for the originals and less processing time.
Looking at the new scan, I don't think that you actually need to scan at 2400 dpi — probably 1200 would be quite enough — try different settings and use the lowest that preserves all relevant details (I mean, you don't need to reproduce the structure of the paper, but all printed lines must be more than one pixel wide). This should also make everything faster and avoid unnecessary inflation of file sizes.
And since you have decided to do such an extensive work, I really recommend to learn some more professional tools than "office" (MS Office Document imaging, I guess) and MS Paint. :-) First of all, it would be really sad to waste the scanning efforts by poor postprocessing. Second, Photoshop or GIMP are not just more suitable, but also more convenient/comfortable. There is quite usefull information in Help:Scanning and Commons:Pearson_Scott_Foresman#Whitepointing_in_GIMP (concerning the last one, unsharpening might need to be done before contrasting, and blurring might be unnecessary — it all depends on the raw image properties).
A nice side effect of well-done contrasting/sharpening/cleaning is a noticeable file size reduction. :-) If you want to reduce it even more, I can recommend reducing the number of colors to something like 16 — it usually has no effect on the image quality for high-resolution scans of drawings. (I have uploaded your new scan converted to grayscale, which led to ~40% file size reduction, and then converted to 16 colors, which additionally reduced it more than twice; check whether you see any differences.)
And please upload in the original resolution! Even though Commons still can't thumbnail large PNGs, this should be fixed in foreseeable future. While it is not fixed, I think that uploading downscaled versions (with appropriate comments) over the originals is a better idea than uploading separate JPEG files, because it keeps everything together, simplifying the usage and maintenance: as soon as the thumbnailing is fixed, the originals can be simply "revert"ed instead of fixing all links to JPEGs back to PNGs, plus it will prevent people from mistakenly making derivatives of the derived JPEGs instead of the original PNGs. And it's easier. :-) (However, in your particular case the JPEGs are already uploaded, so you might find easier just to cross-link them...)
Best wishes! — Mikhail Ryazanov (talk) 05:22, 6 November 2011 (UTC)
Thanks. My nr 1. proposal: scanner: modus professional, 16-bit grey, resolution 800 dpi, + sharpen(K); densitometertest 1x1 pixel, eyedroptest 1x1 pixel; scanzoom: image only (better result than cropping image from full page scan); image format: jpeg, compression level 1 (high quality), encoding standard, ICC-profile included. ONE after-scan image processing action: open in Offic 2010, give maximal contrast (no resizing, no cropping, no rotating) and save; reupload this JPG over the existing commons-file. Exemple of re-upload is File:AduC 001 Montesquieu (1689-1755).JPG. That's not too much work, is that good enough, keeping in mind that the original dimension is 7,5 x 10 cm.? --Havang(nl) (talk) 14:20, 6 November 2011 (UTC)
Well, if you want to go with JPEGs, this is probably the easiest way. I did not notice any quality problems with the new scan, except that 800 dpi seems to be not the native resolution of the scanner, nor a whole fraction of it (some lines show an irregular staircase structure). What's the optical resolution of your scanner? — Mikhail Ryazanov (talk) 05:54, 7 November 2011 (UTC)
Scanner specifications. --Havang(nl) (talk) 16:47, 7 November 2011 (UTC)
Strange — 4800 is divisible by 800. Maybe, these steps on the lines just come from particular irregularities of the original (since they are don't show up everywhere in the scan). Basically, go ahead. :-)
Just one remark: when I directly converted your last JPEG scan to PNG (without any processing), I've got about 3 Mb file size, and about 2 Mb in grayscale (the original JPEG was about 5 Mb). So, you may consider using PNG not only to because it's lossless (your JPEG was actually 98% quality), but also to save bandwidth. The "basic" upload form allows you to copy-paste all information from the old versions in one step (well, two: 1. file name, 2. all the rest into "Summary" fields), so it shouldn't be much more work than reuploading JPEGs. :-) But it's up to you. — Mikhail Ryazanov (talk) 04:25, 9 November 2011 (UTC)
Thanks for all lessons.--Havang(nl) (talk) 12:27, 13 November 2011 (UTC)

File:Flag of the Nordic Council.svg[edit]

Pay attention to copyright
File:Flag of the Nordic Council.svg has been marked as a possible copyright violation. Wikimedia Commons only accepts free content—that is, images and other media files that can be used by anyone, for any purpose. Traditional copyright law does not grant these freedoms, and unless noted otherwise, everything you find on the web is copyrighted and not permitted here. For details on what is acceptable, please read Commons:Licensing. You may find Commons:Copyright rules useful. You can ask questions about Commons policies in Commons:Help desk.

The file you added may soon be deleted. If you believe this file is not a copyright violation, please explain why on the file's talk page.


Afrikaans | العربية | Asturianu | Azərbaycanca | Беларуская (тарашкевіца)‎ | Български | Català | Čeština | Dansk | Deutsch | Deutsch (Sie-Form)‎ | Zazaki | Ελληνικά | English | Español | فارسی | Suomi | Français | Galego | עברית | Magyar | Հայերեն | Italiano | 日本語 | 한국어 | Македонски | മലയാളം | Malti | Plattdüütsch | Nederlands | Norsk nynorsk | Norsk bokmål | Polski | Português | Português do Brasil | Română | Русский | Slovenčina | Slovenščina | Svenska | ไทย | Türkçe | Tiếng Việt | 中文(简体)‎ | 中文(繁體)‎ | +/−

De728631 (talk) 16:11, 22 June 2013 (UTC)

File:Flag of the Nordic Council.svg[edit]

Commons-emblem-issue.svg File:Flag of the Nordic Council.svg has been listed at Commons:Deletion requests so that the community can discuss whether it should be kept or not. We would appreciate it if you could go to voice your opinion about this at its entry.

If you created this file, please note that the fact that it has been proposed for deletion does not necessarily mean that we do not value your kind contribution. It simply means that one person believes that there is some specific problem with it, such as a copyright issue.
Please remember to respond to and – if appropriate – contradict the arguments supporting deletion. Arguments which focus on the nominator will not affect the result of the nomination. Thank you!


Afrikaans | العربية | Беларуская (тарашкевіца)‎ | Català | Čeština | Dansk | Deutsch | Deutsch (Sie-Form)‎ | Zazaki | Ελληνικά | English | Esperanto | Español | Eesti | فارسی | Suomi | Français | Galego | עברית | Magyar | Bahasa Indonesia | Íslenska | Italiano | 日本語 | 한국어 | Македонски | മലയാളം | Plattdüütsch | Nederlands | Norsk nynorsk | Norsk bokmål | Occitan | Polski | Português | Português do Brasil | Română | Русский | Slovenčina | Slovenščina | Српски / srpski | Svenska | Türkçe | Tiếng Việt | 中文 | 中文(简体)‎ | 中文(繁體)‎ | +/−

INeverCry 01:07, 23 June 2013 (UTC)

Copyright status: File:Flag of the Nordic Council.svg[edit]

Български | Català | Čeština | Dansk | Deutsch | Deutsch (Sie-Form)‎ | English | فارسی | Suomi | Français | Magyar | Italiano | 日本語 | Македонски | മലയാളം | Norsk bokmål | Polski | Português | Română | Slovenščina | Svenska | +/−

Warning sign
This media may be deleted.
Thanks for uploading File:Flag of the Nordic Council.svg. I notice that the file page either doesn't contain enough information about the license or it contains contradictory information about the license, so the copyright status is unclear.

If you created this file yourself, then you must provide a valid copyright tag. For example, you can tag it with {{self|GFDL|cc-by-sa-all}} to release it under the multi-license GFDL plus Creative Commons Attribution-ShareAlike All-version license or you can tag it with {{PD-self}} to release it into the public domain. (See Commons:Copyright tags for the full list of license tags that you can use.)

If you did not create the file yourself or if it is a derivative of another work that is possibly subject to copyright protection, then you must specify where you found it (e.g. usually a link to the web page where you got it), you must provide proof that it has a license that is acceptable for Commons (e.g. usually a link to the terms of use for content from that page), and you must add an appropriate license tag. If you did not create the file yourself and the specific source and license information is not available on the web, you must obtain permission through the OTRS system and follow the procedure described there.

Note that any unsourced or improperly licensed files will be deleted one week after they have been marked as lacking proper information, as described in criteria for speedy deletion. If you have uploaded other files, please confirm that you have provided the proper information for those files, too. (You can get a list of all your uploaded files using the Gallery tool.) Thank you.

JuTa 21:04, 30 September 2013 (UTC)

mistake on US debt graph[edit]

Hello, it looks like you left 1980 off the graph [2], which makes the correlation between year and debt incorrect from 1980 on. Is this something you can fix?--Wikimedes (talk) 08:46, 8 October 2013 (UTC)

Thanks a lot for notifying me! Now it should be fine... — Mikhail Ryazanov (talk) 23:57, 8 October 2013 (UTC)
Thanks for correcting it.--Wikimedes (talk) 03:34, 9 October 2013 (UTC)

File:US patent 6987220 Holcombe.png[edit]

Commons-emblem-issue.svg File:US patent 6987220 Holcombe.png has been listed at Commons:Deletion requests so that the community can discuss whether it should be kept or not. We would appreciate it if you could go to voice your opinion about this at its entry.

If you created this file, please note that the fact that it has been proposed for deletion does not necessarily mean that we do not value your kind contribution. It simply means that one person believes that there is some specific problem with it, such as a copyright issue.
Please remember to respond to and – if appropriate – contradict the arguments supporting deletion. Arguments which focus on the nominator will not affect the result of the nomination. Thank you!


Afrikaans | العربية | Беларуская (тарашкевіца)‎ | Català | Čeština | Dansk | Deutsch | Deutsch (Sie-Form)‎ | Zazaki | Ελληνικά | English | Esperanto | Español | Eesti | فارسی | Suomi | Français | Galego | עברית | Magyar | Bahasa Indonesia | Íslenska | Italiano | 日本語 | 한국어 | Македонски | മലയാളം | Plattdüütsch | Nederlands | Norsk nynorsk | Norsk bokmål | Occitan | Polski | Português | Português do Brasil | Română | Русский | Slovenčina | Slovenščina | Српски / srpski | Svenska | Türkçe | Tiếng Việt | 中文 | 中文(简体)‎ | 中文(繁體)‎ | +/−

RJaguar3 (talk) 13:46, 20 March 2014 (UTC)