Commons:Featured picture candidates/File:Rheinkniebrücke zur blauen Stunde.jpg
Jump to navigation
Jump to search
File:Rheinkniebrücke zur blauen Stunde.jpg, not featured[edit]
Voting period is over. Please don't add any new votes.Voting period ends on 12 Jul 2023 at 11:02:49 (UTC)
Visit the nomination page to add or modify image notes.
- Gallery: Commons:Featured pictures/Places/Architecture/Bridges#Germany
Info Rhine knee bridge of the German city Düsseldorf; created, uploaded and nominated by Аныл Озташ -- Anil Ö. (talk) 11:02, 3 July 2023 (UTC)
Support -- Anil Ö. (talk) 11:02, 3 July 2023 (UTC)
Comment Impressive. I would suggest to apply a slight perspective correction – right now the verticals are leaning out and their reflections are leaning in. (If you want a longer explanation, just drop me a note. It’s a pity that the perspective has not already been annotated in the QI process – such issues should be sorted at QI level.) --Aristeas (talk) 09:11, 4 July 2023 (UTC)
- Hi, thanks for your watchful eyes. I hadn't noticed it and if I understood you correctly, it should have been fixed or at least improved with the last correction. As far as the criteria regarding QI are concerned, I agree with you in the matter, but I have to note that in some cases this does not even happen to the extent assumed here - a regurarly look at the most recently listed photos can confirm this from time to time. If I have observed correctly, the QI people pay attention to image quality and other technical details such as white balance, contrast, sharpness and perspective in general and leave other subtleties to the raison d'être of the next levels. I don't know if this is correct, but it seems to be the practice there. However, please let me know if your objection has been resolved by the latest correction or if it was in the right direction. --Anil Ö. (talk) 14:52, 4 July 2023 (UTC)
- Hello Anil Ö., thank you very much for your reply. (1) With my remark about the QI process I did not want to criticize you, I just wondered why the QI reviewer did not spot the perspective issue. Traditionally perspective correction is a typical theme of QI reviews. But you are certainly right that QI reviewers have different activity centres … (2) The perspective correction is indeed difficult in this case. Normally photos with reflections on water are easy: the surface of a body of water is perfectly horizontal, hence the reflection of any point x on a body of water should be exactly below the original point x (share the same x value). But in long exposure photographs of rippled water the contours become very blurred, so it’s not that easy to find the perfect verticals. In addition, I suspect that there is a small barrel distortion caused by the lens used for this photograph (I can’t verify this because the file metadata don’t mention the lens model) which makes it even more difficult to find the perfect verticals. (3) But to show clearly what I mean, I have tried to correct the perspective correction myself. Here you find a copy of my results. My editing is based on your second version, uploaded 21:29, 29 June 2023 and labelled „minimale Farbkorrekturen“. In my version still not all verticals are perfectly vertical (e.g., not all of the street lantern posts), but I have based my correction on a study of all verticals, especially near the left and right border, and so we have to assume that some of the street lantern posts etc. are leaning in reality. I had to rotate the photo first by 1° counterclockwise, then corrected the verticals by a value of +15 and shifted the whole image +2.7 vertically. (These are the values for Adobe ACR and Adobe Lightroom. I do not know how to translate these values for Capture One, maybe somebody else can help here.) Then I had to crop the photo a little bit at the borders – it’s a pity, but every perspective correction causes some empty areas at the borders and hence requires us to crop the photo or to fill some of the “empty” areas (like parts of the sky) with cloning. By the way I have also removed two dustspots. You can use my version as an inspiration for your own editing. (You can also take my version and upload it directly over your version, if you like it.) Hope it helps, --Aristeas (talk) 09:52, 5 July 2023 (UTC)
- Hello Aristeas, I have now found the time to take a closer look at the picture. First of all, I would like to say that I really appreciate the time and effort you have put into editing my picture and the detailed explanations you have given. I have tried to adjust the perspective as shown in your example image. Unfortunately, I have to admit that my edited image still looks different from the example. Despite my efforts, it looks inconsistent to me when I adjust it even further. It seems that adjusting the perspective in this case is more challenging than I had originally thought. --Anil Ö. (talk) 15:42, 6 July 2023 (UTC)
- Hello Anil Ö., thank you very much for your reply. (1) With my remark about the QI process I did not want to criticize you, I just wondered why the QI reviewer did not spot the perspective issue. Traditionally perspective correction is a typical theme of QI reviews. But you are certainly right that QI reviewers have different activity centres … (2) The perspective correction is indeed difficult in this case. Normally photos with reflections on water are easy: the surface of a body of water is perfectly horizontal, hence the reflection of any point x on a body of water should be exactly below the original point x (share the same x value). But in long exposure photographs of rippled water the contours become very blurred, so it’s not that easy to find the perfect verticals. In addition, I suspect that there is a small barrel distortion caused by the lens used for this photograph (I can’t verify this because the file metadata don’t mention the lens model) which makes it even more difficult to find the perfect verticals. (3) But to show clearly what I mean, I have tried to correct the perspective correction myself. Here you find a copy of my results. My editing is based on your second version, uploaded 21:29, 29 June 2023 and labelled „minimale Farbkorrekturen“. In my version still not all verticals are perfectly vertical (e.g., not all of the street lantern posts), but I have based my correction on a study of all verticals, especially near the left and right border, and so we have to assume that some of the street lantern posts etc. are leaning in reality. I had to rotate the photo first by 1° counterclockwise, then corrected the verticals by a value of +15 and shifted the whole image +2.7 vertically. (These are the values for Adobe ACR and Adobe Lightroom. I do not know how to translate these values for Capture One, maybe somebody else can help here.) Then I had to crop the photo a little bit at the borders – it’s a pity, but every perspective correction causes some empty areas at the borders and hence requires us to crop the photo or to fill some of the “empty” areas (like parts of the sky) with cloning. By the way I have also removed two dustspots. You can use my version as an inspiration for your own editing. (You can also take my version and upload it directly over your version, if you like it.) Hope it helps, --Aristeas (talk) 09:52, 5 July 2023 (UTC)
- Comment I would use the snapseed program and pull down from the bottom left corner, and then slightly to the right, so that the buildings below the bridge are perfectly vertical. I think the perspective of the bridge would suit too. Snapseed is an Android program and therefore the photo must be transferred to a device that supports Android.Maybe there are other free perspective correction programs that I don't know about. I would vote for it if you can.Question: Is it allowed for someone else to edit the photo?(⧼Anna Massini alias PROPOLI87⧽) (talk) 14:30, 7 July 2023 (UTC)AnnaːMassini(⧼Anna Massini alias PROPOLI87⧽) (talk) 14:30, 7 July 2023 (UTC)
- Hmm, I don't actually see so crooked that I don't notice it... I have now tried it again and also oriented myself to the buildings under the bridge - thanks for the hint. Anil Ö. (talk) 16:05, 7 July 2023 (UTC)
- Looking at the building on the left at the end of the bridge it is not perpendicular, it is still a little crooked, as are the buildings on the right bank. I think pulling down a little far to the right, getting the water line straight would straighten everything out. It's almost there, but we're not quite there yet. This is my opinion. Sorry. (⧼Anna Massini alias PROPOLI87⧽) (talk) 19:13, 7 July 2023 (UTC)
- Hmm, I don't actually see so crooked that I don't notice it... I have now tried it again and also oriented myself to the buildings under the bridge - thanks for the hint. Anil Ö. (talk) 16:05, 7 July 2023 (UTC)
- I might have found the mistake... how does it look for you now? --Anil Ö. (talk) 19:44, 7 July 2023 (UTC)
- Comment Taking a look at the latest version I am sorry to say so but at the left the verticals are leaning out (see e.g. the street lamp posts on the bridge or the reflected lights on the water). Yes, as I said above, fixing the perspective is not easy in this case. General hint: starting with the original raw image, it’s good to follow always the following order: (1) Fix the lens distortion (barrel or pincushion distortion, Verzeichnung), if present. (2) Rotate the image, if necessary; do this with the help of verticals in the middle of the image, because in the middle the perspective distortion is negligible; when this is done correctly verticals at left and right border should be symmetrical, i.e. both leaning in or out by the same angle, as mirrored on a vertical axis in the middle of the image, like in
// | \\or\\ | //. (3) Fix the vertical perspective distortion so that you get|| | ||. Sorry and hope it helps, --Aristeas (talk) 09:51, 8 July 2023 (UTC)
- Hi, thanks for your watchful eyes. I hadn't noticed it and if I understood you correctly, it should have been fixed or at least improved with the last correction. As far as the criteria regarding QI are concerned, I agree with you in the matter, but I have to note that in some cases this does not even happen to the extent assumed here - a regurarly look at the most recently listed photos can confirm this from time to time. If I have observed correctly, the QI people pay attention to image quality and other technical details such as white balance, contrast, sharpness and perspective in general and leave other subtleties to the raison d'être of the next levels. I don't know if this is correct, but it seems to be the practice there. However, please let me know if your objection has been resolved by the latest correction or if it was in the right direction. --Anil Ö. (talk) 14:52, 4 July 2023 (UTC)
Confirmed results:
- Note to Anil Ö.: Don’t worry, this is just the regular closing of a nomination which got (for some days) no support votes. This is done automatically. You can nominate the photo again (make a /2 nomination) once you have settled the perspective question. --Aristeas (talk) 19:20, 8 July 2023 (UTC)
