Wikipedia:Featured picture candidates/Radcliffe Camera, Oxford

Radcliffe Camera, Oxford

This is a mosaic stitched panorama I took of the Radcliffe Camera in Oxford on the weekend. It is one again extremely high resolution (3137x4605) and detailed, clear and quite an interesting view of a beautiful building. There is an existing FP of this building, but I feel that this FPC is far superior in many ways (resolution, exposure) and the old FP would be a good candidate for de-listing should this one be supported (and perhaps even if it isn't - the existing FP is below resolution standards at 500x667).

• Nominate and support. - Diliff | (Talk) (Contribs) 06:57, 17 October 2006 (UTC)[reply]
• Strong Support. Amazing quality, good encyclopedic value. But is that tilt I see?NauticaShades 07:40, 17 October 2006 (UTC)[reply]
  • I don't think so. I didn't take the photo from exactly straight-on so I think it is a perspective illusion rather than tilt. All of the columns/windows seem to be straight, or at least they're tilting on either side by the roughly same degree due to the perspective. Diliff | (Talk) (Contribs) 08:29, 17 October 2006 (UTC)[reply]
    • You're probably right. NauticaShades 10:34, 17 October 2006 (UTC)[reply]
• Comment I don't know if it is an illusion or not but the building seems deformed, as if the shot were made with a wide angle lens. The effect is not very pleasant for me. -- Alvesgaspar 08:45, 17 October 2006 (UTC)[reply]
  • Well it must be an illusion, as the building is not deformed. ;-) Due to the distance between the viewpoint and the subject, a wide(ish) angle of view is necessary to fit the subject comfortably into the frame. This is unavoidable for the subject. This 'deformity' (it is commonly called perspective distortion) is worse (although different) if you take a photo from ground level. The difference is that we are more used to seeing things from ground level so the distortion we see such photos is generally disregarded. I think this FPC image is a more interesting angle, though. Diliff | (Talk) (Contribs) 09:57, 17 October 2006 (UTC)[reply]
• Oppose Incredibly sharp and I like the lighting, however I agree with Nautica, there is a fairly obvious lean. Also agree with Alvesgaspar about the deformity (especially in comparison to the existing FP) - similar to the Roman baths shot, I'm sure this could have been taken as a single shot, avoiding the distortion and still uploaded at a sufficient res (12 megapixel) to more than justify FPC requirements. Just as a side note - any idea why this section is so much sharper than this bit? --Fir0002 10:19, 17 October 2006 (UTC)[reply]
  • First of all, the lean is not obvious: it's most likely an optical illusion. Secondly, I really don't see the distortion you are talking about. NauticaShades 10:36, 17 October 2006 (UTC)[reply]
  • There isn't a major difference in sharpness between those two crops that I can see, but it could be explained by a slight focus difference between frames (was set to autofocus, not manual) or perhaps slight blur (it was handheld) but it is reasonably negligible. As for the distortion, the difference between the roman baths image and this image is that there would be no significant difference between taking it with a single shot as opposed to a panorama as they were both stitched with rectilinear perspective. The only distortion in the image is the same distortion you would get with a normal rectilinear lens. The roman baths were stitched with a spherical perspective, meaning the same sort of distortion as a fish-eye lens. In comparing to the existing FP, I can't see any significant difference in the disortion (but of course the existing FP is so small it is difficult to see much detail at all). To summarise, this image has all the same attributes as a single rectilinear photo in terms of perspective distortion so that argument is kinda moot. As for the lean, can you demonstrate and quantify it? As I mentioned previously, if there is an actual lean it is pretty negligible. Any perceived lean is more likely to be due to the photo not being taken directly in front of the doors (so that the doors do not line up with the windows or spire exactly). If there is a lean, it is probably less than 0.5% - likely less than other architectural images that have passed FPC. Is this really a reason to oppose? Diliff | (Talk) (Contribs) 10:46, 17 October 2006 (UTC)[reply]
    • No I agree the difference in sharpness is neglible I was just interested. My main concern was the tilt/distortion factor. If you're not using Firefox, get it, and then open your image and the existing image in two tabs and flick b/w the two and I think you'll see what I mean by distortion. The existing one looks fairly natural (to me anyway) and your one seems to bulge down the middle/bottom. As for lean I calculate it to be at least 1 degree horizontally and a lot vertically. Vertical lean could be just perspective but still gives the image an unbalanced look - which is definetly not an attribute diserable for an architectural image! Key points in the example - I think that the center of the door in the bottom should be in alignment with the spire at the top - note center verticle line. The bulge at the base can be see on the left and right lines. Line at base of dome shows horizontal lean. --Fir0002 11:03, 17 October 2006 (UTC)[reply]
      • Well, the crop of the window is not a good basis for comparison because as I mentioned previously, it is not in the centre of the view of the building and as the building is circular, of course it is going to tilt as it is somewhat rotated around. Thats like saying a disc's circumference isn't horizontal when viewed slightly from above. As for the other image, I can see what may be a very slight tilt to the right - I'll re-stitch it tonight (or just rotate the existing image) to see if it will improve things. You are right, vertical lean is perspective and that will be the case with ANY image that is photographed at anything remotely wide angle due to basic laws of physics - any object viewed from angle will result in parallel lines diverging. This can be corrected when stitching by setting the centre point at the horizon but it will then distort other parts of the image as it bends those lines straight. You simply cannot avoid distortion of some kind in photography - you can only minimise it by taking photos with a long telephoto lens at a large distance (not very practical!). I still maintain that it is a pretty petty reason to oppose though, but you're entitled to your opinion ;-). Diliff | (Talk) (Contribs) 11:26, 17 October 2006 (UTC)[reply]
        • OK fair enough, I not pretending to understand all physics behind stitching panos etc, but the fact remains is that the existing FP manages to get a near perfect straight line from spire to center of door. Perhaps you could have moved position a little? Anyway thanks for agreeing to entitle me with my opinion ;-) --Fir0002 22:24, 17 October 2006 (UTC)[reply]
• Support. Note that the picture hasn't been taken from directly in front (compare the sides). This could account for the apparent horizontal lean on the centre window. Either way it's not a big deal. ed g2s • talk 11:14, 17 October 2006 (UTC)[reply]
• Oppose Neutral The shooting position is obviously the origin of the apparent distortion of the building and other geometrical problems. I understand that, no matter what solution was adopted (a series of shots or a single shot with a wide angle lens), the result would be basically the same. So, we can't blame the photographer for his choices, except the position of the camera (but I suspect that nothing can be done about that...). But aesthetical considerations, although subjective, are also relevant. And I think the building looks ugly in this picture. -- Alvesgaspar 12:44, 17 October 2006 (UTC)[reply]
  • I would say that "ugliness" is quite subjective, and I also do not see what you are talking about at all: I just don't see how the Radcliffe looks ugly in this circumstance. NauticaShades 15:14, 17 October 2006 (UTC)[reply]
    • Neither do I. I don't think this image looks significantly different (aside from quality) than the original FP and it received glowing reviews. Obviously standards have increased a little since then but this is one of the landmark sites of Oxford and I think it looks pretty impressive from any angle, but viewing it from above ground level allows a better view of its position relative to the city. In the end, beauty is in the eye of the beholder, but at least it seems the majority can still appreciate this one. Diliff | (Talk) (Contribs) 15:38, 17 October 2006 (UTC)[reply]
  • Not really ugly, I must confess (I've changed my vote to "neutral"). The problem is most of your pictures are superb!... --- Alvesgaspar 16:20, 17 October 2006 (UTC)[reply]
• Comment - Apparentely, there is some freedom in choosing the camera position. At least, it is possible to center the door! Also, it seems that the existing FP was taken from a slightly lower position -- Alvesgaspar 23:05, 17 October 2006 (UTC)[reply]
  • Correct, it was possible to center the door, and I thought I was pretty centered at the time. ;-) The problem was that it was a VERY old tower and there was just barely enough room for people to squeeze past me, so I was fairly limited in my ability to take my time and line everything up perfectly. I am happy to admit this image isn't geometrically aligned perfectly, but I still believe that the pros outweigh the cons ;). Diliff | (Talk) (Contribs) 06:57, 18 October 2006 (UTC)[reply]
• Support Wow, what a lot of talk! I'll just cut through it and say I like it, even with a few non-worrying leans - Adrian Pingstone 13:57, 18 October 2006 (UTC)[reply]
• Support.
  chat} 17:32, 18 October 2006 (UTC)[reply
  ]
• Support - it certainly is better than the one it is replacing. If someone else wants to get a diliff quality picture that doesn't have a one degree slant, I would support that one in this one's place.--Niro5 19:55, 18 October 2006 (UTC)[reply]
• Support. As far as I can tell the distortion is minimal, and may have been unavoidable. --Tewy 00:50, 20 October 2006 (UTC)[reply]
• Support - As far as i know, this photo is good. Arad 18:56, 21 October 2006 (UTC)[reply]
• Support – very nice picture. If the lean is really a problem, all it would take is a slight rotation and crop to make the edges straight again; I could do it in about 3 1/2 seconds if needed. — Editor at Large^(speak) 17:49, 25 October 2006 (UTC)[reply]