Commons:Featured picture candidates/File:Ottawa - ON - Library of Parliament.jpg

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

File:Ottawa - ON - Library of Parliament.jpg, not featured[edit]

Voting period is over. Please don't add any new votes.Voting period ends on 3 Jun 2012 at 14:03:06 (UTC)
Visit the nomination page to add or modify image notes.

Ottawa: Library of Parliament
  •  Info all by Wladyslaw. This is a renomination of Commons:Featured picture candidates/Image:Ottawa - ON - Library of Parliament.jpg from 01/2010. The criticized point (noise) was remedied. -- Wladyslaw (talk) 14:03, 25 May 2012 (UTC)[reply]
  •  Support -- Wladyslaw (talk) 14:03, 25 May 2012 (UTC)[reply]
  •  Oppose I love the colourful brickwork and the picture is detailed and well exposed. But it is distorted. Compare to File:BDP2006.jpg, a similar picture taken just a bit further to the right. The conical turret on the left is warped. The weather vane on the roof with the N/S/E/W compass ring is angled such that it doesn't appear to be parallel to the ground. I suspect that unless this can be photographed from further away, you might have to put up with some degree of perspective distortion because one can't fully correct for the fact that much of the building is shot from close-to and from-below. Colin (talk) 17:51, 25 May 2012 (UTC)[reply]
    • First: I wonder why you compare this picture to File:BDP2006.jpg which is obviously not perpendicular. Second: the weather vane will never be parallel to the ground except for you make a picture directly from the sky over the vane. Third: this building is not photograph-able from further away without distortion lamps and other stuff. You can check it all in maps.google. All in all I can not see a distracting distortion. --Wladyslaw (talk) 18:01, 25 May 2012 (UTC)[reply]
      • Here's what I understand about the issue: I'm still learning... Perspective causes points further from the camera to be smaller and converge. This happens in three dimensions. With a rectilinear lens held perpendicular to the subject, the lens will correct for the fact that objects towards the edges really are further away, by stretching them. When all objects are far away (like the horizon in a landscape) this difference in stretching is zero. But for a nearby building, the difference is clear. This is typically not a problem for a normal angle-of-view lens but the stretching becomes obvious as one approaches the limit of about 110°. If one uses stitching software to compose an image from several frames, some of which are shot at an angle, then the software is able to correct the foreshortening to give the impression that the camera really was perpendicular to the object. Regardless of whether it is done in software or lens, it cannot correct for the shrinkage in the z-axis (depth) (without generating a 3D model). And it cannot correct for the change in viewing angle towards the edge (being below the weather vane rather than perpendicular to it). This isn't usually a problem for the flat face of a building if the camera central and facing the building. But if one can see the sides of towers, turrets or through a circular weather vane, then the fact that the depth and angle aren't corrected becomes very obvious. The turret is warped and no longer looks like it has a horizontal base. The weather vane no longer has an elliptical shape one would accept but has been stretched vertically to a circle. The eye no longer accepts that the loop is parallel to the ground. There is no solution to this problem other than to get further away from the building and/or higher up -- in other words, to reduce the overall angle of view. A compromise is to accept some degree of converging verticals. Perspective is natural. An "architecturally correct" view is only "correct" if the viewpoint is far enough from the subject that top, bottom, left, right and centre are all nearly the same distance from the lens. Attempts to simulate this view with a camera close to the subject will generally only be completely successful for flat objects. If you are using a program like Hugin, you can move the rectilinear projection vertically, and hopefully achieve a natural (rather than architecturally correct) degree of converging verticals, while at the same time minimising the stretching distortions. If your horizontal angle of view is also large, then it may be difficult to achieve any image with minimal distortions or some kind. -- Colin (talk) 11:38, 26 May 2012 (UTC)[reply]
  •  weak oppose At the moment I think something is not ok with geometry of the picture. Even if I could accept a bit of distorsion on the far right, you have to correct some perspective.--Telemaque MySon (talk) 06:55, 26 May 2012 (UTC)[reply]
Confirmed results:
Result: 1 support, 2 oppose, 0 neutral → not featured. /George Chernilevsky talk 18:35, 31 May 2012 (UTC)[reply]
Retrieved from "https://commons.wikimedia.org/w/index.php?title=Commons:Featured_picture_candidates/File:Ottawa_-_ON_-_Library_of_Parliament.jpg&oldid=71905747"