US Custom House

[xpatUSA]

I think the move to more technically correct which might appeal more to architects results in a personally unappealing top heavy effect in this photo. I know Manfred did not have much to work with in this photo as he mentioned he would have used a tilt/shift lens from the start. That said, his results give me a uneasy feeling and to my untrained eyes the columns appear to be converging under the ground. Maybe this is an optical illusion but that does not matter as it is what I see. Perhaps as I learn more about architectural photography that perspective (technically correct, acceptable to architects) may grow on me. Thanks to all responding to this post I know a lot more now than I did last week about all of this and I will be using this knowledge going forward. Thanks Terry - Best regards - Sam

Perhaps a little OT but I recall reading somewhere that ancient Greek columns are not perfectly cylindrical for whatever reason. Someone here should know more about that and whether it bears on this thread or not . . . .

. . . a bit of searching reveals that, in addition, Greek columns could have a curved taper, not just a straight taper. This surely renders theoretical consideration of perspective more complex than the normally simplistic photographic view of 'straight lines only'?

Source: http://www.pagliacco.com/col_greek_doric.html

[Dave Humphries]

While that's interesting Ted (something else I learnt today).

I think whether the columns themselves are curved, tapered or straight is 'pragmatically irrelevant' in this shot - and most shots showing a group of columns.

I say that because I'm fairly our brains are considering the vertical perspective 'distortion' by the average of angles of the group of columns.

When I suggested the shot needed a slight rotation, I was basing that on placing the centre of the foremost column against the vertical edge of my monitor. Equally, if I were correcting the distortion, I wouldn't ever go against the 'edge' of a column, I'd always go to the middle of the column, probably using one of the flutes/grooves that run up the centre. (Assuming it isn't one where they spiral)

[Manfred M]

Ted, as Dave says the shape of the column is irrelevant, what is important that the load bearing structures have to darn close to vertical to effectively carry a load. You don't want to apply a bending moment to a stone column; stone is pretty good at withstanding compression, but doesn't do all that well in tension and of course you want maximize the load bearing capacity of the column.

Any columns I've seen in buildings tend to be vertical. Notable exceptions are the Leaning Tower of Pisa (that was an oops as the soils analysis was wrong) or the minarets at the four corners of the Taj Mahal, which lean outward slightly so that should there be an earthquake, the minarets would fall away from the mausoleum and not damage it.

[Loose Canon]

When considering correcting perspective, one should find an architectural vertical (inside/outside corner of the structure rather than a picture hanging on the wall as an example) in or as close to the center of an image as possible to first determine if a rotation is necessary. The center is where closest to “true” should occur.

Sam, if you haven’t already seen it, this is a pretty good read regarding the discussions in your thread. It primarily for using a TS lens, but necessarily discusses the other points of interest of your thread.

[Sam W]

When considering correcting perspective, one should find an architectural vertical (inside/outside corner of the structure rather than a picture hanging on the wall as an example) in or as close to the center of an image as possible to first determine if a rotation is necessary. The center is where closest to “true” should occur.

Terry - I attempted to do that with my first update (seen in #18). I used the outside corner shown in the middle of the photo. I am misunderstanding something?

Thanks for pointing me to the tutorial, I had already seen it but it is good reference for this post. - Sam

[Loose Canon]

I used the outside corner shown in the middle of the photo. I am misunderstanding something?

I’m not really sure Sam. The statement was meant mostly as a general response rather than specific to your first update.

If you mean the outside upper corner right of the word “House” and on the cantilevered overhang then yes you would have missed something because in the original shot you posted that corner is closer to the left third. Though the crop brought it close to center afterwards.

If you mean the outside corner with the column on the building proper then that is close to center for initial evaluation.

If the original shot was cropped before you posted it then things might be different still.

[xpatUSA]

Ted, as Dave says the shape of the column is irrelevant, what is important that the load bearing structures have to darn close to vertical to effectively carry a load. You don't want to apply a bending moment to a stone column; stone is pretty good at withstanding compression, but doesn't do all that well in tension and of course you want maximize the load bearing capacity of the column.

Any columns I've seen in buildings tend to be vertical. Notable exceptions are the Leaning Tower of Pisa (that was an oops as the soils analysis was wrong) or the minarets at the four corners of the Taj Mahal, which lean outward slightly so that should there be an earthquake, the minarets would fall away from the mausoleum and not damage it.

Manfred,

Thank you for Construction 101

Please read on . . .

Dave, I did say my post was a little OT. I was seeing the column as (to the eye) an area bounded by two vertical lines. Although the columns themselves are vertical as Manfred said, the sides are not necessarily so.

I wondered what the reason for that was and whether that reason is relevant or not to how, say, the vertical corners of a building might be properly or more aesthetically shown in 2D. The base to height ratio did seem relevant to the ancient Greeks and I wonder now about depiction of the Empire state building versus the Pentagon . .

But I don't even remember what the convention is in that respect, so I probably need to re-visit WikiPedia . . .