architecture: slight converging verticals

I have a technique question regarding converging verticals in architecture. I am in the habit of eliminating converging verticals on buildings except when I wis h to emphasize some aspect of the building/composition, in which case I leave st rong converging vertical lines. For this case,I'm mainly referring to rise rathe r than fall. However, I recently showed some poloroids I shot (no converging ver ticals) to a pro here in Japan who told me that usually, a slight hint of conver ging verticals is deliberately left in the shot since it gives the viewer a bett er feeling than a shot devoid of converging verticals. Is this common practice i n the US in architecture photography, or is this just a cultural preference in J apan?

Try to find a copy of the Michael Harris book, "Professional Architectural Photography". In the section titled "Problem Exteriors" he talks about the issue of shooting tall buildings with excessive vertical shift, which creates an apparent divergence of verticals giving the building a top heavy feeling. I've seen this in my own shooting. The solution is to balance your vertical rise with some degree of tilting off vertical (giving convergence) in these situations. Other than solving this type of problem, I'm not aware of deliberate convergence of verticals as a common practice.

I wouldn't say it's "common practice", necessarialy...more an advanced technique. But the bottom line, of course,,is what looks best..and feels right. The prvious response is entirely correct...but this will also depend very much on the lens used, the point of view of the camera relative to the building and the perspective employed. Another trick is when shooting down (especially) at a steep angle...let the verticals converge to increase the drama of the shot. This may mean employing either no corrections..or very slight corrections..depending

I've read in some of the Sinar books that you can "get away with" not fully correcting the verticals of a building in a picture if you'd have to look up or down at more than 20 degrees from the horizontal to see the top or bottom respectively if you were looking in real life. They make out that the brain expects to see some convergence / divergence due to the height etc and to have a fully corrected picture would look false / not true to life.

As others have said though, it's up to you in the end with what you feel best portrays the building or what your client / end user wants.