What is wrong with the perspective in this picture?

http://maps.google.com/mapdata?latitude_e6=40650000&longitude_e6=4221017296&zm=9600&w=600&h=400&cc=US&min_priority=2

and wondered what in fuck you were talking about for a second.

http://maps.google.com/maps?f=q&hl=en&q=chicago&ll=41.88112,-87.631599&spn=0.002077,0.004501&t=k

Another thing that probably plays in here (again, uneducated out-of-my-ass) is that these sorts of satellites aren't geosynchronous, right? And so even a small amount of movement is huge relative to its distance from the group -- which could account for a lot of distortions in its visual angle, especially after compositing.

Slightly?! The satellite(s) must be separated by hundreds of kilometres between each picture to get that kind of change in perspective.

Bingo.

The buildings would be leaning away from each other, not towards each other, if this were the case.

Two pictures seamed together. The helicopyer/sattelite moved inbetween shots.

and the shadows all go in the same direction...

The sun did not move as much.

I have a perspective vanishing point type of question, for all you technical drawers out there.

(also, if there's a website that can teach me about all this feel free to link instead)

Okay, imagine a box with two vanishing points. now imagine another box in the same scene but rotated around the y-axis by some random angle of say 30 degrees. I guess both boxes vanishing points will lie on the same horizon, but they won't have the same vanishing points BUT surely it will be restricted to obey the same perspective aspect as the original box. How do the two sets of vanishing points relate to each other? Is there a designers 'rule' that covers all this ? Or calculations?

Is it something as simple as having the vanishing points the same distance from each other?

If you want accuracy then you probably need to do proper full-on techical construction which involves drawing a plan and projecting from a viewpoint to a horizontal representing the picture plane and putting in heights and ... i forget the details. When I get back home I could remind myself at least enough for some handy google keywords if not coming up with a complete and lucid description myself - or you might wanna hope that someone else more capable comes along.

i found something, thanks to ledge (yes i picked out something from that).

http://www.khulsey.com/perspective_2pt.html

although doesn't mention anything about multiple vp's, i tested some drawings out on paper and using the same Station Point you could figure out the vp's. seems to look okay, not sure if it's exactly right for sure until i draw it really accurate.

yeah that's pretty much what the technique i was on about, good job! and yeah i'm pretty sure that as long as you keep the same station point and picture plane, any VPs you work out should be consistent.

some things it doesn't make clear, choice of picture plane and station point location are as arbitray as the horizon and ground line. moving the picture plane relative to the plan view only changes the scale (further away = larger); moving the station point changes the scale and perspective (further away = smaller and less xtreme perspective). also drawing the side elevation seems kinda redundant, you can just measure the heights directly into the drawing.

yeah i was wondering about those sp and plane positioning, thanks for clearing that up.

anyway, it's all pretty damn clever