How is this shifting vertical perspective effect created?

Asked 4/3/2016

1 views

2 answers

0

I’ve seen images where the viewpoint seems to rise as you look upward through the frame, so the lower part looks like a normal street-level view but higher sections reveal more rooftops and courtyards. How is this effect made? Is it done with a special camera setting, or is it created later in software such as Photoshop?

Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0

Photography Stack Exchange contributor

10y ago

2 Answers

14

This picture is a composition of slices of many photos taken from different heights, likely from a drone. The key to understanding it is to examine the camera's perspective or point of view from different horizontal slices of the composited image.

  • For approximately the bottom quarter of the image, the camera is below the roof lines of the buildings at the lower left and right sides of the image.
  • Starting at about the 3rd or 4th building at the lower left (just past the 2 white cars parked on the side of the street), the camera can see the top of the buildings, and very quickly, is seeing over the buildings into the courtyards.
  • As you scan higher in the image, the camera's perspective shifts higher and higher, seeing more of the tops of buildings. Note, however, that you always can see the side of all buildings facing the camera. Thus, the camera is never directly above the buildings as you go higher.

Effectively, if you "number" the pixel rows from 1 to N starting from the bottom, the pixel row is correlated with the altitude of the camera (not necessarily a linear relation, though).

As far as camera settings, the camera is just taking a sequence of images every so often (perhaps once to several times per second), while the drone flies vertically). This is probably programmed into an intervalometer attached to / controlling the camera. The intervalometer might be a feature of the camera though.

To produce this effect, the images must be combined in post-processing using Photoshop, or another such image editor.


Alternately, it might be the panorama feature of the camera, rather than holding the camera (or mobile phone) vertically and rotating horizontally, the camera is mounted horizontally and moved vertically, again via drone. If that is the case, the camera is performing the automatic "stitching".

Either way, this is a very interesting effect, a bit reminiscent of the street-folding scene in Inception.


Edit: The author of the OP's image, sky.guliveris.lt (a graphics / animation / video / aerial filming services company in Lithuania), has a clip of drone video on Shutterstock.com showing the same town square of the old town section of Vilnius, Lithuania. Based on the footage, I believe the drone also had to fly down the street to incorporate the top portions of the image; they are too far away in the drone footage to have any detail in the composited image if the drone only took off vertically.


Update: In 2017, PetaPixel posted a couple articles regarding these types of images, including a breakdown of how to create them:

Originally by user11924. Source · Licensed CC BY-SA 4.0

user11924

10y ago

0

AI Answer

Generated from our catalog & community — verify before relying on it.

This effect is not a single-camera setting. It’s made by combining multiple photos taken from different heights, often by raising the camera progressively—commonly with a drone.

The final image is built from horizontal slices of those frames, so the viewpoint appears to climb as you move up the picture. In examples like this, people have reported using roughly 18–20 source images. The process is similar in spirit to panorama stitching, but instead of only blending side-to-side overlap, the software/compositing also blends changes in camera height.

So the workflow is generally:

  1. Capture a sequence of photos from increasing heights while keeping the scene aligned as consistently as possible.
  2. Composite or stitch them in software.
  3. Refine the blend manually if needed in Photoshop or similar editing tools.

In short: no special in-camera setting creates it directly; it’s mainly a multi-image composite technique done in post-processing.

UniqueBot

AI

10y ago

Your Answer