I'm new to Blender. I can see right now I don't really have the vocabulary much less any technique with it. Someone suggested that Blender might be or have a way to do something I wanted to to a couple years ago. I never pursued it at the time. To today I have another instance where it would be nice to alter an image to make it appear as if it were taken from say thirty feet to the left of a new image taken today. In this case the original picture was taken over a hundred years ago and any info on the camera and lens is long gone. However, the location of where it was taken is fairly obvious. I'd have taken one from the same spot today but trees have grown between then and now.

The other situation involved two new photos. One taken from one place to capture a view directly behind somewhere else. I'd like to show this background view behind a picture taken from in front of this other spot. The far background works out fine, but the details closer to where the photo was taken appear huge if I overlay it into this other picture taken from farther away. It seems like all the pixels sort of converge to the point where the lens was, so it would involve distorting or warping things as if they converged somewhere farther back.

I haven't got a clue how to do either, but since I first heard of Blender I recently read that all sort of customizations has been done to use if for all sorts of things with moving pictures even. So maybe there are some extensions, plugins, modules or something or at least some terminology for me to look into.



1 Answer 1


I’m thinking you need to convert one (or or possibly both) of the pictures into a 3D model/space.

  • Maybe the lens information is not so long gone as you think. For example, there’s a guy on YouTube who has the very first Kodak, and it still works.
  • Even if it is, if you can go to the actual place and take measurements, then construct a 3D mesh of that environment to scale, lining up the blender camera position and focal length so that the old photograph lines up with the mesh could be quite simple, especially if you have a good idea where the camera was and can place the blender camera in relatively the right position to start off.
  • After that, you need to UV unwrap the mesh based on the camera’s perspective (don’t remember exactly how to do that, but it’s simple), and use the old photograph as a texture for the mesh.
  • Then, you can adjust the camera properties to match the new photograph’s camera, but beware occluded areas becoming visible. You may have to split up the texture and interpolate missing areas (say, a portion of road behind a pole that is now visible).
  • Now, use the compositor and related settings to overlay the render onto the new photograph.

There’s an old tutorial somewhere that explains this visually, so I’ll try to find it.

  • 1
    $\begingroup$ Thanks, I'll do some searches on blender 3d model/space and mesh. $\endgroup$
    – user22240
    Jul 21, 2022 at 13:23
  • $\begingroup$ @user22240 I might rather search for “converting photos to 3-D” but good initiative. You’re welcome, PTL! $\endgroup$
    – TheLabCat
    Jul 21, 2022 at 13:51

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .