We use Camera Map Per Pixel a lot when we add 3D products to an AI generated image. Bij adding this to a curved backplane we can render the product into the AI generated image with schadows and reflections.
Submitted on AHA Ideas (VANTAGE-I-113).
Hello,
Instead of using the Camera Map Per Pixel have you tried with the Camera map modifier? It should give you the same or very close results: https://drive.google.com/uc?id=1dgWNAtZHNO9xNpmAKhDp197kO2k_wYya
Submitted on AHA Ideas.
Unfortunately it doens’t do the same. The left image is with camera pixel and VRay GPY, the right one is the camera map modifier with Vantage. Only the backplate seems to be correct.
Submitted on AHA Ideas.
Hmm, this looks strange. Could you please send me the scene or a similar one that reproduces this issue so I can take a look?
Submitted on AHA Ideas.
Below the link to the demo scene I used for testing this. This scene has now only the camera map per pixel on it.
Submitted on AHA Ideas.
Thank you for the scene! I have managed to make it work with Camera Map modifier. The key part is to have the objects that will be used for the projection to have a good amount of polycount. I used Subdivide with a relatively low value of 10cm to give me enough geometry so that the floor boards are not distorted:https://drive.google.com/uc?id=1sjFzNcXjKnBzFZhEy1-Tmz7hChJyQUpM
Submitted on AHA Ideas.
Thank you for helping out and I’ve indead got it working now with adding subdivisions. It is a bit of a workaround and it add some unnecessary complexity but it works. Again thanks!
Submitted on AHA Ideas.
AHA Ideas status: Pending.
Set during AHA Ideas to Discourse migration.