Cineversity

Hello
I’ve been using Photospheres on my android phone with the Camera app and uploading to Google Photos.
I’m wondering if there is a way these could be brought in as spherical maps into C4d?

Thanks!

Hi Alex,

To my knowledge, these are equirectangular images. Which means they are perfect to use in Cinema 4D while in spherical mode.

Typically you use a Sky Object or just a Sphere.
There are some little changes in the material needed to get good results.

• Check if the image is mirror sided if you are inside of the sphere.
• Set the Material’s Sampling to none. Better quality (typically), and no pole trouble.

I use this for decades by now for my art, often several in one final image, for each point of view (object wise).

Keep in mind that the resolution from a phone might not be sufficient for some parts to the object (convex works better than concave surfaces). I often work with 30K-60K images.

Let me know if you have any further question, I’m happy to help. (Perhaps my “Photography for 3D Artists” series helps, link in my signature)

All the best, stay safe, stay healthy.

Thanks Sassi!
Yes, I will take a look at that series for sure.
One question I have, and maybe this is covered in one of your videos, is if there is a way to do camera projection based on the spherical image?
For example if I wanted to recreate a 3d model of the space (here’s an example : https://photos.app.goo.gl/Re5JoJzBpXJhfa6KA), what would be the best approach? 

Thank you!

Hi Alex,

The sphere gives you from a single point all the information needed, any movement away from that point will lead to distortions. Camera mapping will prevent that. However, as we merged the 3D world into a 2D flat image, even spherical, lots of information is lost. I call that object shadow, as anything behind objects is blocked out with the object. As in Camera Projection, that needs to be replaced to be able to move around, which is a little bit tricky with spherical images, but possible.

The beauty of perfect equirectangular images is that they are typically free of distortion. That is not so much the case for phone-based images. To project anything on objects, no distortion should be present.

I have set up the file:
https://www.amazon.com/clouddrive/share/Tivqs7uLrx32pmAtZoWfpR8LihCldr3WEYXw9VijVo7
But no camera-mapping, I will try this evening to see what I can get out of it.

Cheers

Alex,
As promised, here is the image set into a Spherical Camera Mapping scene.
Scene file and a screenshot:
https://www.amazon.com/clouddrive/share/WFPqBI4WGhUJUbMPJSR44V4vf0j8MzlReVF5c2dmrip

It reminded me of working in VTour (RealViz) back in the early days, I think it came out in ~2006. That was the first one to my knowledge to create “models” with equirectangular images.

Please note that I took only a screenshot from the link you had provided for file 01, and the full version with file 02. I fixed some small stuff on the images.

The Material set to spherical works like a 360ºx180º camera.

I hope that this “sketch” will illustrate some of the problems, but also what can be done as a project “previz” for exhibitions/installations, etc.

Let me know if that worked for you.

BTW: here is a little preview clip:
https://www.amazon.com/clouddrive/share/uv0QWt6kE75k3zGw4jFbVOF5m5prHLGlMhZ5t60IWkT

Have a great weekend, stays safe, stay healthy.

P.S.:

Here is a simplified form of it. I hope it helps to understand the idea of it a little bit more. The null holds the material/Texture Tag, set to Spherical. Any child of it will show the crossing point between surface and projection.
With this, it is easy to find the edges of the room in the image. I ignore the details of the room for now.

Screen-capture and scene file:
https://www.amazon.com/clouddrive/share/zQxvYVhWMV5uyx7Jn5SFM986BozSncVh6uzcl0P0heK

Thanks for you interest in this. I hope the two techniques will help. I’m working on some other pipelines since a while. More about that later this yea, or next(?).

Cheers