Hi! I am working with a cgi pano that I create with 3d tool blender. In this example, I use a room with a checker texture. This is rendered in blender with a panoramic camera in type "Equirectangular" I use the rendered pano in pano2vr (6.1.5 pro) and then create a image-patch, wich later should be changed to video. The settings can be seen in the attached screenshots.
After extracting the patch, I change the camera in blender from panoramic to perspective and change the resolution to the same size of the patch. (in this case, 208 x 208 px) I dont change the position or rotaition of the camera. Then I load the patch into the cameras view (Front) by adding it at the dialog "Background and Images".
I set the FoV value to 35 degrees, according to the FoV-value of the patch in pano2vr. I would expect or lets say wish that the patch-image now seamless overlays with the whole picture, but it does not.
So I like to ask if there are more information available about the setting of the virtual camera in pano2vr, wich is actually taking the shot of the patch, so that i can translate it to the camera in blender. Maybe there are special values of lense distortion i can u, or maybe i need to set values of the sensor size other than the default 32mm? I changed the values a lot to try to get the exact overlay, but I was not lucky. I hope you can understand my description and the screenshots attached to this post are helpfull. If something is missing to get the whole picture, just let me know.
Thank you for any hints and ideas!
Translate camera specs from patch creation to camera settings in blender
-
LichtUndSchatten
- Posts: 58
- Joined: Wed Sep 08, 2010 5:13 pm
- Location: Germany
- Contact:
- Attachments
-
- viewing-parameters.jpg (42.56 KiB) Viewed 1659 times
-
- pano-properties.jpg (31.88 KiB) Viewed 1659 times
-
- pach-in-cam-backgroundimage-blender.jpg (130.24 KiB) Viewed 1659 times
-
- my-patch.jpg (123.08 KiB) Viewed 1659 times
-
Rene Falk Thomasius
- Posts: 4
- Joined: Tue Mar 24, 2020 12:05 pm
- Location: germany
Hi, that's the heart of the poodle. 
You need to be able to transfer the lens focal length used in your panorama rendering 1:1 into PanoVR and you need to be able to adjust the camera height correctly.
This means that the camera height from the 3d scene must match the "virtual camera height" where you create the patch.
You would need an import/export function for the camera data from the 3d tool you are using.
That would be a good start as feature in PanoVR.
So you need to know how the patch (and with what distortion) is projected into the room.
Then you can put it on top of each other with pixel accuracy.
gretnx Rene
You need to be able to transfer the lens focal length used in your panorama rendering 1:1 into PanoVR and you need to be able to adjust the camera height correctly.
This means that the camera height from the 3d scene must match the "virtual camera height" where you create the patch.
You would need an import/export function for the camera data from the 3d tool you are using.
That would be a good start as feature in PanoVR.
So you need to know how the patch (and with what distortion) is projected into the room.
Then you can put it on top of each other with pixel accuracy.
gretnx Rene
-
LichtUndSchatten
- Posts: 58
- Joined: Wed Sep 08, 2010 5:13 pm
- Location: Germany
- Contact:
One more aspect might be interesting. The vertical location of the camera in the room. The camera in my examble stays at 2 m above ground level. If pano2vr uses a default value of the camera location, lets say 1.6 m above zero, then the images will not overlay. Is there a default value? If so, can I change this value in pano2vr?
-
LichtUndSchatten
- Posts: 58
- Joined: Wed Sep 08, 2010 5:13 pm
- Location: Germany
- Contact:
I finally solved it. I had to do this:
1. Change the location of the cameras in the 3d-space to 1.6 m above ground-level
2. Render the complete background-pano again, in the correct aspect ration of 2:1 (it was a full HD in my first try)
3. Create the patches agin.
4. Rotate the camera that records the animation. Here, I use the mode xyz Euler. I use the pan-value of the patch on the Z-rotaion of the cam and the tilt-value +90 for the X-rotaiton of the cam.
1. Change the location of the cameras in the 3d-space to 1.6 m above ground-level
2. Render the complete background-pano again, in the correct aspect ration of 2:1 (it was a full HD in my first try)
3. Create the patches agin.
4. Rotate the camera that records the animation. Here, I use the mode xyz Euler. I use the pan-value of the patch on the Z-rotaion of the cam and the tilt-value +90 for the X-rotaiton of the cam.