I want to create my own HDRI for lighting purposes using a mirrored ball. I want to do this because I’m going to shoot a short video on my phone of a background, motion track it, and insert a character animation. I’ve already done all that, and now want to expand on it.
I want the lighting from the scene where I’m filming from. I understand getting the best HDRIs involve special tri-pod mounts and taking a large number of pictures at various angles. I can’t do this because of the time involved. I need to take a few quick pictures of a mirror ball and be done.
I’ve researched this before and HDR Shop could process the various images of a mirror ball into something useable in blender. HDR Shop appears to be dead now. So is there any other software out there that can handle combining the images of a mirrored ball into something that is usable in blender?
How exact does it need to be? If it’s only for some general ambience lighting I’d just project the images taken from the mirror ball on half spheres, adjust the uv maps until the environment looks correct, then divide the half balls with booleans (I used four balls as I rendered four pictures from 90 degree angles). Then flipping normals inwards and adding emission shader I’d render a 360 image from the center of the sphere. This is not very good example scene as it has those arrayed image planes that have seams between them anyways, but the fence, the ground and the sky give more clear idea about it:
In general there seems to be a lot of tools which can do this. Also interesting when not using a chromeball (but a fisheye lense):
..and actually aside form equirectangular blender can also use mirror ball projection in the environment map of he world shader.. so you might simply use a mirrorball phot directly ??
Not special. Just absolutely any kind of tripod. There is nothing special about camera tripod. It’s way easier to get than a good mirror ball as well. …and cheaper… and you get better quality images with pretty much any camera, since you take a whole bunch of them and combine them… Mirror balls are just faster and more convenient.
You can use some phone app as well even without a tripod google camera has this functionality. It might be tricky to get it for non google phones, but it seems to work fine(takes around 5 minutes to make, maybe less):
If you don’t need it to be really professional, that might be the easiest solution in my opinion. And the next one would be a propper camera and a tripod so you can shoot raw and take multiple exposures. I think a tripod would be usefull for the phone as well.. . And also you sort of need a tripod for the ball either way and then another for the camera while shooting the ball if you need hdr anyway.
If you’re photographing a mirror ball I recommend using a telephoto lens because then you’re further away and take up less space in the image. Also the photo will be closer to orthographic the longer the lens is.
@keihaankarkinen That’s not what I am looking for. I’ve seen your posts around here about the film you’re working on and I respect your knowledge and skills. Are you doing any CGI on top of real life film footage on that project?
@Okidoki yep, saw the first link you posted, not what I am looking for. That was for converting the common rectangular HDRI to a round Mirror ball image. The second link looks promising, can’t wait to check it out!!
Maybe I used the wrong terminology and/or you mis-read, but yes, nothing special about the tri-pod. It’s the mount that attaches the camera to the tri-pod that I am talking about. Maybe it’s called a panaoramic head? Changes the pivot point of the camera from the body to front of the lense, maybe??
What app did you use? I’ve used OpenCamera but it didn’t work well with my last phone, I’ll try it again on my current one!
That’s the idea behind HDR Shop, as I understand it. Take multiple photos of a mirror ball at different exposure settings, then rotate 90 degrees around the mirror ball, take multiple photos again. Combine photo set A into one HDR image, combine photo set B into one HDR image, then run them both thru HDR Shop, which would merge the images together, removing the photographer & equipment
Anyway, I’ve got some ideas from everyone’s input and I’m off to check out that Weta link posted