do it youself HDRI?

[beaker]

yup, that is how many people do it. They explain it on the HDRshop page.
www.debevec.org
www.debevec.org/hdrshop

[Tierackk]

whoops.. i should have adde more information. I mean with a digital camera and no light probe.

[SimonReeves]

Do you think that if you were able to take several pictures of an area, stich them togeather in PS to create a pano...adjust the exposure to all the different levels in HDR Shop or PS and then convert it into a sphere or lat/long or whatever in HDR shop, this would create a decent HDRI?

i suppose you could do....
but with a digi camera, you could buy a fisheye lense, and just take photos of the sky, at different exposures and do it that way

[redfuzz]

Originally posted by Tierackk
...adjust the exposure to all the different levels in HDR Shop or PS
Unless you have taken multiple exposures of the same framing (not panning) it won't work. Adjusting the "exposure" in PS doesn't cut it, it's still only 8bit. HDRI shop will take all the exposures and add them together to give you more range, then you could stick these hdri's together in PS (you'd have to make sure you use an image format with enough color depth that PS can read and write).

g

[FClub_TDurden]

As far as Im concerned....HDRI is a little overrated in the terms of 8bit vrs Float...

Float format only gives you more details in hilites when motion blurred...otherwise you can just use a convolved 8bit map and be fine...most people dont notice the difference...

However, be sure you lock down your camera because when you add the images together in HDRShop....if one is slightly off...your final image will suk...so use a Tripod

[redfuzz]

Originally posted by FClub_TDurden
As far as Im concerned....HDRI is a little overrated in the terms of 8bit vrs Float...

Float format only gives you more details in hilites when motion blurred...otherwise you can just use a convolved 8bit map and be fine...most people dont notice the difference...


Yes, I would tend to agree, I think you and I are on the same page when it comes to all these new bells and whistles. It's a nice look to use as reference, but it should still be faked, for times sake and creative control.
Can you tell me what you mean by convolved ? For the motion blur, where white areas blur to grey, but should stay white, we often just do two render passes. I find that if you break a scene into render passes (spec, diffuce, reflection, etc) you can get so much range out of simple 8bit renders. But we have a feature project coming up where i client has insisted on at least 16bit 3d renders...oh well.

g

[FClub_TDurden]

"what is CONVOLVE?"


Basically its a fancy mathmetical blur....but here is the actual features of what it does and what its used for...

"It is useful if you need to pre-compute a diffuse or rough specular texture map.......Each pixel is replaced with the cosine-weighted integral of the corresponding hemisphere of incident illumination. In non-mathematical terms, the diffuse convolution is the image you'd want to use as an environment-map (indexed by the surface normal) to light a diffuse object in order to show it lit by a light probe. The specular convolution is the image you’d want to use as an environment map (indexed by the reflection vector) to light a rough but shiny object."

Its available in HDRshop (www.debevec.org)

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.