I need some clarity on this:

All those free HDRI sets...most of them are composed of 5 or 6 different files.

Up to now i have been using mostly the 3K HDR file...like the top file in the attached image.  And my work flow is usually to attach this HDR file to BB env sphere in Poser.

As a result i get final renders that are properly lit by the HDR attached to the sphere.

Occasionally i add a light or two of my own to the scene...like an indirect light and maybe a specular...but mostly i just let the scene to be iluminated globally by just the HDR image...and that works quite fine for me in most cases.

My question is...all the other files that come in the set...like the ENV HDR and the Court IBL and the court 8K...what is the purpose of those files?...must those be added also to the scene...or the sphere?...because usually i can add only one...its either one or the other...not all at the same time!...

Or i guess that a better question would be...which one of these files is the one that has to be oficially added?...or does it not matter?...i am pretty sure that each one has a purpose!...i just have to know the differece between the files.

Whic one has to be actually added to the sphere?...

Thx for your time.

Cheers.

The HDR Labs site http://www.hdrlabs.com/sibl/how_it_works.html should tell you everything you need to know about Smart IBL, and what each file is for.

This is an sIBL file set - sIBL is a generic setup that was designed to support different applications and workflows. 

sIBL is explained here - if you haven't read this yet, then please do. It should fill in a lot of blanks. If you have read it, then ask more specific questions and I'll try to explain more.

http://www.hdrlabs.com/sibl/how_it_works.html

The 3 basic ingredients of a smart IBL setup are: - a super high res panorama as camera background

• a high res HDR as specular component for reflection
• a blurred low res HDR as diffuse environment lighting

In Poser, you are using just one EnvSphere for all three purposes. You never need the blurred low res HDR for diffuse lighting, because Poser is perfectly capable of calculating the lighting from the high res HDR image. (Some applications cannot do that, and require two separate images, one for reflection and one for lighting.)

The only real choice then is between the HDR for reflections and the LDR for camera background. 

If you are not going to reveal the EnvSphere in the camera itself with any detail, then the obvious choice is to go with the HDR file. This will give the best results in lighting and reflection because it carries more data about brightness.

The other choice is to use the super high res panorama (JPG) which will give more data about sharp detail (for you to look directly at) but it will not give accurate reflections (too dim) or lighting (missing a lot of light.)

I freely use either and compensate. If I have a super high res JPG and want it for the background, I'm missing some light, so I'll do things like add an infinite (for the sun), or double its brightness with the HSV value set to 2. If I have the high res HDRI (but it's actually not very high resolution) I'll arrange to either use a very wide-angle focal length, or I won't even show it in the camera - I'll insert a backdrop image (on a scaled, one-sided square) to cover up the camera's view of the relatively low resolution HDRI.

Another technique is to use two EnvSpheres, one shifted in position from the other. You "see" one sphere directly (this uses the JPG) and the other provides most of the lighting and reflections (this uses the HDRI).

I'm using a focal length of 27 mm.

The background (where we see the EnvSphere image above the wall) is sharp.

However, the lighting is dark and reflections are dull.

But - my background is blurry. 

I start with the EnvSphere using the HDRI (as in the last image).

I duplicate my EnvSphere and call it EnvBackground. I move this copy 3 inches closer to the camera (zTran = 3 and my Poser Display Units are inches.) It does not have to be exactly 3 inches. But the point is to make as little as possible of the EnvBackground visible - just enough to see only that sphere in the camera. 

Then on the EnvBackground, I replace the HDRI with the super high resolution LDR (JPG) image.

Now I "see" the LDR, but half of the lighting is coming from the HDR.

We can do a little better.

Most people would probably not notice. But if instead of a little stage, that was a lake or ocean, we'd for sure think there is a serious problem with the reflection.

That's why I mostly just use the HDR images and a wide angle camera. 

Here is water looking towards the sun with LDR.

p.s. in hdrlab sets, text file (e.g. filename.ibl) gives sun position in UV.  easy conversion to poser xrot, yrot on inf lite (sun).  ignore zrot.

Keeping an eye on this thread!!

Just a question in passing, but I assume from BB's posts that the HDR image is smaller than the LDR in his example (or why isn't it as sharp?). An HDR image that's 8000 x 4000 should be sharp as a LDR image at 8000x4000, shouldn't it? I mean, it's the same image, just with greater luminance data in 32bit... isn't it?

Reason I'm asking is that I just realised that Carrara has a spherical camera and I can fake HDR in photoshop by stacking images and converting the stack from 8bit to 32bit. I'll probably still play with that, but it'd be nice to learn more of the niceties of HDR along the way. The idea of being able to set up a wide background scene in Carrara (including sun position, colour etc) and project it as a pseudo HDR image in Poser has some scope for play.

Yes an HDR at equal resolution will be equally sharp.

The 8K and 3K mentioned in context of the basketball sIBL set is the width of the images in question. The 8K by 4K image is much sharper than the 3K by 1.5K image.

I do not know for certain why sIBL builders think it's important to make the HDR smaller. I suspect it is because render engines that use HDR images for lighting (other than Poser) perform a more expensive diffuse convolution, and on a very large image this would be extremely slow.

Poser doesn't incur a speed penalty by using huge images, so for us it would be better if the HDR was deep and big. (Deep referring to bit depth being much more than 8 bits per pixel.)

With sIBL sets (as we find them, not as required) we are forced to choose the deep or the big - we seem rarely to be offered both in the same image.

I was reading this thread and it's quite informative. I was wondering , rather than moving the spheres so they are offset, why wouldn't you just shrink the hdr a tad so it's inside the ldr and make it invisable to camera but visable to raytracing. Wouldn't in that case you get the lighting from that sphere but the background from the ldr that is much clearer?

Yes. I tend to forget to mention latest Poser features. If you have the Visible In Camera checkbox, that's a great way to do it. The purpose of this was in fact to give us the ability to use an object (sphere or not) that we don't "see" but supplies HDR IDL and reflections.