I am making some tests now to verify this, but I see some IBL experts online now so I thought I'd alert you now.  I've recently made a couple of dozen IBL lights and corresponding panoramic sphere maps.  As a test I made a compass image, just colors and letters for cardinal directions and assembled a LatLong and Probe in the same way I made the scenes in Terragen. 

It appears that the light object Poser is using does not properly translate the directional images.  That is, East shines light from the East, West shines light from the West, Up shines light from Up, and Down shines light from Down.  However, North shines light from the South and South shines light from the North.  

The light is inside out.

Attached Link: HDRI IBL in Poser using LDRI with Nodes by bagginsbill

Using an 'angular map' with a yellow centre (representing front), and a red perimeter (representing back), with the camera looking towards the -Z direction. A cube placed in front of the camera had a yellow front (side facing +Z direction) and a red back (side facing -Z direction).

There is an interesting tutorial by bagginsbill about HDRI panoramic and IBL, over at RDNA (linked above).

Apply the IBL Probe, and also unwrap it to make a panoramic LatLong, and apply that as a spherical environment.  The sphere when aligned to show the correct directs ought to show North in front of you, East on your left, and so on. 

The light will not match this.  North and South will be reversed. 

I'm not sure which direction is north in Poser, the choice seems arbitrary to me, but front and back are not arbitrary (they are defined by the cameras). Everything seems to work correctly with respect to front-back, left-right, and up-down, and that's all that really matters to me.

I edited my last post to include a link, you may find the thread interesting.

Not sure what you mean by "unwrapped your Angular Map" it is to be applied directly to the IBL. As I understand it the IBL is designed to take an angular map, not other formats.

"Why is the lighting backwards?" Not sure what you mean, here is what I get, posted above, which seems the right way round to me.

I have thought some more about this. Really I suppose it depends on how the image for the IBL was made.

By "figure" in the following, I am talking about a hypothetical figure in a Poser scene at 0 0 0 facing the Front Camera, that can also be imagined in the original scene where the photos were taken, for the purpose of visualising directions.

It seems to me that If the image was made by the mirrored-ball method, then the centre of the image would be pointing away from the (hypothetical) figure, and towards the photographer. The sides of the image would be the background, depicting  the area behind the figure. This would be correct in respect to how Poser handles an IBL map.

With the "Vertical Cross Cube Format" I think it would depend on which direction the key image (the one in the centre of the cross) was taken from. If the key image was taken pointing away from the figure, then you get approximately the same type of mapping as in the mirrored-ball. If the key image was taken pointing towards the figure, then you have the reverse situation and the mapping will be wrong for Poser. I suppose which way you should be facing when you take the key image could be debated, but as the light on the front of the figure is more important than the light on the back, and as the key image is going to have better resolution than the other images (in the final angular map). I would argue that the key image should be taken pointing away from the figure. [edit] In other words if the background is in the North, the key image should be taken pointing South, if the background is in the South the key image should be taken pointing North. That's my theory any way. I'm no expert, in fact I'm very new to this stuff, just trying to work it out as I go along.

OK, I think I see. That's what the original image would be if it was taken as a panoramic photo, yes?

P.S. I added a small edit to the end of my last post.

Yes

Try to set your image as a panoramic background, and use the probe as your IBL.  I'm beginning to think that Poser's IBL is inside out. 

Add colors for East and West  so you can see what I mean.  (try it with the probe and latlong I posted earlier.)  You can line up two directions but not four. 

(I think.  Double checking now.)  I think what you do is create a panorama with North at center.  This is the panoramic background.  Then create another one with south at center.  From this one, make your angular map.  Then mirror that -- quick tests so far look right.  Try this angular map with the corresponding latlong posted below.

It's late here (U.K.) now, about 3am, I have to go to bed before I drop. But I'll try your map tomorrow, and get back to you. One parting shot though, in the past I have worked as per bagginsbill's tutorial, and everything seemed to come out the right way round.

Here is the final render:  

I think I know what'#s causing this for you, muralist, and it's actually not the fault of P7.

The IBL maps used by P7 (and possibly most others - I don't know, since I haven't done an exhaustive survey) are captured using a mirrored ball. This means that, whilst you may be facing north when you take it, the front part of the IBL is actually the south, since it is the reflection of the environment around you. Thus, if you want to use the probe map as an envrienment as well, then you need to do the rotation and reconversion as you did, realigning the mirrored south with the 'true' south.

I don't know if that makes sense just through words, but I'm running late at the moment and dont' have time to fire up poser and show you what I mean.

JonTheCelt

That makes sense.  Now that I've got the right transformation, I'm going to animate a flyaround and test it.

Phew, that went allrigth!

The last time we had this discussion, we almost had a cath-fight!

Bopper.

http://www.youtube.com/watch?v=BPrBxac0v6I

I've uploaded the animations to http://www.veoh.com/channels/3dArt