Hi,

I've posted the following to the Poser Technical forum, but someone suggested I might get a better response if I post it here too. So here goes ...

I am attempting to create some Poser room 'figures' but am coming across some problems with the Poser render engine that are proving frustrating. I wonder if anyone has any advice.

1. To overcome the problems shading flat faces - for example walls - I have split the vertices on my model. However, this leads to curved surfaces requiring lots of polygons to make them appear smooth (increasing the size of the model).

2. Because Poser uses infinite parallel lights I have disabled shadows for the room figure, however this decreases realism when lighting a scene. I could remove a wall, but this becomes harder when you have coving, skirting boards, dado rails and windows all 'crossing' or forming part of a wall.

3. I have found that often when rotating a scene the bounding boxes do not appear to exactly match the location of the objects they represent. Is this normal? Also polygon clipping seems very poor, both in the editor and when rendering.

4. When rendering, large polygons are not clipped correctly leading to gaps in the model. The only way to overcome this seems to be to split the polygons to make them smaller (see render).

5. It appears to be difficult to light this type of model using the Poser lights - do most people find this?

6. Because of the size of walls, etc I use repeating textures to keep texture sizes manageable. This leads to most features in the room being assigned a separate texture map. While this is more flexible, it can lead to longer set up times and goes against the 'norm' of a single texture for a single figure. Is this acceptable for marketplace items and will this annoy people?

All in all, I believe you can get much better results using other 3D packages (see my gallery for a Cinema 4D render of the same room model), but Poser is a big market and people seem to love buying stuff for it - so I am trying to optimise things as much as I can to get good results. What do people think of the Poser model so far?

Finally, the comments above apply to my experiments using Poser 4. Would I be able to get better results using Poser 5?