Need suggestion on render to texture alpha depth issue

hyunxi

Hello, guys. I understand similar problem has been discussed before, but I my problem is a little different. And I think this is a very common issue. Maybe someone has already know a way of dealing with this problem.

I am creating LOD for high-density overlaid alpha planes. I created the low-mesh however I can not create the texture from render to texture. As render to texture only record the first layer of alpha / diffuse information.

I search on the forum, and see similar problem of creating foliage billboard. Render from ontography camera is the way to go.  However this method is not suitable for my case, as the target mesh is more 3d.

Anyone anytoughts?

Eric Chadwick

Distort both into a semi-flat shape, apply planar UVs, use the ortho camera method to bake, then remove the distortion.

3ds Max's modifier stack is perfect for this, just select both high and low meshes, and add a FFD modifier.

This IS 3ds Max, right? (it's best to put your software in the title, see http://polycount.com/discussion/63361/information-about-polycount-new-member-introductions/p1#use)

hyunxi

Eric Chadwick said:

Distort both into a semi-flat shape, apply planar UVs, use the ortho camera method to bake, then remove the distortion.

3ds Max's modifier stack is perfect for this, just select both high and low meshes, and add a FFD modifier.

This IS 3ds Max, right? (it's best to put your software in the title, see http://polycount.com/discussion/63361/information-about-polycount-new-member-introductions/p1#use)

Thank you for the suggestion.
I was think of using bending method to make it flat and render from orthography camera but unfortunately this won't work. Because I have other cylinder  / complicated shape (think the case of hair) models with the same problem. I am trying to look for a general way of solving this problem instead of just work for this specific case. 
I was also thinking of putting 3ds max in the tile , the reason I didn't is that I think people might have some ideas using other software rather than just limited to 3DS max. Also I believe the alpha depth problem is very common in many software's render to texture function (or what ever this function is called in that software).

poopipe

It is (or was) possible using mental ray in max, and is probably possible using Arnold in maya.

Weirdly, simplygon seems to be able to do it - I've not tested it for quality though as I suddenly developed other priorities 

You'll have to do a fair bit of mucking around to make it work.  It might be worth looking up the thread I started on the subject a while back

hyunxi

poopipe said:

It is (or was) possible using mental ray in max, and is probably possible using Arnold in maya.

Weirdly, simplygon seems to be able to do it - I've not tested it for quality though as I suddenly developed other priorities 

You'll have to do a fair bit of mucking around to make it work.  It might be worth looking up the thread I started on the subject a while back

I tested in Simplygon. It can not deliver a correct result. The same result as I did it in the Max. Simplygon / Max can  render correctly if there is only one layer alpha plane. If there are many layers of overlapping planes, both software can only recognize alpha information on the first layer of plane. 

I tried mental-ray in Max, it works on 1 bit alpha but not gradient.