Scanning foliage

swaterik

Hi

I am curious of how to make translucency maps like in quixels library. . Having done some Photogrammetry myself, I tried to make some foliage assets with bluescreen. When scanning vegetation, how do one produce a translucency map? I also wonder how they make their normal maps for their assets. Any one have any answers to this?

Dash-POWER

https://www.allegorithmic.com/blog/your-smartphone-material-scanner

This could be useful if you are scanning single small foliages. If you are using photogrammetry maybe thickness map could generate similar results.

kurt_hectic

Put a really strong light under it, compare in PS with the original photo, difference will be your translucency mask. Not even close to proper scanning, but should work.

swaterik

Dash-POWER said:

https://www.allegorithmic.com/blog/your-smartphone-material-scanner

This could be useful if you are scanning single small foliages. If you are using photogrammetry maybe thickness map could generate similar results.

Thx I guess that would work. Interesting blog post.

m4dcow

I don't know how quixel does their translucency masks but making a lightbox to backlight the foliage and doing a difference of some sort might get you in the right direction,  it also makes generating opacity masks a whole lot easier.

I'm not sure how they do their normals either, but you should check out a program called dabarti capture it is quite cpu intensive, but produces awesome normal and height/depth maps.

Dash-POWER

m4dcow said:

I don't know how quixel does their translucency masks but making a lightbox to backlight the foliage and doing a difference of some sort might get you in the right direction,  it also makes generating opacity masks a whole lot easier.

I'm not sure how they do their normals either, but you should check out a program called dabarti capture it is quite cpu intensive, but produces awesome normal and height/depth maps.

As i know quixel is using large wooden box with led lights on each side. The technique is almost same as Allegorithmic has with paperbox. About translucency don't know but thank you for the Dabarti tool. Didn't know about this. Looks easier then Allegorithmic solution because you don't care about direct light position