Author Topic: Real World Diffuse, Specular, NormalMap and Light  (Read 29837 times)

2016-06-20, 22:05:31
Reply #30

dubcat

  • Active Users
  • **
  • Posts: 425
  • ฅ^•ﻌ•^ฅ meow
    • View Profile
I wasn't 100% pleased with the normal maps, and it was really hard to get both albedo and specular without a Fast Polarization Modulator.

Micro-detail was the first thing I thought about when I saw this on History Channel. They could see crazy details on the Egyptian/Greek artifacts. Cops used this method to see old handwriting that was indented in paper.

I will mount 4/8 lights inside the box and use the "fixed position" method. Fast Polarization Modulator is out of my price range, so I will have to do with only albedo.
It's time to convert the ghetto box into the sexy box!
             ___
    _] [__|OO|
   (____|___|     https://www.twitch.tv/dubca7 / https://soundcloud.com/dubca7 ( ͡° ͜ʖ ͡°) choo choo

2016-06-20, 22:33:15
Reply #31

Juraj

  • Active Users
  • **
  • Posts: 4813
    • View Profile
    • studio website
But was the issue in the nature of the setup or just the way you had it ?

I want to build this in studio room, fixed, nothing portable. So the flash lighting would be of presumably good quality, and imagery will be taken with Nikon D800.

What makes the Modulator superior ? ( overall and for speculars)

The method above would need fixing the bellow ball part for nicely tiled texture, and I would like something more automatic. Or would you just place the ball(s) elsewhere than middle ?
Please follow my new Instagram for latest projects, tips&tricks, short video tutorials and free models
Behance  Probably best updated portfolio of my work
lysfaere.com Please check the new stuff!

2016-06-20, 23:18:07
Reply #32

dubcat

  • Active Users
  • **
  • Posts: 425
  • ฅ^•ﻌ•^ฅ meow
    • View Profile
I want to build this in studio room, fixed, nothing portable.
That changes everything. You only need a modulator if it's portable.
You have to rotate the filter on the camera 90 degree, and this is really hard to do on the portable version, since it's inside a box. With a modulator you just hit the on/off switch.
For your stationary build I would recommend getting some linear polarization sheets, don't even bother with circular polarizers. You take the first picture, turn the circular polarizer on your camera, and everything is unaligned. Just a little wind breeze is enough to ruin the 100% specular picture. You have to use manual focus too, and the circular polarizers messes everything up when you turn them, damn circular polarizers.

But was the issue in the nature of the setup or just the way you had it ?
It was because I used the other normalmap method where you take a picture from each side.

The method above would need fixing the bellow ball part for nicely tiled texture.
The ball is only needed if you don't have any fixed lights, so you can calculate the position based on the highlight.
Since the lights will be fixed on the rig, I only have to tell the program where the lights are (x,y,z). This is done with a .txt file.

I guess you want to build something like this and just turn the ceiling lights off.



« Last Edit: 2016-06-21, 02:53:36 by dubcat »
             ___
    _] [__|OO|
   (____|___|     https://www.twitch.tv/dubca7 / https://soundcloud.com/dubca7 ( ͡° ͜ʖ ͡°) choo choo

2016-06-20, 23:27:01
Reply #33

Juraj

  • Active Users
  • **
  • Posts: 4813
    • View Profile
    • studio website
Yeah, I want pretty much replica of that :- ).

But you got me intrigued with the specular addition.

Do you have any idea/guess how do the Quixel guys process their roughness maps (from what?) ?

One last question, bit of topic, but touching tangentially to what is on first page, does from your observance Corona desaturate Albedo with specular correctly to real-life ? I guess it does, just wondering, since when I needed to use lower specularity for fabrics of low glossiness, I had to likewise desaturate the diffuse otherwise it would saturate too much at angles that otherwise wouldn't. Since the glossiness curve will be remapped, this won't be an issue anymore soon, but is the specular overlayed in way that we can really use the correctly saturated captured albedo in diffuse slot ?

The textures I see in PBR game papers, I always brighter and saturated that even I use, and I think I kind of try to match them well (although still by eye....).
« Last Edit: 2016-06-20, 23:40:32 by Juraj_Talcik »
Please follow my new Instagram for latest projects, tips&tricks, short video tutorials and free models
Behance  Probably best updated portfolio of my work
lysfaere.com Please check the new stuff!

2016-06-20, 23:48:06
Reply #34

dubcat

  • Active Users
  • **
  • Posts: 425
  • ฅ^•ﻌ•^ฅ meow
    • View Profile
I read somewhere that people are generating roughness maps based on pixel intensity from light sources. I bet it's something like that.
They kinda look like 100% specular pictures, the concrete maps have small white stones here and there.
All the Quixel maps are Highpassed with 256px, that's why all the maps are 127 RGB.

EDIT to your edit :P:
I can take some albedo/spec pictures tomorrow and compare.
I want to analyze the new "The Reflector" OSL shader from Rens too.
« Last Edit: 2016-06-20, 23:53:10 by dubcat »
             ___
    _] [__|OO|
   (____|___|     https://www.twitch.tv/dubca7 / https://soundcloud.com/dubca7 ( ͡° ͜ʖ ͡°) choo choo

2016-06-20, 23:57:32
Reply #35

Juraj

  • Active Users
  • **
  • Posts: 4813
    • View Profile
    • studio website
That does sound plausible, at least for non-metals.

Does his shader work under Corona too ? (well since OSL is universal, though I didn't follow if Corona allows it). You want to test his fake rough fresnel :- ) ?

Btw, saw this super toy at EUE from one guy, could replace your mirror ball :- ) https://theta360.com/en/ I for sure am getting one soon.
Please follow my new Instagram for latest projects, tips&tricks, short video tutorials and free models
Behance  Probably best updated portfolio of my work
lysfaere.com Please check the new stuff!

2016-06-21, 00:04:28
Reply #36

dubcat

  • Active Users
  • **
  • Posts: 425
  • ฅ^•ﻌ•^ฅ meow
    • View Profile
You want to test his fake rough fresnel :- ) ?
Yes :P It looks like the sexy PBR shader Blender people use. I need to do the tests in Mr. V :(

Hm, Theta was affordable too.
             ___
    _] [__|OO|
   (____|___|     https://www.twitch.tv/dubca7 / https://soundcloud.com/dubca7 ( ͡° ͜ʖ ͡°) choo choo

2016-06-23, 09:33:24
Reply #37

JakubCech

  • Active Users
  • **
  • Posts: 126
  • jakubcech.net
    • View Profile
    • jakubcech
Lots of useful information, my passport is on its ways so will be able to make my own tests :)
I also ordered these chrome balls http://www.ebay.co.uk/itm/3-Size-Stainless-Steel-Mirror-Sphere-Hollow-Ball-Home-Garden-Ornament-Decoration/221801673314?_trksid=p2054502.c100229.m3211&_trkparms=aid%3D111001%26algo%3DREC.SEED%26ao%3D1%26asc%3D37319%26meid%3Dcbb625eecb8a42bba8432b13f0ec23c4%26pid%3D100229%26rk%3D1%26rkt%3D8%26sd%3D221801673314.
Lets see how they come up, its 18cm diameter for 6 euros :D

Snooker ball method seems okay, but more elaborate than 4 angles flash, comparison would be awesome.

2016-06-23, 14:37:31
Reply #38

maru

  • Corona Team
  • Active Users
  • ****
  • Posts: 13631
  • Marcin
    • View Profile
Looks like some fun content is coming! Can't wait to see some examples and try some ideas myself.
I still have some huge bearing balls somewhere, so I might try with those, or use something else...
Marcin Miodek | chaos-corona.com
3D Support Team Lead - Corona | contact us

2016-06-27, 11:21:43
Reply #39

vazonov11

  • Users
  • *
  • Posts: 1
    • View Profile
    • купить столбики
Looks like some fun content is coming! Can't wait to see some examples and try some ideas myself.
I still have some huge bearing balls somewhere, so I might try with those, or use something else...
and how, we have waited?