Beginner post-processing/Gbuffer question

[skogzort]

Hello,
I know nothing about compositing or post-processing, but I know how to use Photoshop. What I would like to do is to be able to separate out the “shading” in my image from the diffuse color. I used the Gbuffer to export 3 tga files:

diffuse only

RawGI

and RGB.


I opened all 3 files into a layered PS7 document. What I was wanting to do was to somehow combine the diffuse and RawGI layers to create an image that looks like the RGB layer. I tried multiply, hardlight, luminosity etc., but I can not make it work. Is what I’m trying to do possible? How? If not, what would you do with a “diffuse only” channel or a "RawGI" channel? What are some common uses for the other channels? Are there any post-processing tutorials around?

Thanks much for any information

[vlado]

The diffuse channel and the raw light/GI channels need to be multiplied. Some examples of how channels are combined can be found in the Help index:

http://www.spot3d.com/vray/help/VRay...buffer.htm#ex2

All in all, I have found Photoshop not very convenient for compositing work. Corel Photopaint seems better suited for this task, but it would be best to use a real compositing application like Digital Fusion.

Best regards,
Vlado

[skogzort]

Thank you for the response - I read the page, but I want to make sure I understand the formula correctly: If there is no self-illumination, reflection, or refraction to consider (as in my example), then could you say that: GI + Direct = real RGB?

If so then I think my problem is with the way Photoshop “adds” channels. When I use Photoshop to “multiply” I get expected results. For example: if I multiply the Raw GI layer by the Diffuse layer the result IS equal to the “GI only” channel, just like the online manual says it should be. However, when I “add”

GI only:

and Direct:

the result:

is not “real RGB” like I would expect.

Has anyone else been able to use Photoshop to “combine” Gbuffer channels in a way consistent with the examples in the Vray online manual? Does anyone know how Digital Fusion handles the “addition” of two channels – maybe I can mimic it somehow in Photoshop.
Thanks much for the info

[CCS]

I believe I read once upon a time that 'hard light' is closest to a true add. Anyone?

[Seraph135]

Linear Dodge is additive.

[trixian]

Just to comment on the photoshop aspects of this thread.
Hard light is not like add, as it also affects light pixels behind dark ones, while add only adds to their brightness. In photoshop, for some reason beyond common sense, you can only add 2 images together using the calculations feature in the image menu afaik. (this has been the way they have done it since the beginning I think).

[Ramsay]

Just figured out how to do this with combustion, im now a very happy bunny

[rpc212]

Hey Ramsay,

Is there a tutorial you looked up to find out how to composite the g-buffer channels from Vray or just figured it out on your own? If you know of any tuts that would be great if you would share the links or any advice you have about using Vray and Combustion together.

Thanks,

[cpnichols]

There is no add in photoshop... it sucks. Having two screens is approximatly equals to an add. Not perfect, but it can help you get there.

[panthon]

HI !

Ramsay, i´ve had a hard time trying to get all the vray channels working in combustion in order to get the same result as u get with the RGB one, so if u dont mind sharing with us your workflow i would be very pleased!

Thanks in advance!

[rpc212]

Ya lets here the workflow Ramsay!

[Ramsay]

Oh just spotted this... I dont have time today but I will try and get something together tomorrow for you guys

[egz]

There is no add in photoshop... it sucks. Having two screens is approximatly equals to an add. Not perfect, but it can help you get there.

Linear dodge, which was added in Photoshop CS, does the same as Add in After Effects, Fusion etc.