Basic Comping of a Vray render

[schmyvol_eye]

hey cpnichols,

thats more than enough, thanks a mill for the link! I'll try this today and let you guys know if it works or not.

cheers,
y

[cpnichols]

hey cpnichols,

thats more than enough, thanks a mill for the link! I'll try this today and let you guys know if it works or not.

cheers,
y

I looked into a little last night.. but did not try to with my elements. Looks like the latest Windows version is 0.16, and the latest overall version is 0.19. Version 0.16 does not read EXR files so you may need to do it with HDRs. Also it looks like it does a true add which is a bonus.

Chris

[schmyvol_eye]

...so you may need to do it with HDRs

hi cp, thanks for the info. but what do you mean with the above statement? Must I choose to save HDR files in the VFB rollout? Sorry but I'm not too steady when it comes to hdr, g-buffers, tone mapping etc.

but this is why I'm here!

cheers,
y

[cpnichols]

...so you may need to do it with HDRs

hi cp, thanks for the info. but what do you mean with the above statement? Must I choose to save HDR files in the VFB rollout? Sorry but I'm not too steady when it comes to hdr, g-buffers, tone mapping etc.

but this is why I'm here!

cheers,
y

yeap.. you will need to do a full floating point unclamped hdri

[schmyvol_eye]

okay, i will do just that then... i'll post results when I have them.

thanks a mill!
cheers,
y

[schmyvol_eye]

just for an update, unfortunately I havent been able to test this yet due to the workload I'm currently facing, but as soon as I know I'll share my findings!

cheers,
y

[powerandrubber]

Has anyone else tried this film gimp aka cinepaint program? I'm getting nothing but crash and burn.

[powerandrubber]

Looking into it further the windows version is "broken" and won't be fixed until the next version comes out targeted for march 1. -From cinepaint.org.

[cpnichols]

Looking into it further the windows version is "broken" and won't be fixed until the next version comes out targeted for march 1. -From cinepaint.org.

Yeah I saw that... I use film gimp on linux as a cheap crappy version of photoshop. It is a pain to use... just as a warning.

[rpc212]

Something I don't understand is why we are comping floating point images, at least in this tutorial, as opposed to regular 8 bit images (per channel that is)?

I realise there is more info available in HDR images, but what is the point? To get better, more accurate lighting?

How does one adjust shadows just by themselves without influencing anything else in the comped image?

Thanks,

[rpc212]

When saving the EXR, use the raw pixel option. That should give you the full float unclamped color.

In Max 8 is the Use RealPixel RGB Data the same as raw pixel option when saving an Open EXR? This question was asked earlier, but didn't get an answer.

Thanks,

[powerandrubber]

My understanding is that you can't comp non floating point images, the results come out dark like the examples posted that were done in PS. As Chris mentioned, "Yeap... pretty sure the photoshop clamps everything which is why it is looking darker."

[cpnichols]

Pretty simple...

First consider that clamped goes from 0 to 1, and unclampled goes beyond that

If a pixel in your Raw Lighting pass has a value of 2, and a diffuse pass has a value of 0.25 when you multiply them together you sould get 0.5. If it is clampled the lighting pas would be 1 and after you multiply it the color will be 0.25 (darker)... If you rendered it without splitting the passes it would be 0.5 clampled or unclamped, so it would not matter. Compositing is all about math. If you don't have the full range of numbers, you can't composite.

[cpnichols]

So I just saw some demo video of the new After Effects 7.0... Have not used AE in a while, buit it seems that they have some cool new tools that would basically help out all the stuff you need to do this type of comping

- Full 32 bit support
- OpenEXR support

Plus a few other cool tools (more like it's about time), such as proper graph editors, better retimer, etc...

http://www.adobe.com/products/aftere...wfeatures.html

Anyway, it may be worth looking into.

[ness_andrew]

Can't figure out tiny problem - composite g-buffer renders





Hey there. I've nearly figured it out, but my composite is still slightly different from my real_rgb render?
I used rpf type of output all in linear multiply with unchecked clamp output from vray. I have checked to see if my lighting and gi elements are correct and they are, so where am I going wrong?
Thanks
Andrew Ness