Compositing and Post Work - What is it, and why should I care ?

JoeMamma2000

Yeah, I'm not sure. Maybe it's the terminology, or I've just had too many beers, but I'm at a loss on this one.

Maybe someone else can chime it.

ncamp

BC Rice said:

Yeah, I'm not sure. It could be that they all function the same way, but I thought only PNG only rendered the primary image (the alpha). 

When you render in PNG, you're only choosing to render the figure, ya know? Like, not the black background. Maybe you can do that with those other file formats as well? By rendering just the actual rendered image of the figure, you're not forced to manipulate layer styles for blending/compositing, until you get into actual lighting, color matching/correcting and ambient effects. 

If you have the file format of the render room set to PNG, the alpha is controlled by the Render Alpha Channel checkbox in the file format section.  If it is not checked, it renders the background.  If it is checked, it makes the background transparent. 

BC Rice

ncamp said:

BC Rice said:

Yeah, I'm not sure. It could be that they all function the same way, but I thought only PNG only rendered the primary image (the alpha). 

When you render in PNG, you're only choosing to render the figure, ya know? Like, not the black background. Maybe you can do that with those other file formats as well? By rendering just the actual rendered image of the figure, you're not forced to manipulate layer styles for blending/compositing, until you get into actual lighting, color matching/correcting and ambient effects. 

If you have the file format of the render room set to PNG, the alpha is controlled by the Render Alpha Channel checkbox in the file format section.  If it is not checked, it renders the background.  If it is checked, it makes the background transparent. 

Right. So is that will other image formats as well, or just PNG? I was under the impression that it was just PNG, but maybe these other image formats do it as well out of Carrara?  

ncamp

BC Rice said:

ncamp said:

BC Rice said:

Yeah, I'm not sure. It could be that they all function the same way, but I thought only PNG only rendered the primary image (the alpha). 

When you render in PNG, you're only choosing to render the figure, ya know? Like, not the black background. Maybe you can do that with those other file formats as well? By rendering just the actual rendered image of the figure, you're not forced to manipulate layer styles for blending/compositing, until you get into actual lighting, color matching/correcting and ambient effects. 

If you have the file format of the render room set to PNG, the alpha is controlled by the Render Alpha Channel checkbox in the file format section.  If it is not checked, it renders the background.  If it is checked, it makes the background transparent. 

Right. So is that will other image formats as well, or just PNG? I was under the impression that it was just PNG, but maybe these other image formats do it as well out of Carrara? 

I believe that it is true of any format that supports transparency.  I know Tiff and PNG do.  BMP does not.

ncamp 

JoeMamma2000

FifthElement said:

Just a quick note here, as far as I know Carrara does not have real 32 bit output regardless of the format it saves in, it's same "trick" as "ChangeDepth" node in Fusion 

I was just skimming thru this thread looking for something, and I noticed this post, and thought I would clarify.

I think 5thElement is confusing a couple of concepts here, which is easy to do with this subject.

There are two different concepts with digital images, which sound alike but are very different. One is "color space", and the other is "bit depth". And each is described as "xx bits".

Color space describes the "channels" (more like "containers") of color and alpha information. For example, with a Carrara image, you need channels to hold the R, G, B and alpha information. That's called RGBA color space, and it's what you see when you open an image in Photoshop under "channels". And each "channel" is just a grayscale image.

The "bit depth", however, is how many bits are used to describe each pixel of the grayscale image in each of the channels. So for a Carrara image, which is 8 bits per pixel in each channel (256 values per pixel), you have 8 bits per channel times 4 channels (RGBA), which is a total of 32 bits.

So a Carrara image is a 32 bit color space image, with 8 bits per pixel of "bit depth" in each channel. Other types of images have more than 8 bits per pixel in each channel.

So in the context of the discussion we were having, you need a 32 bit color space RGBA image in order to include the alpha channel. 24 bits for the RGB channels, plus another 8 for the A channel.

Easy to get confused, but I hope this helps to clarify.

evilproducer

It helps a great deal Joe.

O.T. Dammit! My left mouse button is crapping out.

FifthElement

 

FifthElement said:

Just a quick note here, as far as I know Carrara does not have real 32 bit per channel output regardless of the format it saves in, it's same "trick" as "ChangeDepth" node in Fusion  

Lol Joe, there you go, I fixed it, I was not talking about color space BTW, and I dont get easily confused, unless i'm drunk or sleepy 

Diomede

There is a plugin for Daz Studio to add some multipass functionality to renders in that program - currently $20 or so

http://www.daz3d.com/mask-and-multipass-toolbox

Here is a tutorial

 

Dartanbeck

This is a cool demonstration by GKDantas - getting SSS via Multipass and Photoshop Elements

cdordoni

LuxCore can also do multipass:

SphericLabs, the developer of LuxCore has given the following info  http://www.daz3d.com/forums/discussion/comment/912192/#Comment_912192

if you go to that link, it has the old build of LuxCore, the new build is here https://app.box.com/s/2g03wx4j6y3lyw7uzhxgq6ifnk3hnxga

You will need to put something like the following in the config section:

batch.periodicsave=15
film.outputs.1.type=RGB
film.outputs.1.filename=image.exr

This means every 15 seconds write an RGB image with filename image.exr. The file will apear in the extension folder.  You can specify a path such as C:/Users/Spheric/Documents/image.exr

 

You could also do stuff like write an HDR and a tonemapped png with

batch.periodicsave=15
film.outputs.1.type=RGB
film.outputs.1.filename=image.hdr
film.outputs.2.type=RGB_TONEMAPPED
film.outputs.2.filename=image.png

 

types can be: RGB, RGBA, RGB_TONEMAPPED, RGBA_TONEMAPPED, ALPHA, DEPTH, POSITION, GEOMETRY_NORMAL, SHADING_NORMAL, MATERIAL_ID, DIRECT_DIFFUSE, DIRECT_GLOSSY, EMISSION, INDIRECT_DIFFUSE, INDIRECT_GLOSSY, INDIRECT_SPECULAR, MATERIAL_ID_MASK, DIRECT_SHADOW_MASK, INDIRECT_SHADOW_MASK, RADIANCE_GROUP, UV, RAYCOUNT, BY_MATERIAL_ID, and IRRADIANCE

Dartanbeck

...and LuxCore is looking promising too. Have you been playing with the multi-pass in it?

I must say, I wasn't looking forward to learning a new way of shading - but it's set up pretty darned nice!

DesertDude

I'm adding to this thread another type of pass I recently learned about - an RGB Light pass. I couldn't find many examples aside from super advanced software, like a Maya/Nuke workflow, but I think I managed to create the same thing using Carrara. My example is a bit exaggerated mainly to show results, but I think one would use this technique to very subtly adjust light intensity and color in post work.

My Nose Lady character is today's model.

So, first image is the "beauty" pass with 3 Distant Lights, A warm Key Light and 2 cool Bounce lights set to lower intensities. To create my custom RGB Light pass I then set the color of each light to a pure color, Red, Green and Blue, as seen in the second image/screen grab. The third  image is the rendering of the custom pass.

I then brought both "beauty" and the custom RGB Light pass into After Effects (and HitFilm to be software agnostic) and created 3 separate comps and isolated the lights using the Shift Channels effect (Channel Swapper in HitFilm). I then comped those over the beauty pass and applied a Hue/Saturation effect (Hue Colorize in HitFilm) to drastically alter the color of the lights to illustrate the outcome.

And just like an RGB Matte you can create secondary and tertiary color schemes to isolate any number of lights to be a super geeky control freak in post.

I hope I am not misunderstaing the technique, kind of flying blind here. Anyone who has more experience with this please chime in and correct me.

Edited: for my notorious typos

cdordoni

DesertDude said:

I'm adding to this thread another type of pass I recently learned about - an RGB Light pass. I couldn't find many examples aside from super advanced software, like a Maya/Nuke workflow, but I think I managed to create the same thing using Carrara. My example is a bit exaggerated mainly to show results, but I think one would use this technique to very subtly adjust light intensity and color in post work.

For a single frame I think I understand how this might work, but what about an animation, Would the lights need to be in the same position from frame to frame (ie. parented to the camera)?

DesertDude

cdordoni, it should work fine for an animation, no parenting needed. Whatever your lights are doing would work fine, whether they are stationary or moving. It's a fancier "matte" for more control for post work, kind of like an RGB matte where you can economically isolate more areas all in one render.

Diomede

Thanks for posting the technique, Desertdude.  Glad I have this thread bookmarked. 

Rashad Carter

This is definitely one of the best threads in the Carrara forum. Very informative!!