Lots of outputs

Hello, there’s this file on blendswap.com

The compositor of this file has a lot of output nodes.


My question: What is the benefit/purpose of having all of these passes output?

thanks

The benefit of passes is that it allows you to apply compositing effects or adjustments to just part of the data that makes up the full image.

For example, with a Cycles render, you might want to apply a setup to help denoise the image, but only on the outputs that show the contribution of indirect lighting.

The screenshot is crap, can’t read the texts on the nodes. Mostly because of jpeg compression but also because the node setup is messy and the view is zoomed out to contain all of them.

One reason to use file output node is to get the passes from Blender to an external application. Some of the passes have data that need to be adjusted before outputting to a specific file format. For example, Z pass has distance value on each pixel but a .png format only stores values in 0-1 range.

Looks as if there are several index outputs, which probably means masks were used in an external application.

One file output node can have multiple inputs. This one probably outputs masks from indices but if used just for selection, object/material indices could be contained in one image by normalizing the values, and could also map those to color.

Could render to multilayer openexr format and everything would be there in one file if the application supports that. That’s also the format Blender uses internally for render layers so if this was kept in Blender, wouldn’t have to bother with any of that when separating rendering from compositing. Multilayer .exr would work the same.

This process definitely, definitely needs to be simplified: you don’t have to do it all in one step, and you probably shouldn’t.

It’s better to do a number of small, discrete steps that are easily repeated, and adjustable individually. (I use the Unix “make” command for this!)

Use MultiLayer OpenEXR for everything, up to the “final cut” which can simply be OpenEXR without the layers.

Only when you have the “fnal cut” in this format should you concern yourself with any “deliverable files,” no matter what movie-format(s) are needed. Starting each time with the same OpenEXR input, which contains “the data” in all its loss-less floating-point precision, you compress and convert and color-balance the data to suit a particular output format and screen resolution.

My question: What is the benefit/purpose of having all of these passes output?
Render out each individual pass to separate image files, load into photoshop and adjust each image for final processing rather than use the blender compositor to combine them