Do many of you do complex scenes in layers ?

montrealfilmguy

I feel it must be quite beneficial to do each part or 2-3 characters with transparant BG and then layer it all up in Affinity photo ?

Then you can do adjustments on each layer or add a fog layer etc...

Tell me your tips and tricks Dazzlers.

FSMCDesigns

I never do. If I had to for a project, I might, but I prefer to get a full render the first time. Then again I am not a big postwork fan either I prefer spending my time making sure the surfaces and materials are right in the app so the renderer can do all the work, I get more satisfaction that way and don't have to second guess my self as much.

There have been plenty of conversations though on doing it the way you mention, so it's whatever works for you. Give it a shot and see how you like it.

Quixotry

If my scene has a lot in it and my graphics card can't take it and ends up dropping things down to CPU, I will usually render my scene in stages: background elements, character, props, and so on in groups. Then they can be layered together in Photoshop. I don't often render scenes that need to be broken down like that, but it's a handy thing to do sometimes and can be a timesaver in some cases.

cstrauss42a

Sometimes I render my lights in separate layers so I can adjust the light intensity in photoshop later to get the desired effect

fred9803

The problem with breaking up scenes is that, particularly with Iray, things have an affect on other things, even at a distance where you think they would't be interacting. Adding/removing scene components is difficult to pull off as lighting, shadows and radiosity all play off each other. So you will often get a mismatch when compliling broken elements because of this. It certainly can be done but only very judiciously.

FSMCDesigns

fred9803 said:

The problem with breaking up scenes is that, particularly with Iray, things have an affect on other things, even at a distance where you think they would't be interacting. Adding/removing scene components is difficult to pull off as lighting, shadows and radiosity all play off each other. So you will often get a mismatch when compliling broken elements because of this. It certainly can be done but only very judiciously.

agreed.

Hylas

I don´t do scenes where that is necessary all too often, but yes, I do it if necessary.

For example, render with 3 characters. Make 3 renders, but in each render only one character has full SubD. Combine in post. Even if the characters are interacting it shouldn´t be too tedious, since the renders are essentially the same; they just have more or fewer details in certain spots.

Spacious

When various items in a scene interact with each other sometimes I will export those items to OBJ, decimate the geometry in Blender, reduce the maps in GIMP, reimport to DAZ to use as placeholders, shadowcasters, reflectors, whatever.  Similar to Hyla's technique above, just taken to more of an extreme.  I've found that only the forground and  focus items really need to look good.  It's a lot of steps and tedious for sure, but it sort of accomplishes the task.

DustRider

Like FSMCDesigns, I tend to prefer to a single render (I sometimes use canvases, but not often). I find Scene Optimizer to be very helpful to get everything to fit on the GPU.

I also think fred9803 brings up some excellent points. Depending on your style, and things like missing reflections and shadows from objects that are turned off could be extremely obvious.

But it's all rather subjective. A lot of people render in layers and create beautiful work. I'd say do what works best for you.

ShelLuser

For a moment I was confused about the question a bit, but... I always make my renders in one go, no layering and afterwards editing. At best I'll use Photoshop (Elements) to brush up some parts if needed (for example when it turns out that I got the lighting a bit off). Sometimes it's simply easier to use PE than letting my computer render the whole thing again for a several hours.

However, when working with more complex scenes (irony would have it that I'm messing with one at the time of writing) I do have the habit of breaking my work apart and then bringing it all together in a single scene. So I'm creating sub-scenes which are small(er) parts of the big setup and then work on those individually without the distraction of the eventual environment or the other props. I then "layer" the whole thing together (I call it layering because I go with one section / area after the other usually working from back to front, but it's obviously not the same thing you're talking about) and eventually I'll have my scene set up after which I then add (or adjust existing) cameras, work some more on the lighting (if needed) and then I'm good to go.

I really grew a huge liking to sub-scenes over the years, also because these allow you to easily re-use stuff at a later time in other renders.