[Secrets of Vzertos] Setting up rendering for Blender.

Hello people!
I want to share with you this half-article, half-educational, which “rips the covers off” the pre-render graphics of the game Vzertos, which I have already written about here a couple of times about this game.
In short, this is essentially a render “like in Diablo 1-2 or Fallout 1-2”.

Scene Setting

A standard Blender scene looks something like this:

Usually I always delete everything at once.To. at the very beginning it interferes with everything. Especially this lamp that shines straight into your eyes! (No)

For Vzertos, it was as if a cameraman came with all the tools and began setting up the set.
The peculiarity is that I needed to get a render of the model in 8 directions. This is mostly a matter of animation, but for convenience a circle object has been added ( Add – Mesh – Circle ) and then, for convenience, the polygon was removed (but not the vertices), an arrow shape of vertices and faces was added.

Why is this object needed?? It acts as a stand that can be rotated, much like sculptors. Before this, of course, you will need to bind the object to this “stand”, but more on that a little later.

Cameraman

We won’t succeed if there is no "eye" t.e. cameras. And here there is a very important point in the settings of both the camera and the location and bindings.

The camera is installed in position (XYZ) = (0.0, -26.0.15.0). Why this particular position? Well, firstly, the numbers are more or less “round”, and I like that. And secondly (but actually the more important part) is the 60 degree angle. But it’s important to note that 60 degrees here is relative to the Z axis, which is vertical and going up. T.e. if you calculate the angle from the “floor”, it will be equal to 30 degrees.

The camera also has its own “dummy” or “gaze center”, which really helped me in some moments. In fact, with its help it is simply more convenient to shift the camera along the Z (for the final image it turns out along the Y axis).

This is what it looks like.

It is also very important https://zodiaccasinoonline.co.uk/ that the camera must have an orthogonal projection and the correct orthogonal size. Here are my settings:

They can change depending on needs. For example, when rendering a location Meditastone the Scale parameter was greater than 100 so that the entire location would fit into the “lens”.

For ease of handling objects, rendering, and overall order in the scene, layers are used (by me).

You can select objects, press [M] and select which layer to send the object to. These are not the same layers as in Photoshop, Gimp or Crete. Here they play the role of “slots” rather than layers. It’s a pity you can’t add more than is available from the very beginning.
This way I separate the “camera” part, the “stand” part and the actual models themselves on other layers.

But there is another layer..

The sun is shining. Or not the sun. Or it doesn’t shine.

The third “system” layer is 2 directional light sources. One of them is the main light of the “sun”, and the second is auxiliary, so that the model is not illuminated from only one side.

Basically everything is ready for installation.
Lights, camera… all that remains is to shout “motor”, but there’s nothing to shoot – there’s no model, and she has no animation. But even if they were, it’s not the end yet. Next is an equally important point..

Render settings

In this project I used the so-called “old-school style”.e. deliberate reduction in the size of the final image, no anti-aliasing, and even thought about limiting the color.

Actually, this is part of the settings: we reduce the size of the final image (specifically in this example it is 38% of 256×256), remove anti-aliasing, and in the Shading tab set Alpha: Transparent (Sky is the standard there).

Can be rendered! But… wait a minute! Where is the “secret of rendering”?

"The Secret of Rendering"

(not really)

What I came to almost after the first versions of the game. Once upon a time this idea was forgotten, but I remembered it when I wanted to make my own DDNT (where, by the way, there is also Vzertos), in that project I was guided by the graphics of Dota 2 and there was a hint, the essence of which was that..

Everything lower, closer to the floor, is darker, everything higher is lighter. Use a smart gradient, dude!

The point is simple. When the picture is rendered, it is as if “empty”.e. yes, there is light and shadows, but something is missing.

You just need to multiply the image by a black and white gradient – an easy idea, but the implementation is not so simple.
If you simply “darken” along the gradient, then it will not be a cake.To. geometrically incorrect.

The next thought was to take the Z coordinate and calculate the gradients based on this coordinate. And that actually helped. Next, we still had to figure out how to install this on each material. FOR EACH ?! WHAT?! YES I HAVE 100500 MODELS ALREADY READY.

Previous experience told me that some kind of post-process method could be used. And it would be nice to put the same material on all objects before rendering, oh, if only Blender was smart… wait a minute! That’s what he can do! Yes, you can put one material on the entire scene without messing up the scene data! This is what I call real layers – you can render different parts of the scene (but apparently from the same camera) and then “compose” them at your discretion.

Any questions regarding this topic are welcome.

Would you be interested in this and that?? About Blender and all sorts of development stuff.