Justintime9 wrote:Mostly I use larger images now. (1000 x 1000, or 2000 x 2000.)
I don't know if this was just lazyness in writing and you actually meant 1024x1024 and 2048x2048, but using 1000x1000 or 2000x2000 is relatively useless. PyPRP exporter will scale those images down to 512x512 (from 1000x1000) or 1024x1024 (from 2000x2000). This is so because graphic cards can only handle textures with sizes of power of two, which means, allowable dimensions are starting from 2^1 = 2 up to 2^11 = 2048.
Values to choose from: 64, 128, 256, 512, 1024, 2048.
You can however mix dimensions, so a texture of size 512x1024 or 2048 x 256 is possible and will not be resized before export.
Regarding your before and after images, well, I don't know how to say that. Please forgive me, but in my opinion, the before image looked more natural. Sure, you tried to follow Tweek's example, and the after image has shadows now, but te room looks so cold, and darker than before. What Tweek did was increasing the intensity of the bluish light and let it add to the orange base lighing of the room. What you implemented was somehow replacing most of the orange base light with cold blue. The floor became darker than before, the ceiling now suddenly seems to emit blue light.
I'm really sorry Justin, I don't want to make you angry or sad, but I think I should mention it before you did too much work in upgrading Tsoidahl Prad.
So, what now? I feel talking and thinking about the tools and workflow - as we currently do - is a good idea, a nice set of tools is a basic need. But don't forget to lean back from time to time, let the image impact you, see how it looks and feels, without looking at the shadows itself, textures, lightmaps. See how a rooms feels, if you feel comfortable in it, if you want to stay in it, explore it, or feel like "get me outta here" instead.
Sure, this is the most difficult task, but in my opinion it's essential for building nice ages. I would underline what Calena said: Take a dark room (in reality), put a lamp into it, and see what the light does. How the shadows look. Note the difference in brightness at the terminator, the line between line and darkness. The sharpness of the line, depending on the distance between lamp, object, and receiver. Have a look at this a long time, not just a few seconds.
And keep in mind that shadows aren't that hard in reality. Not so hard as a rendering engine calculates them. Human eyes distinguish between many brightness levels; we can see at a starry night at about 0.001 lux, up to a bright sunny day, which is about 100,000 lux. On a computer screen, you can display only 256 different levels of brightness. But the eye/brain system can compensate, it interpretes 0=black as "no light" and 255=white as "full brightness". So it's a good idea to make shadows much more subtle than you already did. A shadow in reality, which is ten times darker than the surroundings, measured with a lux-meter, is seen by our eye only around half the brightness difference. We have to keep this in mind when building ages.
But don't let this disaffect you - nice ages *can* be built, and I'm sure you're one of those who will manage it very well.
Oh, and regarding this:
Justintime9 wrote:So I don't really know what to use Ambient occlusion for, or how I would go about using TWO separate lightmaps. Of course if it would give a better result, I'll certainly try it.
I would again recommend to open up the Relativity files. It's explained all there, plus examples.
And some words about lighmaps plus vertex paint: This sounds logical, but is not necessary in most cases. If you already have a lightmap for an object, you can do everything with it. Each brightness level from 0 to 255 (assumed your underlying texture *has* that brightness). For each pixel. You can even colorize an object with it. Mixing both lightmap plus vertex paint will mostly lead to confusion - it can not increase the 0-255 possible levels.
Vertex painting is useful as a cheap alternative to lightmaps, and sometimes easier to "paint" on the object.
It has its uses though, but only for increasing or post-enhancement of object shading. For SHADING, not for SHADOWS. That's a big difference. Tell me if you want to learn more about the difference.