You must be registered to see the links
How much effort goes into image renders? Over time my attention to detail has evolved. There are some things I previously ignored or gave up on, but nowadays take extra time to fix.
Asset Fixing:
Whether it's assets available from the Daz Store, or items I need to create myself, nothing is perfect. Usually I deem their quality as "good enough" and work with what I have, or make quick adjustments where needed.
But as time goes on and I become a better artist, some things I'm just not happy with anymore. Just one example are the baseball pants my characters wear, which has a back pocket that isn't even attached.
Below you can see where I've started fixing the geometry. I'm only focusing on the main problem: not having a floating pocket. If it was going to be an asset I distributed then I would clean it up all the weirdly shaped polygons, but since it has no impact on how it looks after rendering, I'm better off not taking THAT much time on it.
And after some quick dirty work correcting the geometry we now have the pants in one piece. If I were to take this one step further I'd smooth out the corners so they don't look so rigid, but I'll wait to see how it looks in game renders first.
Lighting:
Besides fixing asset shapes and looks, lighting has been the most important aspect I've yet to spend considerable time on. In fact, one of the hardest things I've learned when creating a 3D scene is how to arrange it to get the best quality and efficient rendering times. This is mostly impacted by how much stuff you have in the scene, and the size/intensity/quantity of light sources you're using.
One way I've started to help with my scene setups is to remove part of the structure, like ceilings or walls, and add more consistent lighting sources. Using the default lighting sources for a room (such as ceiling lights and lamps) has been one of the largest mistakes I've made in development.
In the first three episodes I constantly spent time trying to adjust light sources to properly light the characters in a room, and the color/intensity would often change based on how I felt the renders looked.
By creating a standard setup for common scene locations, it'll allow quicker setup and less fiddling around trying to fine-tune things. On top of that, render times are reduced (sometimes significantly) by taking these measures. This is another case where putting in the work upfront pays off in the long run.
Experimentation:
While I learn a fair amount just through my normal scene creation for TaH, my workflow is rather confined to doing the same stuff I'm accustomed to. There isn't much room for me to experiment and rather just do what I always had. But that's where promo and custom renders come into play.
Being able to work on art that's not directly being used in the game gives me some freedom to try different approaches or styles I haven't done before, and can reveal some interesting lessons.
One of these was when working on a commission render that featured a couple characters shrunken in size. What became immediately apparent is thinner parts of the body, like the fingers, nose and ears, looked more akin to a plastic doll. (Don't mind the grainy images as I didn't render them out fully)
This is because of how Daz tries to handle light entering objects. We often think of our bodies as opaque, but light travels through us quite easily. Shine a flashlight into your palm and you can see a faint glow on the other side.
By shrinking the character, this effect was enhanced because it's using a set distance measurement for light to scatter below the skin surface. Since the size of the character shrunk down to 10%, the distance of light scattering below the skin should have also decreased the same percentage, yet it remained static.
The second image here shows what happens after I corrected it, and we see something closer to normal.
While I had an idea how the different settings worked for transparency, light scattering, etc. on skin, this was something new I stumbled into that I never fully visualized and now I have a better understanding when making adjustments to textures.
Final Thoughts:
A lot of problems can be solved by rendering out images longer or doing some postwork with editing software, but these are not the ideal ways to tackle the problems. My hope is to create a more efficient workflow to take most of the tedious stuff out of my hands. It may not necessarily equate to less development time overall when you factor in the extra work upfront, but it will result in higher quality and more consistency at roughly the same amount of time in the end.
