I got an e-mail from Dan, who wanted to know my process for creating a scene - so I figured I would share how the first crime scene was born, with some screen captures of the various stages and of course the final result vs the starting point:
The murder has taken place in the city, so the first thing to do is to find a suitable city asset and decide on where it happened. Satisfied with the location, I then created the main camera and framed the scene.
After creating the camera I added vehicles and arranged them - I actually started out with only police cars, but then added the forensic technicians' van (since are there in the story) and finally added the ambulance to create a focus point and balance the scene out. At this point I also added depth of field to the camera in order to be able to work on the visuals presentation of the scene.
The defalut scene lighting is a generic HDRI map, which had to be replaced with a suitable night-time HDRI for the "ambient" lighting, but those are pretty light as well, so I had to adjust the intensity to give it a deeper, darker feel.
Next up the lighting puzzle started. Seeing that the lighting sets the mood, I started by adding street lights - then moving on to creating read and blue emitters for the emergency lights (which turned out pretty good but needed a lot of adjustment). The scene was simply too crisp by simply adding colored lights, so I decided to add a volumetric fog at the far end of the street where the police cars are to soften the overall look of the render.
Finally, the story dictates that it should be raining - and this was a huge challenge to do in Daz Studio, as there are not a lot of "rain" assets available, and those I did have turned out to be extremely hardware intensive and cumbersome to work with, so I decided to add 3D droplets to the ambulance but to the rest of the rain as post-work in Photoshop.
Here's the original render and the final render from the Visual Novel:
Take a look at the render below - it features our three main character plus a few others from PoP enjoying some time off having a barbeque in the sun:
I made this scene relatively quickly, and there is no denying it does look pretty good, but as much as I hate to admit it, I am not a 3D artist though - which is why I am so incredibly thankful for the two core-component pieces of software I use: Daz Studio and Adobe Photoshop!
Nothing in the scene above is made by me - it is the result of a process of utilizing ready-made assets to the best of my abilities, in a compositionary (is that even a word?) way; combining various props to create an environment, modifying existing characters to give them the look I want, posing said characters (including expressions) and adding subtle details.
This is what the scene looks like when I am editing it in Daz Studio and the perceptive viewer will notice that the only 3D elements are the characters themselves and the barbeque setup. 3D objects are resource intensive, and if the surrounding environment was to be 3D in such detail as displayed in the render, there is no way my trusty PC would be able to handle it, so I cheat by using an HDRI (a spherical image giving the illusion of 3D space and also emitting light)! When using and HDRI in this way (as opposed to just using it for the sky / ambient lighting) is great for single shots like this one - because once all your 3D assets are placed and you rotate or zoom the view, their relative position to the HDRI environment is distorted - making it hard to do different angled shots without having to adjust the 3D elements as well (first major hurdle in Daz Studio).
Above you see an example of combining assets: the light / environment supplied by the HDRI, the barbeque and shelter structure, the steaks (originally a Poser prop I purchased 10 years ago!) and the smoke (a VDB file of steam). My process involves looking at the "complete staged" scene and then identifying if anything can be added for realism; sometimes the devil is in the details...
...you can't really smoke a cigar without smoke... yet, the cigar 3D prop doesn't come with smoke - so I just browse my library and hopefully find something which can be used :). In this shot, the HDRI issues can be spotted if you look at the background; the older dude is zoomed in as opposed to the finished render, but the background is not zoomed in the same way - making it look like the men have moved way in front of their original placement (look at the piece of road and the house in the background). There are probably ways of working around this, but that's currently not in my skillset...
Time for my second gripe with Daz Studio: POSING. The characters you can come up with in DS are insanely good - but it's also kind of "Apple-y" as everything Daz Studio is very proprietary and does not play well (or easily) with others. Unfortunately rigging / posing is quite challenging in DS; you can't just drag and limb and start posing, you have to work very meticulous and in micro steps in order to create life-like poses. I think this is one of the reasons there are so many bad DS renders out there; the software is available for free - but requires training and time to get good results.
The third, and final, major complaint I have with Daz Studio is the one that affects my particular usecase the most: there is no collission detection or surface dynamics to speak of. Which simply means that if two objects intersect, they just blend into one another. This SUCKS for people like me who want to have, say, a hand pressing down on someone's mouth.. without hard-/softbody dynamics and collission detection, you have to be E X T R E M E L Y careful when posing to make it look good. It's impossible to get perfect, and insanely time-consuming to get it to look decent (depending on how high you set the bar of course) - but I've seen my own progression, and although it has taken me the better part of 10 years to get to where I am now, I still see continuous improvement - which is probably why I'm sticking with it!
Speaking of sticking with it, since you've come this far - here's a little reward for your interest in the "creative" process (a tiny re-arrange of some items, major changes for some and clever use of depth-of-field to combat the HDRI inconsistencies):
