Progress Report - October 16, 2020 - Technical Rachaos Work (Patreon)

Hello, everyone!

As I'm sure you've noticed, Liziverse still isn't released. I want to apologize profusely for that.

I've had it fully rendered, edited, and ready to export for 10 days now. The hang-up isn't on my end. Rather, the root of the problem is my Sound Designer's schedule has been an absolute nightmare. Between him working full-time in retail (in America, EG an "essential employee" amid all of the business layoffs and cutbacks due to Covid) and also being a full-time university student (EG taking all of his courses entirely online now) whose professors decided that "online" means "can do twice as much work in the same amount of time", he's only been able to find a few hours a week to work on this. Often times, only on Sundays. And then all of that on top of him currently residing with his parents (normally he lived in the university dorms and had some privacy, but since the universities are closed due to Covid...), which makes even finding time to work on Sundays difficult.

So in a nutshell, and I can't believe I actually have to say this, but despite the fact I work entirely online and remotely, even my work is being affected by Covid. Again, I apologize profusely for this. My sound designer and I are going to have words about our working relationship until this crisis is over - I don't fault him for this, but hanging project production for literal weeks is not an ideal position, especially since I'm just sitting here on my thumbs waiting for him. I don't want a repeat of this for Rachaos.

Speaking of Rachaos, you may have also noticed that today's preview video is a bit more... technical than usual! I unfortunately had to compress its several parts together to be able to upload it somewhere that Patreon will embed it, so it’s not very watchable in the thumbnailed form – click here to watch the full video.

I haven't finished the Rachaos screenplay yet, which is why I haven't posted it to the Google Drive yet. But I am providing an unfinished build of the screenplay at the bottom of this post, because it provides important context to what I have been working on for the past few days.

In a nutshell, there is a single shot in Rachaos that necessitates the use of some generic DOAFantasy characters. Now, I could have just used some other DOAFantasy characters (like Tina and Lisa, for example), but there’s two problems with that:

The first problem is that I am currently in the process of overhauling the DOAFantasy character models, and as such, I don’t have them available. Seeing as I plan to have anywhere from 3 to 6 generic DOAFantasy characters in this one shot, that would mean putting together and faceposing 3 to 6 characters, all as preproduction, all for one shot. And that’s a lot of work.

And the second problem is that I have a very, very large metaproject under the surface, referred to as the DOAFACU (Dead Or Alive Fantasy Aardvarkian Cinematic Universe). It unites all of my DOAFantasy projects (with the exception of Miscellaneous Expenses, which was written before the DOAFACU was imagined) into a single continuity, and I want Rachaos to be a part of it. I already have a end-goal project for DOAFACU written out (the Infinity War of the DOAFACU), and all 14 of the original DOAFantasy characters have roles and personalities defined for it. Using them as generic fodder characters for this one shot of Rachaos causes complications toward that end.

Between those two complications, and the fact that I’ve been wanting a good generic female-character base anyways, as well as wanting to refactor my generic males to use a similar shared-textures system as the newer DOAFantasy models, I decided to take this unexpected wait on Liziverse to take the time to sit down and build a generic-character base.

There are two components to this system: The generic character model itself, and the “generic character randomizer” script that really makes the magic. For this particular project, the generic character model is built directly onto the DOAFantasy DazV5 Prototype 3 (“DOAF P3”) standard, meaning that all outfits and poses made with any DOAF P3 models will also work with them, and that (assuming at least one DOAF P3 model is already spawned in the session file), they will only take a minimal amount of memory (approximately 2MB of memory, as opposed to the 150+MB necessary if it didn’t use the shared textures of the DOAF P3 standard).

But the interesting part, and what is being shown off in this video (and is the focus of this post), is the script. For posterity, I am going to refer to it as “the Generalizer” for the rest of this post. The Generalizer will ultimately be composed of 7 parts, or 7 generators. For Rachaos, 5 generators are planned to be built, and 3 of them are already fully functional. The last 2 generators are unnecessary for Rachaos, and so I will develop them at a later date.

The 7 generators, starting with the 3 already made, ending with the 2 I won’t be making now, are as such:

• 1. Skin-tone generator
• 2. Eye-color generator
• 3. Face-shape generator
• 4. Hair-style generator
• 5. Hair-color generator
• 6. Body-shape (and height) generator
• 7. Face-texture generator

As the video demonstrates, the hair-style and hair-color generators have all of their relevant materials, but I haven’t written the code for them yet – all of those hair-colors I dialed in by hand, for the purposes of this video. The hair-styles themselves are being salvaged from my Generic Dark Elf Project (GDEP), which this new generic-character system is a successor to.

Speaking of the GDEP, the Generalizer has been specifically written to be fully generalizable. Nothing in it is hard-coded, but instead everything is fed through a complex data-structure that resolves based on the animation-set’s name in SFM. The ultimate goal for this is to be able to have multiple character archetypes share the same model-base.

A simple example is Fallen Throne: Using a single generic elf model as a base, I can generate purple-skinned Dark Elves, blue-skinned Ice Elves, red-skinned Blood Elves, chocolate-skinned Wood Elves, or pale-skinned Light Elves. All I have to do is rename the spawned model in SFM to “darkelf” or “bloodelf” or etc. Each has their own Generalizer data structure which dictates how the generators behave – dark elves get purple-skewed skins, blood elves get red-skewed skins, etc.

The code itself is being written in a very generalized fashion, which will in theory make it trivially easy for me to port over to Unity (or whatever game engine I decide to work with), meaning that I can easily carry my generic characters around with me in whatever field I pursue – they’re not constrained to Source Filmmaker.

That’s really all I have for now. My sound designer is confident he will have the next pass of audio ready for me to examine by the weekend. Assuming that it’s solid and ready to ship, then hopefully we’ll have a release in a few days. If not… We might be waiting another week.

I want to apologize for that again.

And as promised, here is the unfinished Rachaos screenplay. I have attached the high-level outline of the whole project to it as well. Hilariously, the outline is also unfinished, but it at least gives a wider sweep than what the screenplay presents. And as always, no promises that it won’t change between now and formal publication to the Google Drive, let alone between then and the final release of the Rachaos project.