17/12/2025 Development Update

Hello! Sorry for the late update, I've got tons of stuff going on here. With that being said, I'm going to keep things short this time.

Progress

So far, the render count comes to a total of 596. Not much else to mention besides that it's going to get a little busier for me in the coming weeks. So, please expect a slower progress for a while. God, I want to upgrade so bad; rendering is the worst part of development for me, and I can't wait to get a new GPU. But given the current market, it's hard to get one without going broke. And of course, if we're going to upgrade, we're going for the best.

Adapting with a mid-range setup

It sounds crazy when I say an RTX 3080 is "mid-range", but when it comes to working with 3D renders, it does seem to be so nowadays. Like, it's not so bad at operating the software I need, but at the same time, the productivity I can get out of this card is quite limited. At the time, the spec of this card is not bad; dare I say, it was rather high-end. And truthfully, for chapter 1, it fulfilled my expectations quite well. But moving forward, my ambitions grew larger... exponentially. I want larger scenes, larger environment, more characters on screen! But, reality is always disappointing. My card has shown signs of fatigue. It was no longer able to keep up with my demands (Que the Good Morning Sunshine Cow meme). And just like most people, I'm not at liberty to spend money as I please. So, like a normal person would do, I began looking for dodgy ways to shave off render time and still try to keep up with quality.

Shit I did to optimise render time:

• Always try to get a closer shot to the character as they speak so I could hide everything else in the scene that is not inside the frame (Totally intended as a filming technique :P)

• Turning the glasses on windows from transparent to coloured, making it unable to show the environment outside, reducing both extra work and render time.

• Utilise object instancing in large environments

  • Manually craft environment using instancing and angle the camera in way to make repeated object looks less noticeable

• Always angle the camera to show as few characters as possible

  • If there's more than 3 characters in the scene, angle the camera to always show only 3 characters. (3 is the limit for me)

  • Hiding characters behind another character in front of the camera

  • If I can't avoid showing only 3 characters, put them out of focus and lower their mesh resolution to base, or, render canvases to manually Photoshop a separately rendered character onto the main render

• Always lower texture resolution on clothing and environment objects

• The use of special interior camera (I don't use them anymore as they mess up the lighting)

• Using prop figures to generate crowds

  • Crowds are always on the lowest texture and mesh resolution. Out of focus to hide the hideousness

  • I used to have entirely black figures so I didn't have to render the texture and change the clothing, but I don't do that anymore, all for the realism (Not gonna go back and redo those renders lmao)

• Render images at 65% convergence or above, and denoise with a third-party denoiser

  • Sometimes I can't even reach 65%, so usually I set a time limit for sets of renders that require more time than the other ones.

• Avoid relying solely on the environment lighting (as in object lighting in the environment), and replace them with spotlights.

• Using low camera angle to avoid showing where crowds are supposed to be (Totally intended as a filming technique, again :P)

  • If you don't show the crowd, the crowd doesn't need to exist... I guess

These are the only things I could've think of, for now. I'm sure I've done more dumb shit just to shave off I few more seconds. Oh, well.

Acknowledgement

Thanks for sticking around this community for however much time you've been here. You are my motivation to keep doing what I'm doing. If you have the time, consider leaving a review on wherever you can find my game. Your support comes in a lot of ways that does not have to be monetary. I love reading comments and constructive criticism about my game. One of my favourite things to see is players trying to predict where the story is going next. When they get it completely wrong, I'd be like "Ha, just you wait". Anyway, this has been the newest dev update. I'll see you in the next one. Catch!