Hello, everyone!
So, out of nowhere recently, I found myself thinking, "I should do a proper retrospective on Shoost!" But when I actually sat down to write it, I couldn't figure out where to start. Everything I wanted to say felt scattered, and the words just wouldn't come together.
Then it hit me. What if I answered questions instead of just writing freely? An interview format might help me get things out more naturally.
The only problem? I didn't have anyone to interview me.
So I made one. I had an AI play the role of interviewer, and went through the whole thing with voice input, talking through everything from how Shoost came to be, to where I want to take it next. It turned out feeling more like an actual interview than I expected, so I hope you enjoy it!
Q: What made you start building Shoost?
Before Shoost, I was working on a piece of software called PlayAniMaker, a VR tool for creating anime. Even back then, I had this ongoing feeling that I wanted to make anime easier to create. Then at another job, someone taught me about anime compositing and post-processing techniques, and I found it genuinely exciting. That's really where the idea came from. I wanted to find a way to make that kind of work more accessible, and that's how Shoost got started.
Q: Shoost has a strong image as a VTuber tool. Was that the plan from the beginning?
Well, not quite! I did think it would be nice if VTubers used it, but I was building it as a creative tool, not specifically as a VTuber application. I did have this idea that just like people use filters to make their selfies look better, there are probably people out there who want to make their own characters look as good as possible.
Q: Does it make you happy when you see people actually using it?
Oh, definitely! When I catch someone using it live in a stream, or I come across a post on X showing off what they made with it, it just makes me happy. Getting survey responses like "Shoost made my stream look so much better and more fun" really does make me feel like it was all worth building.
Q: What's been the hardest part of development?
Keeping the performance light is genuinely tricky. If I just implement an effect the way I want it to look, the processing load gets too heavy, so there's this constant balancing act of getting the visual result right without pushing things too far. Lately, another challenge has been the UI. As the feature set has grown, the interface has gotten more complex, and I do hear from people that it can feel overwhelming at first glance. Finding the right balance between simplicity and the depth that lets people customize things exactly how they want, that's something I think about a lot.
Q: Are there any tools or references that influence your development?
I pay attention to editing tools quite a bit. After Effects, Photoshop, DaVinci Resolve. But honestly, the thing I reference most is how users actually use Shoost. Watching how people work with it gives me more ideas than anything else. I'm also conscious of not just copying what those established tools do, because if I go too far in that direction, people will just say "why not use After Effects?" I want to make sure Shoost stays different from those tools.
Q: How do user feedback and requests factor into development?
I run surveys on FANBOX and Patreon every so often. Things I assumed weren't high priority turn out to be things people really want. Automatic scene loading on startup and the ability to rename layers are recent examples of that. The recording feature was one I thought of as a minor extra, but I got a lot of requests for offline rendering from video creators. I had assumed everyone was recording with OBS, so that caught me off guard a little.
Q: Where do your ideas for new features come from?
Watching people stream is a big one. I'll notice something and think, 'that could be so much more interesting if Shoost did this.' But I don't just act on those instincts immediately. I tend to sit with an idea for a long time, turning it over and over, sometimes even dreaming about it. I like to think things through pretty thoroughly before I start building anything.
Q: What direction do you want to take Shoost in going forward?
I feel like the core anime compositing and post-processing features are mostly in place now, so the next thing I want to focus on is making things more effortless. Like, being able to get something looking really good with just one click. I'm also interested in going beyond graphic effects and taking more advantage of its real-time capabilities. Things like having the visuals react to your voice, your gestures, your mouse and keyboard input.
Q: Last question. How would you describe Shoost in a single phrase?
That's a tough one. But I think the core concept I keep coming back to is 'accelerating creativity.' I want it to be something where anyone can produce quality visuals, and where you just think, hey, this is actually fun to make stuff. If people come away feeling that way, that makes me really happy!
And that's a wrap on the self-interview! It was genuinely nice to put into words things I don't usually articulate out loud. A good excuse to reflect on everything.
Thanks for sticking with Shoost. More to come!