Please download the attachment to get the full videos!
请下载附件来获得完整视频！
I’d like to share some recent progress.
First of all, I believe AI video will definitely become an important part of content creation in the future. As many of you know, I actually started using AI for some content creation quite a while ago, around the end of 2024, but I never really made it a main focus.
In 2025, I also used AI video to support some of my comic projects, but mostly as a supplement to the comics rather than the main content. I’ve also been waiting for video models to improve, and for the right opportunity to create the best possible skinsuit videos.
What I’m showing now are some of the latest improvements.
The first one is a smoother and more coherent process of putting on and taking off the skin. By using multiple key frames from the comic, I can now generate a full video directly, instead of stitching together several separate clips like before.
The second one is voice acting for the comics. At the moment, it’s already possible to generate fairly good voices and sound effects. However, I still feel that the character voices sound a bit too AI-generated, and there isn’t yet a good way to keep a character’s voice consistent. That’s one of the main issues I still need to solve.
At the same time, I’m thinking of using the story of Little Sister 1 as the first test subject for a short video. However, since Little Sister 1 was made back in 2018, some of the character models are already a bit outdated, mainly Scarlett, the brother’s girlfriend.
I’m considering whether I should remake her appearance with a newer and better-looking model. The main character, Nancy, will definitely keep her original look, but for a character like Scarlett, I think an update could work well.
At the same time, I’ve also been thinking that some of my earlier works actually had pretty good ideas and concepts, but the quality of the production back then was a bit rough compared to what I can do now.
If I were to remake some of those older works, would you be interested in that?
I’d really love to hear your thoughts and feedback.
分享一些最近的进展。首先AI视频一定是未来很重要的一个创作内容，大家也知道我很早以前（2024年年底）就开始了使用AI进行一些内容的创作，但是我一直没有将其作为一个重点内容，2025年中间我也会用AI视频为漫画进行一些配合创作，不过更多的是作为漫画的补充。我也一直在等待视频模型的进步，等待一个合适的机会去创作最好的皮物视频。目前展示的是一些新的进展，首先是更加连贯的穿脱皮过程，利用漫画的多个关键帧去直接生成一整个视频，而不是之前那样将几段视频拼在一起。其次是漫画的配音，目前能够生成比较不错的语音和声音效果，不过我觉得角色的声音还是AI感有些严重，以及角色的音色还没有一个合适的办法去固定住，这是目前需要解决的问题。
同时我想将Little Sister 1的故事作为第一个短篇视频的试验题材，不过因为Little Sister 1是我2018年就制作的作品了，角色模型（主要是Scarlett，也就是哥哥的女友）已经有些老旧，我在想要不要用更好看更新的模型重置它的样子，主角（Nancy）肯定是延用本来的样子，不过Scarlett这种角色我觉得可以进行一个更新。
同时我在想，其实我有一些早期的作品的思路和想法还挺好的，但当时的制作相比现在有些粗糙，如果重做某些作品，大家会不会喜欢？
希望得到大家的意见。