The time has come! If I remember correctly this is my first ever dedicated Halloween post/media 🤔. (Always wanted never had time 😝 but this year I found some time with the final phases of Mei s Corruption being a lot of render time I could find some mental load to affect to this) Lara Croft 3D model by Wildeer Studio Sound Design and Music by Mathieu Hellot Lara Croft’s voice by Patapon ANYWAY Let s answer the question right away because this is also and most importantly my first official prod including AI tools 😱! (Yeah I know this is fitting with Halloween because it s a scary subject 😝) I ve posted some WIPs of this project for supporters (https://subscribestar.adult/theropedude) including the first version I tested as a proof of concept and a compilation of some of the sections I didn t use in the final puzzling but that I find cool. TLDR : I love playing with this see you next time 💚. If you have been following me long enough you may recognize the first image of the video because it was indeed the first image of a full 3D video I crafted back in the days 😊(https://the-rope-dude.com/2024/05/19/lara-breast-tease-01/). But you may also know that I ve been playing with AI tools since at least 2023 and I find this kind of tech absolutely amazing. I ve tried a few things and failed a lot but now I feel like the time has come to successfully integrate some of these tools in my workflow and therefore here is the first example that I m confident enough to call a success 😁. Of course it s not perfect and I m not perfectly satisfied with everything but I never am and I m always learning + the tools apparently can t stop to get better and better every 2 months 🤩. I want to share with you how using these tools make me feel (production wise) As a 3D animator I feel like a puppeteer precisely pulling every string of my puppet to make exactly what I want. I have a lot of detailed control but it takes a lot of time. With AI Tools I feel more like an art or movie director trying to guide actors in the direction I want and iterating with several takes until I have the good one . Both have their pros and cons (Mostly control and speed of prod imo) so I m looking for ways to have the best of each world 😁. Also for example a specific camera movement is a lot easier to do in 3D but animating 6 hands groping and playing realistically with boobies and flesh is a nightmare 😝. (And to quote Kurzgesagt : I want to use AI like the align tool in Adobe Illustrator.) How was it made exactly? Of course this is overly simplified but I may explain more in depth if some people are interested but to summarize the process : - Make an image in Blender. - (Optional) Generate a 3/8 sec video with Wan2.2 (I2V) and take the last image of the video. - Make an image in Blender or Edit the previous one with Qwen/Photoshop/Blender(Control-Nets). - Generate a 3/8 sec video with Wan2.2 (FLF2V with the last frame you just edited) - Repeat - Puzzle everything in DaVinci Resolve to make it as seamless as possible. - Upscale/Upframe with Wan2.2/DaVinci Resolve - Edit/Mask/FX/patchwork in Blender/DaVinci Resolve to improve/refine/hide/modify the things hard to manage otherwise. On that matter I want to give a big thanks and shout-out to IchibanFutsuJin (https://civitai.com/user/IchibanFutsuJin/posts) we share a lot of helpful technical knowledge with each other and he does some of the most creative and amazing uses of these tools I ve seen so far 😍. Definitely go check his work! In any case I hope you appreciated this little comeback of Lara Croft and I wish you a great Halloween ^^! Take care 😊💚