This is the first entry into a new Splinterlands Codex by Vin the Archivist. The file, Dossier CA-788, focuses on the subject known as "Arcane Skinwalker", detailed in Vin's words herein.
Although I've been resisting making shorts (and try to resist watching them...), it's clear that with more people on phones than computers, content consumption is fundamentally different than it was in the past. The portrait form factor works better for mobile devices, and the "short" structure works for people's lower attention spans. I'll see if this gets more traction (which could be the content or the format of course) when deciding where to go next.
The inspiration for this video started as I was watching a phenomenal designer called Gossip Goblin who has built an entire world and lore through a incredible set of shorts. I started thinking about how I could adapt this to Splinterlands, as a way to make the lore come to life, and landed on the character of "Vinnie" (or Vin for short), a brawny goblin Archivist who travels Praetoria in search of monsters and heroes who he then records into his Codex.
(1) I first worked on the script. I wanted to get the entire script done in advance so I wouldn't be guessing or improvising too much later on. While I did get a little inspiration using ChatGPT, I actually ended up writing the whole script myself as I had a very specific tone and cadence I was looking for.
(2) With the script written, I moved onto the voice. Vin's voice was the key thing holding everything together. I designed the voice in Elevenlabs, describing it as deep and gravelly with a rough Cockney accent. The inspiration for the character is Vinnie Jones (specifically in his role as Bullet-Tooth Tony in Snatch).
(3) Next I had to create Vin's image. For this I went to Midjourney, which is still my go-to for image generation. After finding the reference image I wanted, I generated alternate poses using Nano Banana in Dzine. With Vin's images in hand, I decided which parts of the script were going to show him vs using alternate "scenery" images.
(4) Next, I worked on the other images, again selecting Midjourney as the primary tool for the job. I used both Vin's image and prompting to maintain some stylistic consistency. To get all the images I needed, I created over 1000 images.
(5) With all images in hand, I used Dzine to do the lip-sync jobs (where Vin's talking) and used a mixture of Dzine (with Kling 2.1/2.5) and Google Veo 3.1 to generate the videos.
(6) For the backtrack, I used Suno. It was a bit difficult getting a track that wasn't too "musical", so I experimented with a few other tools, but with enough trial and error I was able to get a good result from Suno.
(7) Finally I stitched everything together using Davinci Resolve.
I hope you enjoy the content and if you have any feedback or questions, please let me know in the comments.
