All posts
12 posts, newest first. A post can belong to several topics.
November 2026
#1210 Nov 2026 Seven YouTube Shorts in 46 minutes: how an AI drove DaVinci Resolve for me
I went filming at Kellers Ranch and simply handed the clips to an AI. It ran DaVinci Resolve on its own, animated the titles, picked the music and rendered seven finished Shorts in 4K. Here is what happened, explained for everyone who has never heard of “MCP”.
#1110 Nov 2026 brunzel.info: A new project overview, a header image that grows along and a changelog
Three projects turned into six, and the projects page wasn't coping well. So now there's a new overview, a header image that assembles itself, a legal notice and privacy policy, and a changelog for every change.
#108 Nov 2026 ComfyUI render times: Windows 11 vs CachyOS, RTX 3060 Ti vs RTX 5070 Ti
Making the music video “Warum machst du das?” left me with more than 230 render runs on four different setups. The result surprised me: switching the operating system does more than a new graphics card.
#098 Nov 2026 “Warum machst du das?”: lip sync, take two
After the rough cut one thing was obvious: the lips don't match. So, all over again – this time with a clean vocal track, new scene descriptions and two machines. 13 shots are signed off, six are still missing.
#088 Nov 2026 A fresh coat of paint on brunzel.info: my AI projects get their own playground!
Anyone who knows me knows that when a new technology shows up, I'm usually already elbow-deep in it with a wrench, trying out workflows and pushing the limits. Generative AI has moved a hell of a lot in the last few months …
September 2026
#071 Sep 2026 From monster prompt to modular AI video studio: now for ChatGPT and Gemini
A single system prompt can get you surprisingly far. But sooner or later the same thing almost always happens: it gets longer, more contradictory and harder to maintain. That's why I built the Video Production System V8.1 Modular – for ChatGPT projects and Gemini Gems.
August 2026
#0628 Aug 2026 My experience with the cloud platform “RunPod”
Welcome back to the Brain Dump! Today it's about something for everyone who hits their hardware limits with local AI models: cloud GPUs, or more precisely the RunPod platform – what it costs, what it saves, and what I learned the hard way.
July 2026
#055 Jul 2026 Installing ComfyUI cleanly on Windows 11: the ultimate Conda guide (troubleshooting included)
Once you get deeper into generative AI, there's hardly any way around ComfyUI. The node-based interface is extremely powerful, but rarely forgives a messy system. If you don't want to wreck your regular system Python with clashing dependencies, Conda is the way to go …
June 2026
#0430 Jun 2026 Local voice cloning: installing Voicebox cleanly with Conda on CachyOS/Arch
If you're looking for a genuinely local, open-source alternative to ElevenLabs, it's hard to get around Voicebox (by Jamie Pine) right now. Here's how to set up its Python backend cleanly in Conda on CachyOS and run the studio in your browser.
#0328 Jun 2026 When the server starts sweating: the heatwave in HorstTower's monitoring
Summer in Germany usually means more than outdoor pools and ice cream – it also puts our hardware to a real test. I looked at a week of monitoring data from my server HorstTower: rising temperatures, a self-inflicted data gap and a Sunday rush.
#022 Jun 2026 “Acht-Drei”: a 3D comic and a joint against the quirks of the video AI LTX 2.3
Sometimes you have to get creative – not just when writing lyrics, but when outsmarting algorithms, too. My new track "Acht-Drei" (German for "Eight-Three") is an emotional trip back to the year I was born, 1983 …
#011 Jun 2026 When the frontend becomes the final boss: the search for the perfect blog design and the right words
It's a well-known phenomenon among tech enthusiasts, admins and creative minds: the infrastructure in the background is usually up faster than you can spell "web hosting". The server is purring, the backend is stable …
No posts in this topic yet.