Making Battle Reports Into Videos Without Losing Our Minds
We had one of those very relatable hobby-media moments recently: the kind where excitement collides head-on with the reality of time.
Stas dropped a fresh video into chat, and the immediate reaction was exactly what you want to hear after publishing something: it was fun to watch, people replayed it, and the overall vibe was yes, this works. That kind of response is always a nice boost, especially when you’re still figuring out what your own style is supposed to be.
But there was also the other side of it — the side everyone who has ever tried to turn hobby time into content knows very well.
“Either I’ll find some other formula for this, or I’ll stop, because I spend as much time making one video as playing three games.”
And honestly? Fair.
That one sentence probably sums up the eternal struggle of hobby content better than anything else. We love having photos, videos, reports, little highlights from games, all that stuff. It’s great to look back on, great to share, and great for building the atmosphere around the hobby. But the production side can very quickly eat the time that was supposed to go into actually playing.
Looking for a style that feels right
What we liked most in that exchange was how clear the real goal was: not just “make more videos,” but find a personal formula.
That matters a lot. There’s a huge difference between copying a format because it exists and slowly discovering what kind of pacing, editing, and presentation actually feels natural. Short highlights? Calm cinematic shots? Fast cuts with jokes? A narrated recap? Something halfway between a battle report and a memory reel?
That part takes time, and probably a bit of trial and error. Unfortunately, trial and error also takes… more time.
Still, when a video gets the kind of response Stas got — enthusiastic, immediate, and very genuine — it feels like a good sign that the search is going in the right direction.
The most 2025 solution imaginable
Then Michał came in with what might be the most modern answer possible.
Instead of manually suffering through every second of editing, he described a workflow using ffmpeg to extract frames from a recording and then handing those frames over to GPT with a prompt along the lines of:
Based on these frames, generate ffmpeg commands that will splice together the most visually interesting and impactful segments into a single highlight video. Choose smooth transitions between clips. Also, write a short, engaging narration script for a text-to-speech model to read over the final highlight video, matching its tone and pacing.
Which is, we have to admit, both slightly terrifying and extremely funny.
Also: kind of brilliant.
There’s something very hobbyist about this approach. Not in the “perfect polished studio pipeline” sense, but in the classic “we refuse to spend six hours doing something by hand if there is a weird tool-assisted shortcut available” sense. That energy is deeply familiar.
The important part: people genuinely enjoyed it
The best bit in the whole conversation wasn’t even the technical side. It was the reaction:
- “No, seriously, do this.”
- “I love it.”
- “It’s wonderful!”
- “I’m already watching it for the second time.”
- “It’s really fun to watch!”
That’s the stuff that matters.
When you’re making hobby content, especially in an experimental format, it’s easy to focus only on what still feels rough from the creator’s side. You see every awkward cut, every thing you’d do differently next time, every minute spent exporting instead of rolling dice. Meanwhile, everyone else is just there enjoying the final result.
And maybe that’s the takeaway here: if the format is fun to watch, then it’s worth trying to make the process lighter rather than abandoning it outright.
Probably the real challenge
So this feels like the actual challenge now:
how do we keep the charm, keep the personal style, and cut down the production time enough that making a video doesn’t cost three games?
That’s a much better problem to have than “nobody cared.”
We’re very curious where this goes next. Maybe the answer is shorter edits. Maybe a repeatable template. Maybe AI-assisted clip selection. Maybe some hybrid that keeps the personal touch while removing the most tedious parts.
Whatever the final formula ends up being, one thing seems clear already: there’s definitely something here.
And if a video makes people immediately want to watch it again, that’s a pretty strong sign not to stop.
If we keep experimenting with this side of the hobby, we’ll gladly share more — especially if we manage to find that sweet spot between cool to watch and doesn’t consume an entire campaign’s worth of free time.