Trilium promo video

[ricolandia]

I've been meaning to make a promotional video for Trilium for a while, and with this new release( a great, great one!!), I see it's time to finally move forward.

What I already have so far:

1. The script
2. The base 3D scene, built in Blender
3. The kinetic typography template in Blender
4. Voice-over
5. Soundtrack

What's still missing:

6. Kinetic typography — I already have the Blender template for the text motion; I just need to animate each text block and render them out with an alpha channel to overlay on the videos.

7. Screen captures for each scene described in the script. I was planning to do all of them myself, in a separate Docker instance on my homelab (I've already installed and populated it).

But… I thought it would be fair to invite the community to participate, so I paused my production. After all, I don't want to make this decision alone — I'm just an enthusiastic user! :)

My proposal, if you agree:

7.1 First, review the script and let me know if this is the direction that developers, contributors, and the wider community would like to see.

7.2 Once the final script is approved, we could proceed in one of two ways:

a) Each interested user could record one of the screen captures (in FullHD or 4K) and make it available, so I can composite it into the Blender 3D scenes.

— or —

b) If you prefer, I can make all the source files available in a public repository. Let me know which path you'd rather take.

The script is attached below.

Final Script.md

Rendered Frame

[adoriandoran]

That's actually great news! Trilium is still flying a bit under the radar when it comes to promotion. I checked the script, and it seems to cover the big picture and the key features that differentiate it from other applications.

But I think some other aspects deserve attention as well, such as the text editor capabilities, Markdown notes, spreadsheets (the note type, not the collection), mindmaps, PDF annotations, canvas, and geo maps. These would show users who have never used Trilium before that it’s a feature-rich application, not just a self-hosted open-source note-taking app that works with text. I believe showcasing these visually in the background during the voice-over would be sufficient; they don’t necessarily need to be mentioned in the script.

I also think we should create a few new demo notes, since the existing ones are heavily outdated. They might be useful for the video.

[ricolandia]

Hi @adoriandoran, I really like the improvements! If you put together these demo notes, I can import them into my instance and take screenshots of the results. Can you do that?

I'll wait a couple of days to see if anyone else has suggestions. After that, I'll rewrite the script with these new scenes and share it with you before the next milestone.

Thank you!

[adoriandoran]

Sure, I’ll take care of the demos over the next few days.

[Staenly]

This sounds like an amazing initiative! I very like it. However, I’d like to share my thoughts on the timing of the release. Nowadays, having a mobile app is almost a deal-breaker for many users when choosing a note-taking tool. Since Trilium currently lacks a stable mobile app, some people might lose interest once they realize that. Therefore, I’d suggest holding off on releasing the video until the official mobile app is out, which seems to be happening very soon

[ricolandia]

Hi @Staenly, good point. An official version would attract more users. I personally am satisfied with the pocket version by @Nriver, but I totally understand what you're saying.

We also still have the synchronization bottleneck. A one-click sync service is needed, as already outlined in discussions #9572 and #9084, when we talked about having something that is easy for the average user.

We should also consider unifying the repository for the scripts (which we call plugins, add-ons... or do we fully adopt the term "scripts" once and for all?). I've tried something that I am using for my personal organization. It is mentioned here: #9954.

In any case, nothing prevents us from creating scenes based on what we already have, and since the scenes are "modular", we can insert and reorder them on the Blender timeline (VSE). I'll move on with what we have and await considerations like this one to deliver a "final video".

Once all the scenes are ready with what we have, I will share the source files with everyone (I just need you to tell me where to upload the files), and then we will have material that can be used by all.

I am also willing to create reusable material in .svg (via InkScape) or via Penpot, to have a promotional repository that maintains visual consistency. I'll leave that up to you. This is my contribution.

[eliandoran]

@Staenly , personally I appreciate all marketing efforts because we are most certainly lacking in that department. 😅
Regarding both mobile and a one-click service, both are on our roadmap. For mobile, I expect an Android version to come sooner rather than later.

So feel free to start your work on it; the script itself seems nice.

[ricolandia]

Hello @eliandoran. Yes, I’m proceeding with the plan. We’re going to have something modular, which we can add to and reuse.

[mdagnolops]

For the public script's CLONE scene (0:08–0:16), I'd make the duplicated note the main event and let the title settle before the edit highlights both locations. Animating the label, cursor and two highlights together could hide the feature the voice-over is explaining.

If this modular cut is still in progress, one small comparison could help: keep your eight-second scene and Blender composition, but compare the current text entrance with a short masked reveal followed by a stationary title. Start the clone demonstration after the title settles; those are proposed timing choices, not a readability result I've measured.

I build 1human, a free reference network for humans and agents. Our original Hello reveal study is CC0: borrow the motion idea in your existing Blender template, without changing renderers or joining anything. If you try that comparison on this public scene, I'm happy to review whether it makes the clone action easier to follow. I reviewed the written script, not a rendered cut; no private project files needed.

[ricolandia]

Thanks for the careful read, and for keeping it specific to the CLONE beat.

Your sequencing point is right, and it matches the intent of the cut: the screen capture is the hero and the kinetic text is a confirmation, not a co-star. What I want to avoid is exactly what you flagged: title, cursor and the two highlights all moving at once.

Concretely, the plan for the Blender edit is: push-in on the "iMac" > the title reveals and settles (~1s) > only then the clone action (cursor duplicating the note) > the two locations glow in the last third, with the title stationary. I'll also run your comparison on that 8-second scene: a short masked reveal followed by a stationary title vs. the current entrance, same composition, no renderer change.

One heads-up on timing: I'm wrapping up the source files and will make the public repository (CC-BY-4.0) available shortly with the script, art and demo setup. Production itself is paused on my side for ~2 weeks (heavy workload), so I'll run the masked-reveal comparison and share the rendered CLONE beat when I resume. I'd be glad to hear then whether the clone action reads better.

Thanks for the CC0 offer: the masked-reveal idea lands... and I'll note the reference in the repo credits. Nothing to join or change tooling-wise.

[mdagnolops]

That sounds like the right sequence: the title settles, the clone action gets its own beat, and the highlights confirm the result. When you resume and share the public CLONE cut, I can review the two versions at phone viewing size, focusing on whether the duplication is clear. Thanks for planning the credit; your existing Blender workflow is a good place to test it.