How to make a serious philosophical short with subtitles

Last updated 2026-08-17 · 6 min read

What is the short version?

A philosophical short is a single argument, delivered slowly, with pictures that stay out of the way. Its whole credibility comes from restraint — the voice, the pacing and the imagery all have to resist being interesting on their own.

1. Go to Vizard Agent and say the question the short answers.
2. Say the language for the narration and the subtitles.
3. Say the register and the length.

What do you need before you start?

The idea and a register. Vizard Agent writes, narrates and illustrates from your framing, so the thing that decides whether the short works is the question you pose — a sharp question produces a sharp argument, and a vague one produces platitudes.

• The question. "Why you should never get angry at someone who disrespects you" is a real subject.
• The tradition or thinker. Naming one as inspiration gives the writing a spine rather than a mood.
• The language. Vizard Agent searches for a narrator in it and subtitles in the same language.
• The register. Calm, deep, unhurried. This format fails when it sounds motivational.
• The length. Forty-five seconds is the working size for a single argument.

What do you type into Vizard Agent?

Pose the question and name the influence. Vizard Agent writes the argument itself, so the two things worth specifying are the idea it should defend and the tone of voice it should defend it in — everything visual follows from those.

Prompt

Variants worth knowing:

• Your own script. Paste it and Vizard Agent narrates and illustrates your words instead.
• Abstract rather than literal imagery. Often stronger for an argument than illustration of the example.
• A series. Keep them in one project so the narrator and the look stay consistent.

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real philosophical short. The audio comes first and everything visual is built against measurements taken from it, which is what keeps the pictures changing on the argument's turns rather than on a timer.

1. Checks the available generation capabilities and their pricing.
2. Reads the image generation options before planning any scenes.
3. Checks the voice options in the target language and searches for a serious, calm, deep male voice.
4. Reads the caption capability's options so the subtitle style is decided before the render.
5. Generates the narration in the chosen voice.
6. Probes the voiceover's exact duration.
7. Runs transcription on its own narration to extract word-level timings.
8. Reads the transcribed timings before building anything against them.
9. Generates the nine illustrated scenes concurrently, then searches the music library twice for a slow, thought-provoking bed.

Step 7 is the habit this whole format depends on. Subtitles on a philosophical short are read rather than heard, and taking the timings from the recording rather than estimating them is what makes the words appear exactly as they are spoken.

What does the result look like?

From the run this page is written from, probed on the delivered file: 1080x1920, H.264, 25fps, 45.0 seconds, AAC audio. Vertical, forty-five seconds, nine illustrated scenes under a calm narration, with subtitles timed to the read and an ambient score beneath it.

Vizard Agent generated all nine scenes concurrently rather than one after another, which is why a nine-image short takes minutes rather than most of an hour to produce.

Forty-five seconds across nine scenes is five seconds a picture. Vizard Agent holds them at that length deliberately: fast cutting reads as motivational content, and the slowness is a large part of what makes this format feel serious.

This category was not separately measured, so treat the timing as a range rather than a promise: comparable work runs a median of 28 to 38 minutes end to end. Across all projects the median cost by tier is Flash 47, Pro 55, Max 242, Ultra 263 credits.

When does this not work well?

This format sounds authoritative by construction, which is exactly its risk. Vizard Agent will deliver a calm, slow, well-illustrated argument for whatever position you hand it, and none of that delivery makes the position true or the attribution to a thinker accurate.

• Attributing an idea to a thinker is a claim. Check that the philosopher actually held the view before naming them.
• Serious and vacuous is the failure mode. A slow read and a calm voice over stock wisdom is still stock wisdom, just delivered more convincingly.
• Illustrated scenes can be too literal. Abstract imagery often serves an argument better than a picture of the example it uses, and Vizard Agent will do either.
• Subtitles change the register. They help reach and they push the piece towards the social conventions of the format, which is a trade rather than a free gain.
• One argument per short. A second idea in the same forty-five seconds means neither one is given room to land.
• Advice about anger and relationships is not neutral. What you assert about how people should feel is your responsibility.

How do you fix a result that came back wrong?

Say what to change. Vizard Agent keeps the narration, its word timings, every generated scene and the subtitle file separately, so a re-read or a replaced image is a change to one layer with everything else still aligned to the same numbers.

• "The voice is too warm." A colder, slower read, everything re-timed.
• "Scene five is too literal." That image regenerated more abstractly.
• "Bigger subtitles." A restyle over the same cut.

How does Vizard Agent compare to doing it yourself?

By hand this is writing the argument, generating a voice for it, finding nine images that illustrate without undercutting it, and then timing the subtitles word by word — all for a forty-five second video that only does anything for a channel as one of many.

Common questions

Does it work in my language? Yes. Vizard Agent searches for a narrator in it and generates the subtitles in the same language.

Can I supply my own script? Yes. Paste it and Vizard Agent narrates and illustrates your words rather than writing its own.

How many scenes for forty-five seconds? Around nine. Each holds for about five seconds, which suits a slow read.

Will the subtitles match the voice exactly? Yes. They come from word timings measured on the recording itself.

Can I make a series? Yes, and this format only works as a series. Keep them in one project and Vizard Agent holds the narrator and the visual style steady across every short.

Should the imagery be literal? Usually not. Ask Vizard Agent for abstract scenes when the argument is about a feeling rather than a situation.

Does it add music? Yes. Vizard Agent searches for a slow, ambient bed and mixes it well under the narration.

How long should one be? Forty-five seconds for a single argument. Two arguments in one short means neither one lands.