# How to use CharacterOS

CharacterOS is an early preview from Froyla Video Creation Studio, a product of Elisda. You choose one of five things, and the app then shows only the steps for that choice, one question at a time. You can leave at any step and come back: your answers are saved. The price is always shown before anything is made, and nothing is charged until you press the last button. Making characters and videos is being switched on step by step. Results come from AI and vary.

Page: https://video.froyla.com/characteros/help

## The five choices

Do 01 and 02 first. You need a character (01) and its voice (02) before you can make a talking video (03), a scene video (04) or a short film (05).

- 01 Create a character: teach CharacterOS a person from videos.
- 02 Give a character a voice: record or upload speech.
- 03 Make a talking video (needs 01 and 02): the character says your words.
- 04 Make a scene video (needs 01 and 02): describe a scene, in any shape.
- 05 Make a short film (needs 01 and 02): several characters and many scenes.

## Create a character

You need 4 to 6 videos of the same person (5 to 20 minutes in all), 6 pictures of them, and the person's agreement.

1. Name: give the character a name.
2. Videos: add 4 to 6 videos. Clear light, the face easy to see, different poses and expressions. Each video gets a tick when it works, or a short note on what to change.
3. Pictures: add six pictures, one for each slot: face looking straight on, side view, full body standing, smiling, serious, and a typical gesture or pose.
4. Consent: tick the box and read a short sentence to the camera. This is how CharacterOS checks that the person in the videos agrees to be copied.
5. Check and learn: choose Standard. Deep, which also learns how the person moves, is not available yet. The character is saved when you press the last button.

Who can be a character: you, or another adult who agrees. Not children, and not public figures.

## Give a character a voice

You need 1 to 10 minutes of clear speech from the person, and a saved character: only that person speaking, little background noise. A trial voice that is not kept is deleted after a day.

1. Pick the character.
2. Upload a recording (an audio or video file of up to 100 MB).
3. Type a line and play it to hear the voice. You can listen up to 6 times.
4. Keep it. A voice costs $1.50 to make and $1.00 a month while you keep it; the monthly fee starts after the first 30 days. The price is shown before you press the button, and nothing is charged until then. You can release it at any time and the charge stops. If credit cannot cover a monthly fee there are 14 days to add credit before the voice is released. The recording is kept, so a voice can be made again.

## Make a talking video

1. Pick the character (it needs a kept voice).
2. Type what they should say, up to 600 characters. They say it in their own voice, and the video is as long as that takes.
3. Choose which of their pictures to use and, if you like, a mood.
4. Check the words and the price, then press "Pay and make the video". If we cannot make it, your credit is given back.

The video is marked "Made with Froyla AI" and kept for 30 days, so download it if you want to keep it.

## Make a scene video

1. Pick one to three characters (each needs a kept voice).
2. Describe the scene, up to 600 characters.
3. Choose the shape (wide, tall or square) and the length, from 3 to 20 seconds. You see the price as you choose.
4. Check the summary and the price, then press "Pay and make the video". If we cannot make it, your credit is given back.

Each character keeps their own face and clothes, and the profile you wrote for them on the About them screen is used too. A scene over 10 seconds is made in two parts, one after the other: the second part starts from the last picture of the first, so the place, the light and where everyone stands carry on. A scene can take several minutes. Scene videos can have sound too, if you want it: a few words said by one of the characters in the scene, and a few words for the music, both optional. The words must fit the scene (about 13 characters for each second), and the price adds the speech and the music. The video is marked "Made with Froyla AI" and kept for 30 days.

## Make a short film

1. Pick the cast (up to 6 characters, each with a kept voice).
2. Build your scenes. For each scene write the script (what happens), the image prompt (how it looks) and the audio (the words spoken), choose who is in it (1 to 3) and how long it lasts (3 to 10 seconds). Add as many scenes as you like: the film is as long as all its scenes together. Tick "Start where the scene before ended" on a scene that should carry on from the one before it: it is then made after that scene, from its last picture, so the place and the light carry over (it takes longer, because those scenes are made one after the other). You can also tick "Render together with the scene before": scenes next to each other that last 10 seconds in all (with at most 3 different characters) are then made in one render and cut apart afterwards, so they look like one take and each scene can still be downloaded on its own. The price is the same. It is a new option: look at the result before you rely on it for a long film. If you want a start, ask for a first draft from an idea and change it.
3. Approve the story: you see each shot, who is in it, what they say, and the price (the production price plus the platform usage fee of $1.00 for each finished minute, at least $10 for a film). Set the most you want to spend. Nothing is charged until you press "Pay and make the film".
4. Watch it being made. A film takes about 20 to 40 minutes and you can leave the page. If we cannot make it, your credit is given back.
5. Your film: play it and download it. The voices are played over the pictures (the mouths are not synced). The film is marked "Made with Froyla AI" and kept for 30 days. Not happy with one scene? Press "Remake" next to it, change its script, image prompt, audio, people or length, and pay for that scene only (its production price plus the platform usage fee for its seconds, no $10 minimum). Only that scene is made again; the rest of the film stays as it is and the whole film is joined again. A remade scene is made on its own. If it cannot be made, your credit is given back and the film is untouched. You can remake scenes for 30 days, as long as the film is kept.

## Questions and answers

**What does it cost?** The price is shown before you press the last button on every screen that spends money. It depends on what you make (length, size, how many scenes). A voice is $1.50 once and $1.00 a month while kept.

**What if something fails?** CharacterOS tells you what happened and whether you were charged, and you can try the step again.

**How long are my videos kept?** 30 days. Every video (the ones you upload to teach a character, and the videos and films made) is deleted for good 30 days after it was made. Each finished video shows its deletion date next to its Download button. Characters, their pictures and their voice recordings stay until you delete them.

**Will the character match the person exactly every time?** No. Results come from AI and vary. Each video is checked against the character's pictures, you are told if it looks less like them, and you can remake any scene.

**Why was my video or picture refused?** The message names the item and says what to change, for example a face that is hard to see or a picture that repeats another.

**Can I change my mind?** Yes. Go back at any step. A draft is kept, so you can leave and return.

**How do I delete a character?** Open the character, press Delete and confirm. Its videos, voice and learned files are deleted with it. You can delete all your data at any time.

**Can I use someone else's face or voice?** Only an adult who agrees, with their consent recorded in the app. Not children and not public figures.

**Where do I find what I made?** Under "My characters, videos and films" on the start screen.

**I still need help.** Contact us at https://video.froyla.com/contact and say what you were trying to do.
