Image, Video and Voice Words: Text-to-Image to Voice Cloning

AI can now make pictures, videos, music and voices. These tools come with their own words: text-to-image, aspect ratio, upscale, voice cloning, deepfake. This lesson explains them in plain English, so you can follow creative AI lessons and use the tools safely.

What You'll Learn

• What "generative AI" and the "text-to-something" words mean
• Common image words: aspect ratio, resolution, upscale, inpainting
• Common video words: image-to-video, frame, clip length
• Common voice words: TTS, voiceover, voice cloning, dubbing
• Safety words: watermark, deepfake, AI label

Generative AI

Meaning: AI that creates something new, like text, images, audio or video, instead of only sorting or predicting things.

Example: ChatGPT writing a poem, or an image tool drawing a picture from your description.

Text-to-image, text-to-video, text-to-speech

These names tell you what goes in and what comes out.

• Text-to-image: you type a description, you get a picture.
• Image-to-image: you upload a picture and describe a change, you get a new picture.
• Text-to-video: you type a description, you get a short video clip.
• Image-to-video: you upload a picture and the AI makes it move.
• Text-to-speech (TTS): you type text, you get a spoken voice.
• Speech-to-text: you speak, you get typed text. Also called transcription.

Image words

Aspect ratio: the shape of an image or video, written as width:height.

• 16:9 is wide, like a YouTube video or a laptop screen.
• 9:16 is tall, like Shorts, Reels and TikTok.
• 1:1 is square.

Resolution: how many pixels (tiny dots) an image has. More pixels means sharper. Video resolution is often written as 720p, 1080p (Full HD) or 4K.

Upscale: making an image or video bigger and sharper with AI.

Inpainting: changing only one part of an image, like removing a person from the background or changing a shirt's color. Outpainting extends the image beyond its edges.

Style: the look of the image, like "watercolor", "3D cartoon" or "realistic photo".

Seed: a number some tools use as the starting point for a picture. Using the same seed and prompt can give you a similar result again.

Diffusion and noise

Meaning: most image and video generators use a method called diffusion. The AI starts from random noise (like TV static) and removes it step by step until a picture appears that matches your prompt.

Why you'll hear it: in lessons and news about how image tools work. "Noise" here does not mean sound; it means random dots. How Diffusion Models Work explains it with pictures.

Video words

Frame: one still picture in a video. Videos usually show 24 to 60 frames per second (fps).

Clip: a short piece of video. Many AI video tools make clips of a few seconds at a time.

First frame / last frame: some tools let you upload the starting picture, the ending picture or both, and the AI fills in the movement between them.

B-roll: extra footage that plays over a voice to show what is being talked about.

Render / export: turning your edited project into a final video file.

Voice and audio words

Voiceover: a voice talking over a video without the speaker on screen.

Voice cloning: making an AI copy of a real person's voice from recordings of it. Only clone a voice with that person's clear permission.

Dubbing: replacing the speech in a video with another language, often keeping a similar voice.

Lip sync: making mouth movements in a video match new audio.

Stem: one separate part of a song, like only the vocals or only the drums.

AI Voice & Audio teaches these tools hands-on.

Safety and honesty words

Watermark: a mark that shows where content came from. It can be visible, like a logo in the corner of a free AI video, or invisible, hidden in the file so software can detect that AI made it.

Deepfake: realistic fake video, audio or images of a real person doing or saying something they never did. Making deepfakes of real people without consent can be harmful and, in many places, illegal.

AI label / disclosure: a note telling viewers that content was made or changed with AI. YouTube and Instagram ask creators to label realistic AI content.

Commercial use: using something to make money, like ads, client work or selling it. Many free AI plans do not allow it, so check the tool's terms.

A quick reference

Key Takeaways

• Generative AI creates new text, images, audio and video; "text-to-X" names tell you what goes in and what comes out.
• Aspect ratio is the shape, resolution is the sharpness, and upscaling makes images bigger and sharper.
• Diffusion tools turn random noise into a picture step by step.
• TTS turns text into a voice; voice cloning copies a real voice and needs consent.
• Watermarks, AI labels and commercial-use rules help you use creative AI honestly.