# Chat Box Source: https://learn.diffusion.studio/docs/chat-box The chat box configures how your AI conversations work.

## Controls at a Glance When enabled, the agent will create a step-by-step task list before acting on your request. After each step, it will decide what to do next. Recommended for complex, multi-step tasks. For simple tasks like generating an image, it's faster to keep this off. The agent has access to 16+ tools. Here you can choose which tool it's allowed to use. Set the aspect ratio for generated images and videos. Not all aspect ratios work with every tool - if your chosen, one isn't supported, the closest available option will be used.Set to `Auto` to let the system decide based on your prompt or chat context. Sets the length of generated videos.If your requested length isn't supported,it will use the nearest available option. When enabled, generated videos will include audio. This takes slightly longer than generating without audio. > Not all aspect ratios or durations are available for every tool. If an option isn’t supported, the closest available option will be used automatically. *** ## Optimizing prompts Within the Tool configuration, you'll also find a switch for `Prompt Optimization`.

When enabled, the agent uses an LLM to restructure your prompts so they work best with each tool. In other words, it automatically applies best practices for prompt writing. > Tip: For high-fidelity prompting and precise prompt engineering, keep this disabled so your exact wording is preserved. *** ## Defining tool the access At the bottom of the same dropdown, you'll see `Configure agent`.

The agent can access multiple models for the each tool. E.g. the tool "Text to image" might use `Flux.1 Kontext Max` or `GPT Image 1` . Click on a model to select it, **double click to make it your primary selection** > There’s no single “best” model — it depends on your style preferences, desired output speed, and budget. As a research company, we continuously monitor the generative space and make state-of-the-art models available whenever possible. ## Next Steps See end-to-end workflows. Create a Game Trailer # Camera & Composition Prompts Source: https://learn.diffusion.studio/docs/examples/camera-prompts Complete reference guide for cinematic camera shots, angles, and composition keywords for prompting. ## Shot Types ### Extreme Close-up

Description: Very tight framing on a detail (eye, hand, object).
Example Prompt: An \[Extreme close-up] of a warrior's eye glowing with fire, cinematic, dark fantasy

### Close-up

Description: Focuses on a face or object for detail and emotion.
Example Prompt: A \[Close-up shot] of a knight's face wet with rain, cinematic, 16:9

### Medium Shot

Description: Shows subject from waist up. Balanced framing for dialogue or body language.
Example Prompt: A \[Medium shot] of a cyberpunk hacker typing at neon screens

### Full Shot

Description: Captures the entire subject head-to-toe.
Example Prompt: A \[Full shot] of a samurai standing in neon rain, cinematic, 16:9

### Wide Shot

Description: Includes the subject and surrounding environment.
Example Prompt: A \[Wide shot] of a lone traveler crossing a desert with ruins, cinematic

### Extreme Wide Shot

Description: Subject is tiny against a massive environment.
Example Prompt: An \[Extreme wide shot] of an army marching through vast mountains, epic scale

### Establishing Shot

Description: Introduces the setting or environment.
Example Prompt: An \[Establishing shot] of a futuristic city skyline glowing with neon

*** ## Camera Angles ### Low-angle Shot

Description: Looking up at subject. Suggests power, dominance, or epic scale.
Example Prompt: A \[Low-angle shot] of a colossal dragon roaring against storm clouds

### High-angle Shot

Description: Looking down. Suggests weakness or vulnerability.
Example Prompt: A \[High-angle shot] of a child holding a lantern in a dark forest

### Eye-level Shot

Description: Neutral perspective at subject's eye level.
Example Prompt: An \[Eye-level shot] portrait of a medieval queen in candlelight

### Bird's-eye View

Description: Directly from above, almost vertical.
Example Prompt: A \[Bird's-eye view] of a city under attack, cinematic

### Worm's-eye View

Description: Extreme low shot looking upwards. Maximizes scale.
Example Prompt: A \[Worm's-eye view] of a skyscraper glowing in neon

### Dutch Angle

Description: Tilted frame. Creates tension or unease.
Example Prompt: A \[Dutch angle] view of a futuristic alley glowing with neon lights

### Over-the-shoulder Shot

Description: Shows perspective from behind a character's shoulder.
Example Prompt: An \[Over-the-shoulder shot] of a detective overlooking a crime scene

*** ## Camera Movement ### Static Shot

Description: Camera does not move. Clean framing.
Example Prompt: A \[Static shot] of an astronaut floating in space, cinematic

### Slow Push-in

Description: Subtle forward camera move. Builds intensity.
Example Prompt: A \[Slow push-in] on a sorcerer chanting in a ruined temple

### Zoom-in / Zoom-out

Description: Optical zoom for focus or scale.
Example Prompt: A \[Zoom-out] revealing a massive spaceship above a city

### Dolly Shot

Description: Camera tracks subject smoothly.
Example Prompt: A \[Dolly shot] following a knight running across a battlefield

### Crane Shot

Description: Camera moves up/down dramatically, like on a crane.
Example Prompt: A \[Crane shot] revealing a medieval castle from above the forest

### Tracking Shot

Description: Camera follows subject's movement. Dynamic and cinematic.
Example Prompt: A \[Tracking shot] of a car speeding through neon streets, cyberpunk

### 360-degree Shot

Description: Camera circles completely around subject.
Example Prompt: A \[360-degree shot] around a warrior in battle, cinematic

*** ## Composition & Framing ### Rule of Thirds

Description: Subject aligned with grid intersections. Balanced and cinematic.
Example Prompt: A portrait framed with \[Rule of thirds], cinematic lighting

### Centered Composition

Description: Subject centered. Creates symmetry and focus.
Example Prompt: A \[Centered composition] of a throne in a vast hall, epic cinematic

### Symmetrical Composition

Description: Balanced, mirrored framing. Harmonious or surreal.
Example Prompt: A \[Symmetrical composition] of a cathedral interior glowing with candles

### Silhouette Shot

Description: Subject shown in shadow against strong light.
Example Prompt: A \[Silhouette shot] of a cowboy against a sunset horizon, cinematic

### Foreground Focus

Description: Object in front is emphasized, background blurred.
Example Prompt: A \[Foreground focus] of a sword in the sand, blurred desert background

### Deep Focus

Description: Foreground and background both sharp.
Example Prompt: A \[Deep focus] shot of a soldier in ruins, entire city in sharp detail

*** ## Lighting Modifiers ### Backlight

Description: Light shines from behind subject. Creates dramatic rim.
Example Prompt: A \[Backlight] shot of a dancer on stage, glowing outline

### Rim Light

Description: Thin glowing outline of subject edges.
Example Prompt: A portrait with \[Rim light] highlighting the silhouette, cinematic

### Rembrandt Lighting

Description: Classic triangle light on cheek. Dramatic portraits.
Example Prompt: A portrait with \[Rembrandt lighting], cinematic photography style

*** ## Best Practices * **Combine scene + camera tag + style** for best results * **Use 2–3 camera tags max** per prompt to avoid conflicts * **Compare results:** try the same scene with and without a tag * **Tags can stack** for cinematic results: * `[Extreme wide shot] + [Bird's-eye view]` → epic landscapes * `[Close-up] + [Dutch angle]` → intense character drama * `[Tracking shot] + [Low-angle]` → dynamic action sequences > Pro Tip: Start with a basic prompt, then add one camera technique at a time to see how it affects your results. This helps you understand each technique's impact. # Create a Game Trailer Source: https://learn.diffusion.studio/docs/examples/create-game-trailer Learn how to create an engaging game trailer using Diffusion Studio Pro and AI-powered generation *** ## What You'll Make

A 40-second sci-fi military trailer with 4 scenes: underground metro, city chaos, stairwell climb, and rooftop overwatch. **Time needed**: 30-45 minutes *** ## Pick Your model Lets start int he canvas view and configure our tools.

### Option A: Manual * Find FLUX\_KONTEXT\_MAX in the list * If you don't see it, click "Configure Agent" and search for it * Double-click to select it ### Option B: Auto-select * Turn on "Auto-Select" at the top * Tell the agent: "use flux-kontext-max model" ## Settings * Turn OFF prompt optimization (prompts here are optimized; enable it for your own projects). * Aspect Ratio: 16:9 * Duration: 10 seconds * Audio: Auto-select ON *** # Generate Images You'll create 4 base images first, then animate them. Each gets 4 variations - pick your favourite.

## Scene 1: Underground Metro Copy this prompt: ```json { "prompt": "Photorealistic third-person shooter perspective from the POV of a soldier descending into a futuristic metro station. The squad leader moves ahead down wet steps into a dim corridor filled with flickering red lights, steam, and debris. The foreground shows the following soldier's rifle partially raised. Subtle reflections on flooded rails, scattered trash, sparks in the distance. HUD: lower-left mini-map, top-right health and ammo, faint crosshair in center. Same camera angle, same gear, cinematic neon-blue and orange palette with consistent lighting and tension.", "negative_prompt": "bad anatomy, unreadable HUD, low detail, weird lighting, distorted perspective", "seed": 87 } ``` ## Scene 2: City Street

```json { "prompt": "Photorealistic third-person shooter screenshot showing two soldiers advancing through a rain-soaked futuristic city street at night. The camera is positioned from the POV of the second soldier, with their rifle partially visible in the lower foreground. Ahead, a matte-black armored soldier (the squad leader) walks through neon reflections, flaming wrecks, and swarming alien drones overhead. Explosions light up the wet street and Chinese neon signs. HUD overlays: bottom-left mini-map, top-right ammo and health, subtle crosshair. Cinematic rain, motion blur, high-fidelity lighting, blue-orange contrast.", "negative_prompt": "text overlays, watermarks, broken anatomy, missing limbs, distorted weapons, cartoon artifacts", "seed": 87 } ``` ## Scene 3: Stairwell ```json { "prompt": "Photorealistic third-person shooter camera angle from the POV of a soldier ascending a dark, narrow stairwell. The main soldier climbs slowly ahead with a flashlight beam cutting through the dust. Part of the rifle and gloves of the following soldier are visible in the lower frame. The stairwell is grimy concrete, with red emergency lights flickering and wet steps reflecting the flashlights. HUD overlays: mini-map, ammo, and subtle mission objective. Shadows flicker with muzzle movement. Matches same urban combat style, lighting, and equipment from city street scene.", "negative_prompt": "cartoon effects, broken walls, disfigured limbs, poor shadow detail, low realism", "seed": 87 } ``` ## Scene 4: rooftop ```json { "prompt": "Photorealistic third-person shooter view from the POV of a soldier standing on a high rooftop, partially showing their weapon in the lower frame. Ahead, the squad leader in matte-black tactical armor stands at the edge of the rooftop, looking out over a futuristic war-torn city in flames. Neon-lit towers rise into smoke-filled skies; alien drones swarm above, and explosions light distant streets. Rain falls steadily. HUD shows mini-map lower left, crosshair centered, ammo and status top-right. Same dark, cinematic vibe as city chaos shot, with consistent character and lighting.", "negative_prompt": "distorted arms, broken gear, low resolution, text glitches, incorrect lighting", "seed": 87 } ``` ## Animate Everything Now switch to the animation mdoel and bring your images to life.

## Switch Models Switch to **SEEDANCE-1-PRO** (same method as before) ## Animate Each Scene ### Scene 1 - Underground: ```json { "prompt": "Dimly lit underground metro tunnel. Third-person POV behind soldier, following another down wet stairs into a flickering corridor. Water on tracks reflects red lights. Steam bursts from vents, sparks fall from ceiling. Camera dips slightly on each step. Same soldier armor and gear. Subtle flashlight beam moves. Realistic ambient effects.", "duration_seconds": 10, "seed": 87, "audio": "Dripping water, footstep echoes, flickering light buzz, distant subway screech, radio ping" } ``` ### Scene 2 - Street: ```json { "prompt": "Futuristic war zone at night, rain falling on neon-lit street. Cinematic third-person shooter view from behind a soldier, walking slowly through flaming wreckage and broken cars. The main soldier ahead walks forward with weapon ready. Drones fly overhead, glowing Chinese signs reflect on wet ground, small explosions light the fog. Camera slightly sways with steps. Consistent with sci-fi military game. Subtle motion blur, rain particles.", "duration_seconds": 10, "seed": 87, "audio": "Heavy rain, low thunder, distant drone hum, boots on wet pavement, subtle gunfire in distance, faint radio static" } ``` ### Scene 3 - Stairwell: ```json { "prompt": "Dark stairwell inside ruined building. Third-person view from soldier's POV, showing rifle and hands. Ahead, the main soldier ascends slowly with flashlight scanning steps. Red emergency lights flicker on walls. Concrete is wet and reflective. Dust particles in beam. Camera shakes slightly with steps. Ultra-realistic indoor combat scene.", "duration_seconds": 10, "seed": 87, "audio": "Boots on wet stairs, flashlight hum, faint breathing, dripping water, distant rumble upstairs" } ``` ### Scene 4 - Rooftop: ```json { "prompt": "High above a burning futuristic city at night. POV from behind a soldier with rifle visible, standing on wet rooftop. The main soldier ahead scans horizon filled with smoke, neon-lit towers, and flying drones. Flames flicker far below. Camera slowly drifts to the right, wind moves loose wires. Blue-orange dusk light and cinematic rain particles.", "duration_seconds": 10, "seed": 87, "audio": "Wind gusts, faint city rumble, distant sirens, drones buzzing, rooftop puddle splash" } ``` ## Put It Together

## Approve Your Videos For each animation: 1. Click the video to open sidebar 2. Check the quality 3. Click **"Approve"** to send to Timeline 4. Repeat for all 4 scenes ## Go to timeline'

Head to **Timeline** - you'll see all 4 scenes ready to go. *** ## Pro Tips * Pick any seed (e.g., 87) and use the same one for all image generations * Same camera angle (third-person POV) * Blue-orange lighting * Minimal movement * 10 second per sccene * 16:9 aspect ratio * 24 FPS * Choose images that look good Together * Make sure lightning flows between scenes * Keep HUD elements consistent *** # You're Done! You now have a 40 - second game trailer with: * Consitent visual style * Professional audio * Smooth scene transitions * Cinematic quality *Same as the preview video at the top!* *** ## Template Prompts Here are some reusable prompts for game trailer creation: ```text Create a dramatic opening shot of [game setting] with [mood/lighting] and [specific elements]. The scene should establish the game's atmosphere and genre. ``` ```text Generate a [duration]-second clip of [action description] with [specific effects] and [camera movement]. The scene should be fast-paced and exciting. ``` ```text Create a dramatic close-up of [character description] with [lighting] and [expression]. The shot should convey the character's personality and role. ``` ```text Add a [transition type] effect between [scene A] and [scene B] with [duration] and [easing]. ``` ## Next Steps ### Try These Next **Mix it up:** * Day vs night versions * Different weather (snow, fog) * New camera angles **Go deeper:** * Keep the same soldier throughout * Add story elements * Layer multiple audio tracks **Share it:** * Optimize for social media * Use in your game * Create marketing versions ### Quick Tips Save your working prompts for future use and iterations. Generate multiple versions at once to compare options. Always check quality before approving final renders. Think about audio when planning scenes for better integration. > **Keep everything consistent** - That's it! You've got everything you need to make awesome game trailers. 🎬 # Ghostly Fashion Model — Images & Animations Source: https://learn.diffusion.studio/docs/examples/fashion-model This guide shows how to create a **series of cinematic fashion shots** featuring a female model with a sharp bob haircut and a long black leather coat, set against a solid red backdrop.\ We'll cover both **image generation** and **animation** using Diffusion Studio Pro.

*** ## Model & Settings

Model: Flux.1 Kontext Max
Aspect Ratio: 16:9
Prompt Optimization: Disabled (prompts are pre-optimized)

Model: Seedance 1.0 Pro
Duration: 5 seconds
Aspect Ratio: 16:9
Audio: Off
Prompt Optimization: Disabled (prompts are pre-optimized)

*** ## Image Generation Prompts > Go to the Canvas page, select the correct settings as described above, put in the prompt, and let the magic begin: ### 1. Fashion Portrait Fashion photo of a female model against a solid matte red backdrop. Wearing a long, black leather coat. She has a sharp black bob haircut with bangs and a neutral expression. Studio lighting is soft and even, with slight shadows for depth. Realistic skin textures, high-fashion editorial mood. Cinematic composition, ghostly presence, faded film effect, with her hair and figure in strong motion blur.

*** ### 2. Close-Up > Select the generated image as a reference for other prompts to keep details consistent. Close-up fashion portrait of the same female model against a solid matte red backdrop. She wears a long black leather coat, visible only at the collar and shoulders. Her sharp black bob haircut with straight bangs covers her forehead. Her intense eyes and neutral expression are clearly visible. Soft, even studio lighting with subtle shadows enhances the realism of her skin texture. High-fashion editorial mood, cinematic composition with ghostly presence. Faded film effect, strong motion blur streaking through her hair and facial contours.

*** ### 3. Cinematic Overhead with Horse > Don't forget to select the first generated image as a reference to keep consistency. Cinematic overhead shot from high above. The female model lies stretched out on a solid matte red backdrop, wearing a long black leather coat. Her sharp black bob haircut with bangs frames her face as she gazes upward. Beside her, a majestic white horse lies gracefully, mane flowing across the red surface. Soft, even studio lighting with subtle shadows emphasizes textures of the coat, her skin, and the horse's mane. High-fashion surreal atmosphere, ghostly presence, faded film effect. Slight motion blur in the hair and mane creates a dreamlike cinematic composition.

*** ## Animation Prompts Once images are generated, use the animation settings above to convert them to video.\ Here are the ready-to-use prompts: > Don't forget to select the correct image that you want to animate. ### 1. Model with Horse Interaction \[Static medium shot] in a professional photo studio with a solid red background. A female model with a sharp black bob haircut and bangs, wearing a long black leather coat, gently turns toward a white horse beside her. She slowly raises her hand and delicately touches the horse's head near its cheek. The horse subtly moves its head toward her touch, mane shifting slightly. Background stays solid red, camera fixed on model. Realistic cinematic lighting with soft highlights and controlled shadows.

*** ### 2. Hair Flip Animation \[Static medium shot] in a professional photo studio, female model centered in frame. Sharp black bob haircut with bangs, long black leather coat. She subtly tilts her head while flipping her hair, causing the bob to swing dynamically. Realistic cinematic lighting with soft highlights and shadows. Neutral, unchanging background.

*** ### 3. Sharp Head Turn \[Static medium shot] in a professional photo studio. Female model with sharp black bob haircut and bangs stands centered, wearing a long black leather coat. She quickly turns her head toward the camera, her bob swinging sharply before settling. Camera remains fixed and focused. Realistic cinematic lighting with soft, controlled shadows. Neutral, unchanging background.

*** ## Results By following this workflow, you can generate and animate: * A **static editorial portrait** * A **dynamic close-up with ghostly motion blur** * A **surreal horse interaction scene** All unified in the same **cinematic, high-fashion ghostly style**. # Prompt Templates for Diffusion Studio Pro Source: https://learn.diffusion.studio/docs/examples/prompt-templates Ready-to-paste prompt templates for T2I, T2V, I2I, and I2V in Diffusion Studio Pro > Copy any block below into the **chat box**. Replace bracketed fields as needed. *** ## Text-to-Image (T2I) ### Base Template ```json { "model": "flux.1", "seed": 42, "style": "studio", "prompt": "[describe the subject clearly and concretely]", "negative_prompt": "deformation, extra limbs, text warping, severe noise", "guidance": 7.5, "steps": 28, "sampler": "DPM++ 2M Karras", "resolution": "1024x1024", "aspect_ratio": "1:1", "lighting": "soft diffused studio", "camera": { "focal_length_mm": 85, "aperture_f": 2.8 }, "postprocess": { "sharpen": 0.0, "film_grain": 0.0 }, "output": "png" } ``` ### Example: Ferrari Model ```json { "model": "flux.1", "seed": 56, "style": "studio", "prompt": "Ferrari 296 GTB in Rosso Corsa, seamless light-grey studio, soft top light, crisp paint reflections, black wheels with yellow calipers, centered three-quarter view", "negative_prompt": "cartoon, matte paint, heavy grain, extra wheels, warped logos", "guidance": 7.0, "steps": 30, "sampler": "DPM++ 2M Karras", "resolution": "1536x1024", "aspect_ratio": "3:2", "lighting": "soft box, rim fill", "camera": { "focal_length_mm": 105, "aperture_f": 4.0 }, "postprocess": { "sharpen": 0.05, "film_grain": 0.0 }, "output": "png" } ``` ## Quick Workflows ```text Set the project resolution to [WIDTH]x[HEIGHT] and conform all clips. Keep aspect ratio; letterbox/pillarbox if required. ``` ```text Export the current timeline as MP4 (H.264 video, AAC audio), bitrate [20 Mbps], audio [48 kHz stereo, 192 kbps]. ``` ```text Conform the sequence to [24] fps. Use optical flow where needed; keep audio pitch. ``` ## Text-to-Video (T2V) ### Base Template ```json { "model": "seedance-1-pro", "prompt": "[describe the action and setting in 1–2 sentences]", "negative_prompt": "object drift, relighting, logo changes, text morphing", "duration_s": 5, "fps": 24, "seed": 42, "aspect_ratio": "16:9", "camera": { "yaw_total_deg": 0, "pitch_total_deg": 0, "roll_total_deg": 0, "zoom": 0.0 }, "motion_strength": 0.35 } ``` ### Example: Runner ```json { "model": "seedance-1-pro", "prompt": "Night city street, lone runner cutting through light rain; neon reflections on wet asphalt; moody fog; cinematic shallow depth of field", "negative_prompt": "face distortion, clothing flicker, relighting, color shift", "duration_s": 10, "fps": 24, "seed": 77, "aspect_ratio": "16:9", "camera": { "yaw_total_deg": 8, "pitch_total_deg": -2, "roll_total_deg": 0, "zoom": 0.0 }, "motion_strength": 0.4 } ``` ### Quick Workflows ```text Stabilize the selected clip with optical flow. Minimize crop; preserve original fps and audio pitch. ``` ```text Time-remap the selected clip to [80%] speed with optical-flow interpolation; ripple maintain duration. ``` ```text Reduce noise slightly and add mild detail recovery; avoid halos and hue shifts. ``` ## Image-to-Image (I2I) ### Base Template ```json { "model": "flux.1", "seed": 101, "prompt": "[describe your intended change while keeping the subject]", "negative_prompt": "identity change, extra objects, background swap, text warp", "image_strength": 0.35, "preserve_color": true, "sampler": "DPM++ 2M Karras", "steps": 24, "resolution": "1024x1024" } ``` ### Example: Outfit Change ```json { "model": "flux.1", "seed": 88, "prompt": "Same person and pose, swap to black suit (matte), clean studio, neutral background, crisp edges", "negative_prompt": "face drift, new jewelry, background change, extra props", "image_strength": 0.30, "preserve_color": false, "steps": 26, "resolution": "1536x1024" } ``` ### Quick Workflows ```text Apply the [LoRA name] style at weight [0.7]. Keep composition and identity unchanged. ``` ```text Blur the background only; keep subject sharp. Radius [12] px; maintain color and contrast. ``` ```text Posterize to 5 color bands; add subtle film grain 5%; keep edges clean. ``` ## Image-to-Video (I2V) ### Base Template ```json { "model": "seedance-1-pro", "prompt": "Preserve the uploaded image exactly as is; no edits or new elements. Subject-lock stabilization. Shallow depth of field, soft studio lighting, 180° shutter. Smooth ease-in-out.", "negative_prompt": "subject deformation, texture crawl, reflection drift, color shift, zoom, dolly, jitter, flicker, background change", "duration_s": 8.0, "fps": 24, "seed": 42, "image_guidance": 0.995, "motion_strength": 0.06, "camera": { "yaw_total_deg": -4.0, "pitch_total_deg": 0.0, "roll_total_deg": 0.0, "zoom": 0.0, "dolly_frac": 0.0, "easing": "ease-in-out" } } ``` ### Example: Orbit (Left) ```json { "model": "seedance-1-pro", "prompt": "Freeze the subject 100%; car must not change. Camera-only ultra-slow orbit left; identical lighting and reflections.", "negative_prompt": "subject movement, relighting, color shift, zoom, dolly, jitter, flicker", "duration_s": 8.0, "fps": 24, "seed": 99, "image_guidance": 0.995, "motion_strength": 0.06, "camera": { "yaw_total_deg": -4.0, "pitch_total_deg": 0.0, "roll_total_deg": 0.0, "zoom": 0.0, "dolly_frac": 0.0, "easing": "ease-in-out" } } ``` ### Quick Workflows ```text Create a seamless 2-beat camera loop from the current move; total duration 5 s; match in/out velocity. ``` ```text Add a gentle push-in: 2% dolly over 5 s; no zoom; keep framing centered. ``` ```text Freeze the last frame for 0.75 s and add a soft audio sting marker. ``` ## Quick Tips * Use a fixed seed to reproduce results when iterating. * For rigid subjects (cars/products), set image\_guidance 0.98–0.995 in I2V. * Keep negative prompts short but protective (logos, text, background). # Timeline Source: https://learn.diffusion.studio/docs/timeline Arrange, edit, and refine your media with precision using the timeline. Timeline overview demo

## Layout Video tracks sit at the top, audio tracks at the bottom. Each track shows clips with thumbnails and duration so you can see your sequence clearly. The ruler shows seconds and frames. Click anywhere on it to move the playhead to that time. The playhead marks the current position. Drag it to scrub or place new clips exactly where you want them. *** ## Media Pool ### Adding Assets There are three ways to move assets into the Media Pool: Approve from Canvas

1. **From Canvas** – Click on an image or video in the Canvas, open the sidebar, and click **Approve**. Approve from Chat

2. **From Chat** – On the Timeline page, generated media has an **Approve** button under it. Clicking it adds the asset. Upload from Local

3. **From Computer** – Click the upload arrow at the top of the Media Pool and select your files. ### Deleting Assets Delete from Timeline

* **From Timeline** – Select a clip and press **Backspace** or click the trash icon.

* **Delete a Track** – Open the track menu (⋮ icon) and choose **Delete track**. * **From Media Pool** – Right-click an asset and select **Delete**. * **From Canvas** – Select an element and press **Backspace**. *** ## Key Features ### Adding Clips Add Clips

Drag clips from the Media Pool into the Timeline. You can move several files at once. ### Trimming & Splitting Split Clips

Drag a clip’s edge to trim. To split, place the playhead and click the scissors icon or press `Cmd + B`. ### Clip Properties

Select a clip to open its settings in the Config Panel. Adjust timing, scale, effects, and audio. ### Clip Actions

For audio, use **Silence Remove** to cut quiet sections automatically. ### Export Frame

To export a frame, place a playhead and click the camera icon, frame will be directly exported and also added to the media pool. *** ## Track Controls Each track has its own icons for managing visibility, gaps, and order:

* **Move Track (Handle)** – Drag a track up or down.

* **Close Gaps (⇔)** – Remove empty spaces between clips.

* **Toggle Visibility (Eye)** – Show or hide the entire track.

* **Track Options (⋮)** – Set a track as primary or delete it. Resize Page

* **All panels** are resizable and toggleable. Hide bars

* **Hide** them individually—or hide all at once for a distraction-free view. *** ## Next Steps Bring in media from chat, uploads, or your computer. Trim, move, and split clips to structure your video. Use track controls, transitions, and effects to polish. Save frames or render the full timeline. # Welcome Source: https://learn.diffusion.studio/docs/welcome Start your learning journey with beginner-friendly resources that will have you creating in no time. When you open Diffusion Studio Pro for the first time, make sure to select a backup folder. All generated assets will be saved there.

> **Be aware:** Assets you created in previous projects will no longer be available if you change the backup folder later. Let's jump into the editing UI by creating a new project.

## Interface Overview The Diffusion Studio Pro interface is organized into 5 main panels, each serving a specific function: Common video editing tools such as transitions, captions, shapes, and more The media panel, where all assets for your video are stored The canvas, showing the current state of your composition; it can be panned and repositioned A non-linear timeline that supports an infinite number of layers and clips A chat sidebar for generating assets or manipulating the composition

The chat, timeline, and edit panels can be hidden anytime for a better timeline experience. In the status bar, you'll also find the option to switch to the canvas view.

## When to Use Which View? Best for generating single assets (images, videos, music) in parallel. Chats are non-blocking, i.e. you can submit prompts without waiting for the result of the previous prompt. The canvas also lets you organize assets in whatever structure fits your project. Best for composing assets into a full video. The chat can access the current state of the composition and supports multi-step instructions. > **Rule of thumb:** use the timeline for agentic workflows and the canvas for single-asset generation. ## Next Steps Choose a location on your computer where all generated assets will be stored. Start a new project to explore the interface and begin creating content. Discover how to use AI-powered generation through the chat interface. Learn more about Chat Box Check out our examples to see what's possible with Diffusion Studio Pro. Create a Game Trailer