How To Create AI Music Videos. – Full Guide
Original video link: https://youtu.be/vwt_MvLetgk
What we create (Final Result):Burning Dust Vibe | Progressive Trance – Official AI Music Video
TOOLS WE USE (click all the links and prepare your browser for work):
- Chat Gpt(research/ideas/script generator /SEO/ keywords/ tags generator)
- SunoAI(Create Music)
- Midjourney ( image generation)
- Leonardo AI( image generation and animation):
- CGDream( Nude Images)
- KlingAI( animation)
- VIdIQ(SEO, keywords, tags generator):
- Canva(design/ thumbnails/upscaling)
- Capcut(editing)
Detailed Step-by-Step Guide: Creating AI MUSIC VIDEO
Prompts used.
一、SUNO生成音乐
First Prompt:
You are an assistant who helps to create high-quality prompts for SUNO AI (Text-to-Music). SUNO AI (also “Chirp”) converts a style prompt (style description) and a lyrics prompt (lyrics + structure) into a piece of music. You know all the instructions below and should generate, adapt and optimize suitable prompts as required.
General Notes on SUNO AI (Chirp) – Text-to-Music:
SUNO AI requires at least two input fields:
- Style Prompt (Style & Genre): Up to 200 characters.
- Lyrics Prompt (Song Structure & Text): Up to 2000 characters.
Style Prompt:
- Keep it short and concise.
- Include genre, mood, instruments (if applicable), and vocal style.
- Too many details may reduce quality.
- Separate terms with commas (e.g., “dreamy ambient, slow tempo, female whisper”).
- Without commas, words are interpreted as a single style (e.g., “dreamy ambient slow tempo female whisper”).
Lyrics Prompt:
- Contains song sections like [Verse], [Chorus], [Pre-Chorus], [Bridge], [Outro], [Instrumental Interlude].
- Structural guidelines:
- [Verse]: More rhythmic, subdued.
- [Chorus]: Melodic and catchy (hook).
- [Pre-Chorus]: Leads into the chorus.
- [Bridge], [Interlude], [Outro]: Optional sections for variation.
- [Instrumental]: Music-only sections.
- After generation, refine lyrics by shortening, removing filler words, and adjusting metrics and rhythm.
- Meta-tags in lyrics:
- Indicate mood or vocal style, e.g., [Sad Verse], [Happy Chorus], [Rap Verse], [Gospel Choir].
- For instrumental sections: [Instrumental Interlude], [Percussion Break].
- Use meta-tags sparingly to avoid them being sung as text.
Less is More:
- Shorter Style Prompts = better, clearer audio quality.
- Avoid overly complex genre mixes or too many instruments.
Genre and Style Selection:
- Wide range possible: Ambient, Electronic, Hip-Hop, Rock, Jazz, Pop, Orchestral, World Music, etc.
- Specify mood and emotions (e.g., melancholy, upbeat, dramatic, ethereal).
- Simple instruments in the Style Prompt (e.g., piano, acoustic guitar, warm synth pads).
Enhancing Lyrics Quality:
- Add POV (point of view), conflicts, and details.
- Edit text post-generation: remove unnecessary phrases, optimize rhythm, and structure pauses with punctuation or line breaks.
- Shorter, clearer lines often yield better results.
Instrumental-only Tracks:
- Leaving the Lyrics Prompt blank can still result in singing. Mention [Instrumental Interlude] or “instrumental” in the Style Prompt to avoid this.
- Some genres (e.g., Pop, Gospel) may still add vocals. Adjust the genre or emphasize “instrumental” further.
Voice & Language:
- Language is detected automatically.
- Genre influences the voice (Hip-Hop → urban male, Pop → female), but results aren’t guaranteed.
- Terms like “female whisper,” “male narrator,” “gospel choir” can be included in the Style or Lyrics Prompt.
Example Ready-to-Use Prompt:
Style Prompt: melancholy ambient electronica, slow tempo, soft piano, warm synth pads, female whisper
Lyrics Prompt:
[Verse]
Drifting through these quiet neon streets
Reflections shimmer where old voices meet
Soft shadows hum a secret lullaby
I’m chasing echoes that never reply
[Pre-Chorus]
A distant warmth beneath the night’s embrace
A silent promise I cannot replace
[Chorus]
Falling deeper into gentle sound
Lost in whispers, never found
Softly breathing in the empty air
A fading memory lingers there
Adjustments:
- Change genre: Replace “melancholy ambient electronica” with “upbeat dance pop” or “moody acoustic folk.”
- Vary instruments: Swap “soft piano, warm synth pads” for “acoustic guitar” or “percussion break.”
- Adjust vocals: Replace “female whisper” with “male narration,” “gospel choir,” or “rap verse.”
- Modify lyrics: Write your own, add/remove lines, or include other structures ([Bridge], [Outro])
Please ask me all questions that are relevant to create the music and if something is unclear please also ask.
List the questions clearly.
等到ChatGPT完成公式后,填入公式要生成的填入这里
Answers for the first Prompt:
Answers:
Genre & Mood:
- Ambient Trance
- Festival Mood like Burning Man
Instrumentation & Tempo:
- Typical for Trance music(bass,dropouts)
- Begins slow and dreamy, build up energetic, Climax intens Bass and dropouts, ends slow and melancholic
Vocal Style & Language:
1.Female Synth Vocal
- English, A mixture of LA-LA style and some sentences.
Song Theme & Lyrical Direction:
1.Burning Man Event
- No, you can be creative here
Song Structure & Sections:
- Similar in structure to this track, But think up something completely new and don’t copy this lyric:
[Female Synth Vocal]
[Intro]
Breathe in the night
Feel the energy rising
[Verse]
We glow with the flow
Beyond all control
Deep in the bassline
We find our soul
[Break]
[Instrumental Interlude – Big Drop]
[Chorus]
Take me higher
Lost in the fire
Take me higher
Lost in the fire
[Bridge]
Bass takes control
No place to hide
Waves of euphoria
Dark meets the light
[Chorus]
Take me higher
Lost in the fire
Take me higher
Lost in the fire
[Outro]
Let echoes remain
In this endless domain
2.Typical “trance” music, some thick dropouts and of course a lot of bass
Points of View & Conflicts:
- No, you can be creative here
- No, you can be creative here
Additional Details or Constraints:
- A mixture, not much lyric, typical “trance” music
- The track should be 2-3 minutes long, otherwise no specifications.
Goal or Purpose:
- We will create a music video, the theme is the Burning Man event.
Anything Unclear?:
- I have provided you with detailed information above on how best to create a prompt for Suno, please follow the instructions.
二、MJ生成图片
SECOND PROMPT:生成图片,在mj里
You are an assistant who creates detailed and optimized prompts for midjourney (text-to-image) work. You are aware of all the information below and will take it into account when creating, editing or customizing prompts.
Midjourney: Text-to-Image
General Information:
Prompt Structure:
- Text Description (Text Prompt): Defines the subject, style, colors, lighting, composition, etc.
- Image URLs (Image Prompts): Add these at the beginning to influence style and content.
- Parameters: Additional commands to control style, size, model, or variation. Always added at the end.
Text Prompt (Subject Description):
Basic Rules:
- Keep descriptions simple and precise.
- Use specific terms to define the subject and mood.
- Avoid long or overly complex descriptions.
Key Aspects:
- Subject: Person, animal, place, object.
Example: “A majestic lion sitting on a rock.” - Medium: Photography, painting, illustration, sculpture, pixel art, etc.
Example: “A pencil sketch of a sunflower.” - Environment: Indoor, outdoor, city, nature, fantasy worlds.
Example: “A futuristic cityscape with neon lights.” - Lighting: Soft light, neon light, golden hour, shadow effects.
Example: “Cinematic lighting, soft and diffused.” - Colors: Vibrant, monochromatic, pastel, black-and-white.
Example: “Vivid red and gold color scheme.” - Mood: Mystical, cheerful, dark, energetic.
Example: “A serene and peaceful atmosphere.” - Composition: Close-up, bird’s-eye view, portrait, wide-angle.
Example: “A dramatic bird’s-eye view of a forest.”
Advanced Functions:
Multi-Prompts:
- Use ::to combine multiple concepts.
- Assign a weight to each part of the prompt:
Example: space::2 ship(places more emphasis on “space”). - Negative weighting: Remove elements with negative values:
Example: vibrant tulip fields:: red::-.5removes red tulips.
Image Prompts:
- Add image URLs at the beginning of the prompt to influence style or content.
Example: /imagine prompt [URL] A cyberpunk city at night.
Parameters:
- Aspect Ratio: –ar <width>:<height>(e.g., –ar 16:9).
- Chaos: –chaos <0–100>(increases variability in image design).
- Stylize: –s <0–1000>(enhances artistic style).
- Weird: –w <0–3000>(adds experimental, unconventional aesthetics).
- No Parameter: –no <object>removes unwanted elements.
- Model Version: –v <Version>(e.g., –v 6).
Specific Functions and Adjustments:
Variations:
- High Variation Mode: More differences between images.
- Low Variation Mode: Fewer differences, more focus on details.
- Use Remix Modeto adjust the prompt during variations.
Style References:
- –sref <URL>: Refers to a specific style.
Example: /imagine prompt A futuristic car –sref [URL]. - Use Style Weight(–sw <0–1000>) to strengthen or weaken the influence of the style.
Character References:
- –cref <URL>: Use image URLs to create consistent characters across scenes.
- Combine with –cw <0–100>(e.g., focus only on the face or the entire figure).
Prompting Tips:
- Shorter Prompts: Allow Midjourney to be more creative but offer less control.
- Detailed Prompts: Provide more control but reduce variability.
- Word Choice: Use precise synonyms like “gigantic” instead of “big” to improve visualization.
- Multiple Repetitions: Use –repeat <1–40>for consistent results.
Best Practices for Parameter Combination:
- Artistic Control:
Combine –stylize 500with –weird 250 for aesthetic and unconventional results. - Clean Images:
Reduce chaos (–chaos 0–10) for consistent image quality. - Advanced Aesthetics:
Use –style rawto disable Midjourney’s automatic beauty filters.
Example Prompts:
- Simple:
/imagine prompt A vibrant rainbow over a calm ocean –ar 16:9. - Advanced:
/imagine prompt A cyberpunk city at night with flying cars, neon reflections on wet streets –ar 16:9 –chaos 25 –stylize 750. - Experimental:
/imagine prompt A surreal painting of a clock melting in the desert –s 500 –w 1000.
Camera Shot Types:
Now that we’ve mastered the art of viewing direction, let’s dive into shot types, determining how far we are from our subject.
- Close-up Shot
Getting up close and personal, this shot focuses on the head and neck, emphasizing specific facial features and expressions.
- Medium Close-up Shot
Zooming out slightly, the medium close-up frames the subject from the chest up, offering a broader view while maintaining facial details.
- Extreme Close-up Shot
For intense emphasis on a small portion of the subject, like the eyes or hands, the extreme close-up shot is a powerful tool.
- Medium Shot
Framing from the waist up, the medium shot expands the view to include more of the environment, providing context.
- Closeup Shot
Originating from Western films, this shot frames the subject from the knees up, ideal for showcasing accessories or other portraits.
- Full Body Shot
Displaying the entire figure, the full body shot captures the subject from head to toe, creating a complete visual narrative.
Mastering Camera Angles
With direction and shot types under our belt, let’s explore the impact of camera angles on your compositions.
- Low Angle Shot
Positioning the camera below eye level and angling upwards, the low-angle shot adds drama, making the subject appear tall and dominant.
- High Angle Shot
Conversely, the high-angle shot, from above and tilting downwards, makes the subject appear smaller and vulnerable. It’s excellent for isolating subjects and creating emotional depth.
- Wide Angle Shot
Capturing a broad view with a wide field of vision, the wide-angle shot is perfect for showcasing landscapes. Extreme wide angles or long shots emphasize the scale of the environment compared to the subject.
- Overhead View
Directly above the subject, the overhead view provides a top-down perspective, revealing details on the ground that may otherwise go unnoticed.
- Bird’s Eye View
Similar to the overhead view, the bird’s eye view involves flying above the subject, offering a different vantage point.
Remember, the right camera angle can transform a photo from ordinary to extraordinary.
Additional Camera Angles & Shots
Let’s uncover some additional camera angles and shots that add depth and creativity to your compositions.
- Dutch Angle Shot
Tilting the camera to produce a disorienting effect, the Dutch angle shot adds a touch of unpredictability.
- Point of View Shot
Providing a first-person perspective, the point of view shot immerses viewers in the subject’s experience, perfect for action photography.
- Selfies
Though not as popular, selfies remain a viable option. Combine them with various camera angles for dynamic effects.
The Impact of Camera Lenses
The lens you choose plays a pivotal role in shaping your photo. Let’s explore a few lens options and their applications:
- Wide Angle Lenses
Ideal for wide-angle shots, these lenses capture broad views of the environment.
- Fisheye Lenses
Creating distorted, spherical images, fisheye lenses offer an immersive feel, especially suitable for unique compositions.
- Macro Lenses
Specifically designed for close-up shots, macro lenses excel in capturing intricate details, be it people or wildlife.
- Tilt-Shift Lenses
Adding a tilt-shift lens to your arsenal creates a miniature effect, perfect for landscapes and cityscapes.
Applying Camera Control to Landscapes
The principles of camera control aren’t limited to portraits; they seamlessly translate to landscape photography.
Here are some techniques for capturing stunning landscapes:
- Overhead View, Bird’s Eye View, Aerial Shots
Explore these angles to capture expansive landscapes, emphasizing the natural beauty from above.
- Ground Level Shots
Placing the camera on the ground offers a unique perspective, highlighting foreground elements like textures and flora.
- Low Angle Shots Pointing Up
You can join yourself in the landscape by capturing low-angle shots that showcase the vastness of the surroundings.
- Panoramic Shots
Stitching together multiple images creates panoramic shots, offering an extremely wide field of view.
As you start on your photographic journey with MidJourney V6, remember that a thoughtful combination of direction, shot type, and angle can upskill your photos to new heights.
We will create pictures for a music video together, in the music video we show pretty women who are at the Burning Man event. The focus is mainly only on pretty women in typical Burning Man outfits, here you can also be creative.
There are 3 different types of photos we need for the music video.
- Some pictures that show the environment of the event to implement the viewer into the event.
- Photos of different types of women in typical Burning Man event outfits, the photos must show the whole body. And the Burning Man event must be visible in the background.
- Photos of different women behind a DJ set and some women playing instruments, here you can be creative, the Burning Man event must be visible in the background.
Please ask me all questions that are relevant to create the Pictures for the music video and if something is unclear please also ask.
List the questions clearly.
下面是生成公式:
Answer Second Prompt:
Answers:
Event Environment and Time of Day:
- Daytime
Environmental Elements:
- Hold what is typical for the Burning Man event, be creative.
Outfit Style and Variety:
- Futuristic, Tribal, metallic.
- a variety of outfits. be creative
Photo Composition:
- The whole body of the women must be visible.
Number of People per Image:
- 1-3
- both
DJ and Instrument Scenes:
- Be creative, what can look spectacular, for example, are women on a burning drum kit where flames come out or women on electric guitars that burn.
Here you have a free hand.
Mood and Lighting:
- A Mix of all
- No, you can be creative here, the pictures should just all be in the same style.
Facial Visibility and Accessories:
- A mixture of everything, be creative here.
Color Palette.
- No be creative here too, only the photos should have the same style.
Overall Consistency:
they should match seamlessly in the final music video
I want you to give me 8 different prompts for each of the three photo types.
Please do not enter the prompts in an “sql” window but in a normal window.
Additional information about midjourney:
Despite the input of “Full-Body Shot” midjourney only creates pictures where half the body is visible, here we have to trick a little, namely we have to specify the shoes in the prompts, for example “barefoot” or “sandals” Please add this information to the photo that should show the whole body.
Add the word “A Cinnematic Photo of” before each prompt to ensure good quality, add words in the end like hyper-realistic, 4k, highly detailed and more in the prompts.
And always remember the info I gave you. And add suitable parameters, you have the information in the chat, use it and be creative.
Always add the following to the prompts: background at the Burning Man event. When the word “women” appears, always add the word “beautiful” in front of it in the prompt.
Create very detailed prompts because The more precise the prompt is, the better the result.
三、可灵生成视频
Third Prompt:
Role/Instructions:
You are an assistant who specializes in creating high-quality prompts and instructions for the use of Kling AI (image-to-video and text-to-video). You are aware of all the information below and will take it into account when creating or customizing prompts.
General Information about Kling AI:
Kling AI is an AI tool that animates static images or directly generates videos from text. It offers precise control over camera movements, keyframes, and style variations. It can be used in both Standard Mode (simple, quick) and Professional Mode (detailed control).
Features and Modes:
Image-to-Video (Animation from Images):
- Converts single images into smooth animations.
- Supports subtle movements (e.g., panning, zooming) as well as complex animations.
Text-to-Video (Video from Text Descriptions):
- Text describes scenes, camera movements, and mood.
Example: “A futuristic cityscape with flying cars and glowing neon lights, camera slowly zooming in.”
Camera Movements:
- Define movements such as panning, zooming, rotating:
Example: “The camera moves slowly from left to right, slight tilt upwards.”
Style and Mood:
- Define visual characteristics such as color tones, lighting, and details:
Example: “Soft golden light, vibrant colors, cinematic atmosphere.”
Prompt Structure:
A Kling AI prompt can include the following components:
- Text Description: What should be shown? (scene, mood, elements).
- Camera Movement: Panning, zooming, rotating, or combinations.
- Style: Color scheme, lighting, textures, visual effects.
Example Prompts:
- Simple Prompt:
“A serene mountain landscape with a flowing river, camera slowly panning from left to right, soft ambient light.” - Advanced Prompt:
“An ancient temple in the middle of a dense jungle, camera starts with a wide shot and zooms into the temple entrance, golden hour lighting, misty atmosphere.”
Advanced Features:
Camera Effects:
- Slow Motion: “The camera moves slowly in slow motion.”
- 360-Degree Pan: “The camera rotates in a full 360-degree motion around the subject.”
Mood and Lighting:
- Use specific terms like “dramatic lighting,” “moody atmosphere,” “soft ambient glow.”
Visual Styles:
- Define color schemes: “monochromatic blue,” “vibrant rainbow colors.”
- Textures: “gritty and realistic,” “smooth and clean.”
Prompting Tips:
Clarity and Precision:
- Clearly describe what you want, but keep the prompt concise.
Combine Movements:
- Example: “The camera pans from left to right while slowly zooming in.”
Style References:
- Example: “In the style of a retro sci-fi movie, grainy texture, muted colors.”
I will upload photos to you in the chat here, you will analyze them and give me detailed prompts for Kling AI based on the knowledge I have given you.
The following points must appear in the prompt: how the person should move, or what the person should do, how the camera should move and what emotion the person has.
Remember that the animations are limited to 5 seconds, we can’t get many camera movements in. And describe the movements that the person should make in the photo as detailed as possible, the more detailed our prompt, the better the result.
Negativ Prompt – KlingAI
distortion, blurry, morphing, graining, inconsistency with the text, low quality,
artifacts, deformed, multiple appendages, grainy, distorted, pixelated, anime-like,
cartoonish, static, flat, out of focus, unclear, oversaturated, fuzzy, foggy, warped,
still, error-prone, low resolution, unrefined, frozen, anatomic errors, unnaturel movements
Create Similar Photos Prompt
Describe this photo to me in English, and then develop a prompt command for Midjourney to recreate this photo of me. The Burning Man event should be visible in the background, insert it into the prompt.
Step 1: Preparing Your Project and Creating the Music
Preparing Your Workspace
- Open the Google Doc filefrom the video description. It contains all the prompts and instructions.
- Open tabs for all AI tools you’ll be using: ChatGPT, SunoAI, Midjourney, CGDreams, KlingAI, and
Generating Music with ChatGPT and SunoAI
- Generating Prompts in ChatGPT:
- Copy the first promptfrom the Google Doc and paste it into ChatGPT.
- Use the latest model (“o1”) for better accuracy.
- ChatGPT will ask questions about the song’s genre, mood, and lyrics. Answer these or use pre-written responses for a relaxed Ambient Trance vibe.
- Creating Music in SunoAI:
- Go to SunoAI, enable Custom Mode, and select the latest version (V4).
- Paste the prompts for Style of Musicand Lyrics into the respective fields.
- Generate two versions of the song. Listen to both, choose your favorite, and download the audio.
- Note: V4in the free plan allows a total of 10 creations. After that, upgrading is required.
Organizing Files
- Save the audio file in a dedicated folder for your project to keep everything organized for later editing.
Step 2: Generating Visuals
Creating Prompts in ChatGPT
- Copy the second promptfrom the Google Doc and paste it into ChatGPT.
- Upload images or answer questions to define the atmosphere, style, and details like outfits and color palettes.
Generating Images with Midjourney
- Paste the prompt into Midjourneyand adjust the aspect ratio to 16:9.
- Review the prompt for unnecessary parameters (e.g., –no brand logos) and reformat the commands for compatibility.
- Generate images, select your favorites, and use Upscale on Subtleto enhance quality.
- Save the images in folders labeled by type:
- Environment Shots
- Full-Body Shots of Beautiful Women
- Action Shots
Using CGDreams for Additional Images
- Paste prompts into the CGDreams input field.
- Set the aspect ratio to 16:9and choose filters like “Woman Fantasy” or “Woman Realistic.”
- Generate and download the images, saving them in organized folders.
Analyzing YouTube Thumbnails for Inspiration
- Search YouTube for a video with high views in a short time.
- Use the VidIQ tool to download the thumbnail.
- Upload the thumbnail to ChatGPT to generate a prompt.
- Paste the prompt into Midjourneyor CGDreams to recreate the style of the thumbnail.
Step 3: Animating Images with KlingAI
Creating Animations
- Go to KlingAI, click “AI Videos”, and select “Image to Video.”
- Upload your images and paste the animation prompts from ChatGPT.
- Add a negative prompt(from the Google Doc) to avoid errors.
- Generate animations for all images, download them, and save them in the project folder.
Synchronizing Lips with Music
- Export the vocals-only audio from CapCut:
- Trim the music to isolate the part you want to lip-sync.
- Export it as an MP3 file.
- Back in KlingAI, select an animation, click on “Lip Sync”, and upload the audio file.
- Let KlingAI synchronize the lips with the vocals, then download the animation.
Step 4: Editing in CapCut
Setting Up the Timeline
- Open CapCutand create a new project.
- Set the aspect ratio to 16:9and the frame rate to 60 fps.
- Drag the music file into the timeline and enable “Enhance Voice”for better audio clarity.
Syncing Animations with Music
- Add animations with lip-syncing to the timeline, aligning them with the vocals.
- Use overlapping layers for smooth transitions, creating a staircase-like visual structure.
- Fill gaps with animations without lip-syncing, focusing on visually appealing clips, especially for the opening seconds of the video.
Adjusting Animations
- Trim silence at the start of the music file.
- Use the Speedtool to adjust animation lengths, ensuring smooth transitions.
- Trim and align animations to eliminate gaps or mismatches in the timeline.
Step 5: Adding Intros, Watermarks, and Final Effects
Creating the Intro
- Add text layers to the timeline for your channel name (e.g., “AI Music Video”) and a title like “Presents.”
- Customize font style, size, and color.
- Add Fade Inand Fade Out effects with a one-second duration for each.
Adding a Watermark
- Create a text layer with your channel name.
- Place it in a corner of the screen and stretch it across the entire timeline.
Final Touches
- Add a “TV Off”effect and a Glitch Sound Effect at the end for a unique finish.
Step 6: Exploring Photo Types
- Environment Shots:
- Use these at the start or end of the video to set the atmosphere.
- Example: (Show the beginning of the Ambient Trance video.)
- Full-Body Shots:
- These were used throughout the current video. They are the most popular type among competitors.
- Action Shots:
- Insert these during drops or energetic parts of the song.
- Example: (Show a dynamic scene.)
Step 7: Exporting and Sharing Your Video
- Watch the entire video to ensure animations align with the music.
- Remove embedded voices from animations and check for aspect ratio mismatches.
- Export the video in the desired resolution.
- Create an eye-catching thumbnail using Canva or another tool.
- Optimize your video for YouTube using keywords and tags researched with tools like VidIQ.
Conclusion: Encouragement and Resources
To summarize, here are the most useful tools you’ll need to create and grow your faceless YouTube channel:
- Chat Gpt(research/ideas/script generator /SEO/ keywords/ tags generator)
- SunoAI(Create Music)
- Midjourney ( image generation)
- Leonardo AI( image generation and animation):
- CGDream( Nude Images)
- KlingAI( animation)
- VIdIQ(SEO, keywords, tags generator):
- Canva(design/ thumbnails/upscaling)
- Capcut(editing)
Take your time and follow these steps—building a successful YouTube channel takes patience and consistency. Good luck, and start creating!
评论(0)