MakeBestMusic: A Quick Start Guide

Getting Started

Once you open the "Create Music" interface, you will find that the feature supports generating both Songs and Instrumental music.

Getting Started

Let's begin with how to create a song. You have two options to choose from: Simple Mode or Custom Mode

• 1

  Simple Mode

  What it does: Requires only a brief musical description (e.g., "A smooth piece of jazz").

  Result: MakeBestMusic will automatically generate both the lyrics and the full song for you.

• 2

  Custom Mode

  What it does: Gives you full control over the creative process.

  Lyrics Input: You can input your own complete and structured lyrics. If you need inspiration, simply click the "AI Lyrics" button, and MakeBestMusic will generate them for you.

  

  Style Prompt (Crucial Step): This step is essential for the quality of the generated music. You need to describe the desired style as detailed as possible. Your prompt can include:

  • Musical Genre (e.g., Pop, Rock, EDM)
  • Key and Tempo
  • Instrumentation (e.g., acoustic guitar, analog synth pad)
  • Chord Progression
  • Vocal Accompaniment and Harmonies

  Random Style: If you don't have a specific preference, click the "Random Style" button. These styles are curated by our team and are designed to consistently produce pleasant, high-quality music.

  
• 3

  Advanced Options

  If you wish to make more detailed adjustments to your track, please open the "Advanced Options" section.

  Advanced Adjustment Parameters

  • Exclude Styles: Specify music genres or instruments that you want the AI to avoid during song creation. Note on limitation: due to the underlying AI model, signature instruments or vocal methods closely associated with the main style you prompt may not be fully excluded.
  • Vocal Gender: Use this to specify the gender of the generated vocals, allowing more precise control over the overall aesthetic of the music.
  • Weirdness: This controls the complexity of the chord progressions. A higher value results in more complex—and potentially more experimental—chord structures. Important:setting this value too low or too high may cause the generated music to deviate from your desired outcome.
  • Style Influence: Controls how closely the generated track adheres to the Style Prompt you entered. A higher value means the resulting music will be a closer match to your descriptive text. Important:setting this value too low or too high may cause the generated music to deviate from your desired outcome.

  

  Finally, enter your Song Title and click the "Create" button to generate your music.

Generating Instrumental Music

To create instrumental music, the process is similar to creating a song, but requires one key modification.

First, after selecting either Simple Mode or Custom Mode, you must toggle the Instrumental switch to the ON position.

• Simple Mode: Simply provide a brief description of the instrumental piece you want (e.g., "A smooth piece of jazz"). MakeBestMusic will automatically generate the instrumental track for you.
• Custom Mode: You can input detailed style prompts based on your vision. These prompts can include specific musical genres, key, tempo, desired instrumentation, and chord progressions.

Should you require more refined control, you can make subtle adjustments using the previously mentioned "Advanced Options."

Finding Inspiration

If you are looking for creative inspiration, you can browse excellent tracks created by others in the Showcase areas, such as the homepage or the "For You" section.

You can listen to the tracks there. If you find a piece you particularly enjoy, simply click the "Create Similar" button on the track to instantly generate a new piece of music with a similar style and characteristics.

After using the Extend feature to generate a longer track, you will receive two new extended songs simultaneously. These tracks will be clearly labeled with the "Extend Song" tag for easy identification.

Get Whole Song

Please note that the output generated by the "Extend" feature is not a full song; it is only the extension segment of the original track.

If you are satisfied with the generated extension, navigate to the "More Options" menu (usually indicated by the ... icon) on the extended track, locate the "Get Whole Song" action, and click it. This will instantly merge the original track and the extension into one complete song.

The final, merged track will be labeled with the "Merge Song" tag.

If your goal is to achieve an even longer song, you can reuse the "Extend" feature on this newly merged track to continue increasing its duration.

Creating a Voice Persona

Let's begin by explaining how to create and save your personal voice profile (Persona).

• Select Mode:Start by selecting Custom Mode.
• Access Feature:Click the Persona function, then choose Create Persona.

Extracting Your Voice

After creating the persona, you will need to upload an audio file to extract its vocal traits. This voice can then be used to generate the vocals for the other songs you create.

• Name and Describe:Once your audio is uploaded, create a Name for the persona and provide a detailed description of the voice.
• Tip:The more specific and detailed your descriptive terms are, the more accurately the AI will be able to extract a voice similar to your original audio.

Once the persona is created, you can simply click it to apply that specific voice to any new song you generate.

Saving a Voice from an Existing Track

If you are satisfied with a voice generated during the song creation process, you can save it as a Persona for one-click access in all future creative endeavors.

• How to:select From Library and expand the Select a Song option.
• Create Persona:Choose the song that contains the voice you like and select Create Persona.
• Access:Once saved, you can readily call up the saved persona voice whenever you produce a new track.

How To Use？-Singing Originals with Your Persona

To generate a song using your own created voice profile:

• Select a Persona (your custom voice profile) from your list of created characters.
• In the song configuration, input the complete lyric structure, detailed style description, and song title.

Copyright Notice:If the audio source you upload is a copyrighted commercial song, the system may prevent the upload. We recommend using your own audio sources or publicly available, licensed audio.

Five Core Capabilities

The Cover feature includes the following five core functions:

• Style Transformation:Instantly convert the original track's style into any genre you desire, easily exploring the music's potential across different musical styles.
• Instrument Replacement:Supports the free addition of various instrument sounds—such as guitar, drums, synthesizers, and more—to the arrangement, better aligning the production with your creative vision.
• Vocal Alteration:Flexibly adjust the song's vocals. This includes converting the voice to different types (e.g., child's voice, male voice, female voice) or completely removing the vocals to retain only the instrumental track.
• Lyric Adaptation:Allows you to modify the original lyrics or substitute them with entirely new ones. The new lyrics will be sung to perfectly match the original melody.
• Quick Song Generation:Simply upload a short humming or melody snippet and the system will intelligently expand it into a structurally complete song or instrumental track, significantly lowering the barrier to music creation.

How To Use？- Singing Originals with Your Persona

Accessing the Cover Feature

You can start using the Cover feature and select your reference track through one of the following two methods:

• From the Creation Menu:Navigate to the "Create Music" function menu, switch to "Custom" mode, and click the "Cover" button.
• From the Library: In the Library song list, find the track you wish to adapt, and select the "Cover" option from its "More Options" menu (...).

After accessing the feature, please select a song to serve as your reference audio source.

Tips：If you wish to use an audio file from your computer or mobile device as the source audio for your Cover creation, simply click the "Upload" button in the top right corner of the Library to submit your owned audio. For best results, clear articulation and melodic lines are advised; please be aware that copyrighted songs are strictly prohibited.

Extracting Your Voice

After creating the persona, you will need to upload an audio file to extract its vocal traits. This voice can then be used to generate the vocals for the other songs you create.

• Name and Describe:Once your audio is uploaded, create a Name for the persona and provide a detailed description of the voice.
• Tip:The more specific and detailed your descriptive terms are, the more accurately the AI will be able to extract a voice similar to your original audio.

Generating a Cover Track

After selecting your Cover song, the system will automatically populate the configuration fields with the reference song's lyrics, style, and title. You are free to adjust these settings according to your preferences, but please note that the generated song will retain the original reference song's vocal melody, regardless of the style changes.

Tip: If you need to modify the music genre, instrumentation, key, tempo, or vocal type, you can make these changes in the Styles section.

Advanced Parameter Influence

Pay close attention to two key parameters in the Advanced Options:

• Style Influence:A higher value means the generated music style will be a closer match to the style prompts you entered in the Styles field.
• Audio Influence:A higher value means the generated music will more closely resemble the characteristics of the original source audio (e.g., rhythm and structure).
• Important Warning:Setting either of these values above 90% may lead to distortion or undesirable issues in the generated music.

Step-by-Step: From Hum to Song

• Record Melody:First, record an audio segment of your hummed melody.
• Upload Audio:Navigate to the "Create Music" menu, click the "Upload" icon in the top right corner of the Library, and submit your hummed audio.
• Select Cover:Switch to "Custom" mode, click the "Cover" function button, and select the audio file you just uploaded as the source.
• Finalize Configuration:Input the desired lyrics and detailed style prompts.
• Submit Task:Click the "Create" button to submit the song generation task. In approximately 30 seconds, you will receive a complete song.
• Please Note:MakeBestMusic makes minor adjustments based on the hummed melody, so the generated song's melody is not guaranteed to be exactly the same as your original hummed audio

Generating Instrumental Music from Humming

If you wish to generate Instrumental music from your hummed melody:

• Adjust Mode:Switch the mode to Instrumental. Input prompts in the Styles configuration, such as: "violin solo" (if you desire an instrument solo).
• Remove Vocal Descriptions (Crucial):You must manually delete any vocal-related prompts from the Styles configuration. This is necessary because the AI automatically adds vocal descriptions when it detects humming, which interferes with instrumental generation.

Getting Started

Navigate to the "AI Voice Cover" menu to begin your creative journey!

Step 1

Select the audio to be covered

Select the song you wish to cover from your local device (such as your computer or mobile phone), or choose a track you have already created on our platform from the "Create Music" Library.

Song Requirements:

• Format:Mp3 / Wav / Flac files
• Length:Under 7 minutes
• Size:Under 30MB

Step 2

Select a Voice Model

Please select the Voice Model you wish to use for the cover song. You can either use your own cloned voice (Persona) or choose a suitable model from the Public Voice Library.

Step 3

Select Key Shifting

You can choose whether and how to adjust the song's singing key to optimize the performance of your selected Voice Model.

• No Key Conversion:The song's key will not be altered.
• High Pitch:The Voice Model will sing in the vocal timbre's high range.
• Mellow / Soothing:The Voice Model will sing in the vocal timbre's mid-low range.
• Natural:The Voice Model will sing in the vocal timbre's natural mid-range.

Final Step

Submit Task

Finally, click the "Cover" button to submit the cover generation task. In just a few minutes, you will receive your vocal-swapped song.

How to clone your own voice model?

Step 1

Access Voice Cloning Page

Navigate to the "AI Voice Cover" menu and click the "Clone My Voice" button to access the voice cloning configuration page.

Step 2

Configure Voice Model Details

On the configuration page, you need to complete the following actions:

• Upload Voice Image:Upload an image so you can clearly identify and distinguish this voice model.
• Set Name:Assign a unique name to your voice model.
• Select Gender:Choose the vocal gender for this voice.

Step 3

Select Cloning Fidelity

Please select your desired "Cloning Fidelity." The higher the cloning accuracy, the more similar the AI voice clone will be, and the better the AI audio cover effect will be.

AI Voice Cloning Features and Rules:

• Cloning Accuracy:The higher the cloning accuracy, the more similar the AI voice clone will be, and the better the AI audio cover effect will be.
• Basic Needs:The [Basic] and [Standard] modes require fewer source materials for cloning and can meet your fundamental needs.
• Advanced Modes:The [High-Fidelity] and [Premium] modes allow the AI to clone your voice with greater precision, significantly improving similarity. They also include audio refinement features like automatic auto-tune removal and noise reduction, resulting in a more accurate, clearer, and natural audio cover.

Step 4

Upload Your Voice Sample

You can provide your voice sample using one of the following two methods:

• Direct Recording:Record an a cappella (unaccompanied) audio segment of your singing.
• Local Upload:Select and upload an audio file from your computer, mobile phone, or other device.

Step 5

Select Training Audio

Please select the audio files you wish to use for voice model training.

Tip:Ensure the audio you select is clear, free of noise, and accurately represents the vocal characteristics you want to clone for the best possible results.

Submit Cloning Task

Finally, click the "Clone Now" button to submit the voice cloning task. After a processing period, you will receive your exclusive voice model for use in your music creation.

Final Step

Submit Cloning Task

Finally, click the "Clone Now" button to submit the voice cloning task. After a processing period, you will receive your exclusive voice model for use in your music creation.

Note: If the audio source you upload is a copyrighted commercial song, the system may prevent the upload. We recommend using your own audio sources or publicly available, licensed audio.

Getting Started

Navigate to the "Split Music" menu to split your music.

Step 1

Select Source Music

You can select the music file you need to split from your local device (such as a computer or mobile phone).

Alternatively, if you wish to split a song already created on our platform, navigate to the "Create Music" menu, find the desired song, and click the "Split" operation button within its "More Options" menu (...).

Step 2

Select Splitting Mode

Please choose the splitting mode that suits your needs:

• Normal:Get up to 5 separate instrument and vocal tracks (5 credits).
• Advanced:Get up to 12 separate instrument, vocal, and MIDI tracks (10 credits).

Final Step

Submit Task

Finally, click the "Split" button to submit the splitting task. After a few minutes, you will receive your separated audio tracks.

Meta Tags Overview

Meta Tags serve as core instructions in music AI creation. They guide the AI in constructing the song structure and determining the musical style, thus achieving fine-grained control over music and lyric generation. You can view them as explicit prompts given to the AI, ensuring it accurately understands and generates the specific type of song you envision.

How to Use Meta Tags Effectively

Skilled application of Meta Tags can significantly enhance the quality and precision of AI-generated songs. Here is how to apply Meta Tags effectively:

• Song Structure: Meta Tags help define the song's structure. For instance, you can use them to specify the various parts of the song, such as the Intro, Verse, Chorus, and Outro. This ensures the AI understands the song's flow and progression.
• Dominant Style: Meta Tags can guide the AI to create music in a specific style or genre, such as Pop, Rock, or Jazz. You can even use Meta Tags to mix and match different styles to create unique music.
• Integrating Dynamics: Dynamic elements like volume, tempo, and emotion can all be controlled through Meta Tags. You can instruct the AI to make certain parts of the song softer or more energetic, thus utilizing Meta Tags to add a professional touch to your music.

By proficiently using Meta Tags, you can gain full control over the music creation process, ensuring the AI accurately generates the masterpiece you have in mind. Therefore, whether you are a novice or an experienced musician, integrating Meta Tags into your workflow will elevate your MakeBestMusic experience to a new level.

Music Structure and Style Meta Tags

Structural Meta Tags

Basic Structural Tags define the conventional sections of a song:

• [Intro]: Introduction
• [Verse]: Main Storytelling Section
• [Chorus]: Hook/Main Idea
• [Bridge]: Transitional/Contrast Section
• [Interlude]: Intermission/Break
• [Pre-Chorus]: Pre-Chorus/Build-up to Chorus
• [Outro]: Conclusion/Ending

Special Structural Tags are often used in genres like Electronic Music:

• [Hook]: Catchy/Memorable Part
• [Break/Breakdown]: Pause/Instrumentation Stop
• [Sample Section]: Segment featuring sampling
• [Refrain]: Repeated Phrase/Line
• [Build-up]: Section increasing in intensity
• [Drop]: Release of energy/beat
• [Climax]: Highest Point of Intensity

Transitional Tags are used to indicate connections or endings between sections:

• [Lead-in]: Introduction to a section
• [Fade out]: Gradually decreases volume to the end
• [Instrument+solo]: Instrumental Solo (e.g., [Guitar solo])

Genre and Style Meta Tags

The list of style and genre tags available to guide AI music creation includes: Acoustic, African, Alternative metal, Alternative pop, Ambient, Atlanta rap, Ballad, Baroque, Blues, Boom bap, Cello, Chill, Christian & Gospel, Christmas, Country & Americana, Dance & Electronic, EDM, Girl group, Gospel, Hardcore rap, Heavy metal, Hip hop, Indie, Indie rock, J-pop, Jazz, K-pop, Lo-fi, Orchestra, Party, Piano, Pop, Pop-Rock, Post-Hardcore, Punk Rock, R&B, Soul, Rap, Reggae, Rock, Romantic, Soul, Synth, Synth pop, Techno.

Vocal Meta Tags

These tags are used to specify the voice type, range, and combination of vocals within a song: male bass, female mezzo, female child voice, child voice, teen voice, cartoon child voice, children's choir, male & female duet, male choir, mixed choir, child harmony, female lead vocal with child backing choir, female mezzo solo with layered robotic harmony, teen lead vocal with children's choir outro, male & female duet pop styles.

Lyric Construction Template

Add a section tag before each segment of lyrics (e.g., [Verse], [Chorus]).

• You can specify the exact number of bars within instrumental section tags like intros and interludes (e.g., [Intro 16bars]).
• You can also specify the instrument to highlight or the vocal category to use within a section tag for more precise control over the generated music (e.g., [Verse 1: piano solo, female vocal]).
• To ensure specific instructions are followed, use square brackets ([]) to define song sections.

Example: Lyric Structure

Style Prompt Construction Template

When writing style prompts, try to be as detailed as possible. The more detailed your prompt is, the higher the quality of the generated song will be, and the better it will align with your expectations.

Your style prompt should ideally include the following points:

• Precise Music Genre: If you haven't decided on a specific genre, look for inspiration in the genre and style list below.
  Note: Ensure the music style is consistent with the lyrical content's tone (e.g., avoid pairing the [Heavy Metal] style with soft lyrics like "The flowers are so fragrant and beautiful").
• Tempo: e.g., 70 BPM. A higher BPM value results in faster music.
• Time Signature: 4/4 time is the default setting and is suitable for most popular song structures.
• Specified Vocal Category or Singing Style: e.g., Male/Female vocal, raspy texture, soulful R&B style, etc.
• Detailed Instrumentation: e.g., Epic orchestral score, Soaring string section, Thundering timpani, Heroic brass fanfare.

  Note: Ensure the choice of instruments is consistent with the music style's tone (e.g., Electronic music focuses on digital sounds like synthesizers and drum machines; Rock focuses on electric guitars, bass, and drums). A rational combination of instruments effectively strengthens the style expression and prevents acoustic clashes.

Combining Attributes:

When you want to combine multiple attributes (such as style, key, and instrument emphasis), use commas (,) to separate them and clarify your intent.

Your style prompt should ideally include the following points:

• Example: pop ballad, C major, 4/4 time, 70 BPM, female vocal, soft drumbeats, lyrical piano, warm synthesizers...

This method helps the AI distinguish between various stylistic elements.

Universal Formula Template:

1 Precise Music Genre + 1 or More Core Instruments + 1 Clear Vocal Type (with optional singing method)

Tips and Restrictions:

• Experiment with different lyric structures and style prompts to see what works best for your writing style.
• Incorporate elements of emotion, speed, and "catchiness" to enhance the music's overall appeal.
• Remember, the more specific your instructions are, the more likely the AI is to produce the results you desire.
• Restriction: Do not enter two or more music style prompts simultaneously in the Styles field, as this may cause errors in the generated song.
• Inherent Instability: Due to the underlying training data of the AI model, the song generation process has inherent instabilities. Therefore, the system cannot guarantee 100% recognition and execution of all instructions you provide.