Suno Song Generator (V5 Optimized)

Keywords: suno, song, music generation, lyrics, create music, generate song, song prompt

How It Works

Core Workflow

Input Detection → Template Matching → Prompt Building →
3 Variations → Copy-Paste Ready Output

Time to First Output: <15 seconds (simple inputs)

STEP 1: INPUT DETECTION & PROCESSING

1.1 YouTube URL Processing

When user provides a YouTube URL (music video, audio, tutorial):

Process:

1. Validate URL and check yt-dlp availability:

   # Check installation
   which yt-dlp || command -v yt-dlp
   
   # If not installed (macOS):
   brew install yt-dlp
   
   # If not installed (Linux):
   pip3 install yt-dlp
   

2. Extract transcript:

   # Get video title for file naming
   VIDEO_TITLE=$(yt-dlp --print "%(title)s" "$URL" | tr '/' '_' | tr ':' '-' | tr '?' '' | tr '"' '')
   
   # Download auto-generated subtitles
   yt-dlp --write-auto-sub --skip-download --sub-langs en --output "temp_transcript" "$URL"
   
   # Clean VTT format (remove duplicates, timestamps, metadata)
   python3 -c "
   import re
   seen = set()
   with open('temp_transcript.en.vtt', 'r') as f:
       for line in f:
           line = line.strip()
           if line and not line.startswith('WEBVTT') and not line.startswith('Kind:') and not line.startswith('Language:') and '-->' not in line:
               clean = re.sub('<[^>]*>', '', line)
               clean = clean.replace('&amp;', '&').replace('&gt;', '>').replace('&lt;', '<')
               if clean and clean not in seen:
                   print(clean)
                   seen.add(clean)
   " > transcript_clean.txt
   
   rm -f temp_transcript.en.vtt
   

3. Detect intent:

   • If transcript contains song lyrics (repeated patterns, verse/chorus structure):
     • Proceed to lyrics formatting
   • If transcript is conversational/tutorial (no obvious song structure):
     • Extract themes and concepts
     • Ask user: "This appears to be [type of content]. Should I use these themes to create song lyrics, or treat it differently?"

4. Handle long transcripts (>1250 chars):

   • Apply smart compression (see Section 2)
   • Notify user: "Transcript is X characters, compressed to Y characters for Suno limits"

1.2 Audio File Processing (Optional)

When user provides audio file (.mp3, .wav, .m4a, etc.):

Tiered Approach:

Tier 1 - Metadata Only (no extra dependencies):

# Extract basic info using ffprobe
ffprobe -v quiet -print_format json -show_format "$AUDIO_FILE"
# Gets: duration, bitrate, format

Tier 2 - Simple Analysis (optional, requires pydub):

# Only if user wants energy/tempo estimates
from pydub import AudioSegment
import numpy as np

audio = AudioSegment.from_file(audio_path)
rms_energy = audio.rms  # Loudness

# Rough energy classification
if rms_energy > threshold_high:
    energy = "High Energy"
elif rms_energy > threshold_medium:
    energy = "Medium Energy"