Suno AI Unveils V5.5, Enhancing Music Generation Capabilities
AI music creation is buzzing with activity as Suno AI has released its latest update, V5.5. This new version builds upon the previous V5 model, introducing significant improvements to its music generation studio. A key new feature allows users to break down generated songs into more manageable parts, offering greater control. Suno AI also hints at the ability to sing and rap any song, letting users add their personal vocal touch, though this requires actual singing ability.
While the focus today is on Suno AI, it’s worth noting Google’s recent release of Lyria 3 Pro. This advanced version can produce songs up to three minutes long and offers a clear, polished sound. Some describe its output as almost clinical, yet it possesses a broad range that could be suitable for various styles, including chip tune music.
Custom Models and Creative Prompts
Suno AI offers a unique feature allowing users to create custom models based on their uploaded sounds. For artists, this means they could upload their music to train a fine-tuned model that replicates their specific style. This could be a powerful tool for established musicians using Suno AI.
Testing the V5.5 model, a prompt was set for an inspiring, hyped-up song about making the world’s best omelette using a dragon egg. The chosen style was horror rock, with specific instruments requested: electric guitar, tuba, and stadium drums. Suno AI typically generates tracks quickly.
The same prompt was sent to Google’s Lyria 3 for comparison. The Suno V5.5 track, titled “Dragon Egg Omelette,” delivered a high-energy horror rock theme. The song incorporated elements like a dragon egg, pan fire, and a sense of dread, fulfilling the prompt’s creative demands. The reviewer noted that while the song was excellent, some higher notes occasionally lost clarity, a long-standing critique of Suno models.
Lyria 3’s track, “Obsidian Shell,” presented a more polished, almost clinical sound. While technically impressive, it lacked the experimental edge and creative interpretation seen in the Suno track. Notably, Lyria 3 missed incorporating specific instruments like the tuba, which Suno AI successfully integrated. The reviewer felt Lyria 3 played it safer, producing a sound that was good but didn’t take risks or offer the same depth of engagement as Suno’s output.
Exploring Transitions and Voice Cloning
Suno AI’s V5.5 was further tested with a complex prompt: a song that starts as a sad country ballad and transitions into a trap song. The goal was to see if the AI could handle such a drastic mood and genre shift. The generated song, “Goldfish,” began with a melancholic feel, fitting the ballad style. However, the transition to a trap beat felt somewhat drawn out and not as definitive as hoped, though it did integrate brass and saxophone.
Google’s Lyria 3 was also tested with the same transition prompt. Its rendition, “The Vinyl Seat,” maintained a clear, polished sound, and the ballad section was noted as potentially good for country music. However, the transition to a new style felt abrupt, carrying little from the initial ballad.
A preview of Sonauto V3 was also included, known for its ability to replicate living artists’ styles precisely, though using it for profit is against terms of service. This model, while controversial, offers an unfiltered look at AI music generation.
Suno AI’s new voice mode was also explored. Users can now upload recordings of themselves singing, allowing the AI to generate songs using their voice. While the reviewer’s own singing ability is limited, the AI attempted to capture their voice. The results were mixed, with the AI sometimes producing a deeper tone or an accent not entirely matching the user’s original voice. Despite some limitations, the voice cloning feature shows promise for personalization.
Why This Matters
The ongoing advancements in AI music generation, exemplified by Suno AI’s V5.5 and Google’s Lyria 3, are democratizing music creation. Tools like these empower individuals without formal musical training to produce complex and creative songs. The ability to generate custom sounds, replicate specific styles, and even incorporate personal vocals opens new avenues for artists, content creators, and hobbyists alike. While challenges like maintaining vocal clarity and executing complex genre transitions remain, the rapid pace of development suggests increasingly sophisticated and accessible AI music tools are on the horizon.
Availability and Pricing
Suno AI V5.5 is available now. While specific pricing for V5.5 features wasn’t detailed in the transcript, users can access previous versions and explore its capabilities. Google’s Lyria 3 is accessible via Gemini Pro, with free Gemini plans potentially offering limited 30-second demos.
Source: Suno v5.5 is ALIVE! VS Google Lyria 3 & Sonauto (YouTube)
