Since the previous thread was closed, I had to start a new thread to raise the question of language support in music generation again.
Thank you for the detailed clarification and roadmap. The additional context is appreciated.
One concrete issue remains insufficiently addressed: language support and linguistic quality in music generation.
Over recent updates, several previously usable languages appear to have been deprioritized or degraded. In particular, Turkish currently exhibits severe pronunciation and prosody issues, to a degree that makes it unusable for serious musical or lyrical applications. This is not a subjective preference but a functional limitation.
Beyond pronunciation, the broader concern is reduced flexibility in music generation especially regarding vocal articulation, language-specific phrasing, and stylistic control. These aspects were central to Tunee’s value for advanced users working with multilingual or culturally specific material.
For users operating in non-English or non-Western contexts, this raises an important strategic question:
Are multilingual accuracy and language-specific vocal quality still active priorities, and can we expect previously supported languages including Turkish to be restored to a usable standard?
Clarifying whether these limitations are temporary side effects of the current optimization cycle or a deliberate scope reduction would significantly help users assess whether continued investment in the platform remains viable for their use cases.