I would like you to offer neutral spanish accent. Today virtual vocalists uses argentinian accent.

Dear Udio Team,

I recently tried generating a Kiswahili song using your platform and I wanted to share my experience with some feedback.

Firstly, I want to commend you on the platform’s excellent features, particularly the ability to use two voices (different genders) and the integration of instruments. These features made the process highly engaging, and I’m impressed with the overall flexibility in music creation.

However, I encountered an issue with the Kiswahili lyrics generation. The Swahili accent used by the AI sounds somewhat unnatural, resembling a foreigner speaking Kiswahili, rather than a native accent. This made it difficult to achieve the authentic tone I was aiming for in my song. I would appreciate it if the accent and natural flow of Kiswahili could be improved to sound more native.

I’m confident that with this adjustment, Udio can provide even more accurate and authentic experiences for Swahili-speaking users.

Thank you for the great work you're doing, and I look forward to future improvements!

Best regards,

Michael

you could improve support for other languages a little, especially Brazilian portuguese. I think its bit artificial and misinterpreted

Udio 1.5 and 1.0 cannot create a single clip of a Cantonese song with correct pronunciation.

For mandarin song creation in 1.5, it is very inconsistent, sometimes the vocals sing mandarin but then some words are cantonese.

Sometimes it provide clip with perfect pronunciation of mandarin but the arrangements (instrumental )are the worst. Sometimes it all became spoken words

if UDIO keep thinking these two languages are the same, it will become impossible to create a mandarin and Cantonese song in the future with UDIO

For example Estonian language in Udio has a noticeable accent, making it less usable (especially when compared to Suno).

Improving this, along with other smaller languages, would make Udio much more appealing to lot of users.

It’s currently holding me back from committing to a yearly subscription..

Korean pronunciation is weird from version 1.5,It's awkward for Koreans to hear the pronunciation. Especially when it's a quiet song

It used to be, if you wanted a song in french or japanese for example, it would generate lyrics in that specific language, but now when I tried to do that, it would automatically turn to english.

Is there going to be more language support visa vie arabic/hebrew and data for said styles? also, I know its' been said but is there a way that we can have our uploaded audio on the main website because as it stands, unpublished songs that get enough likes to go on daily/monthly do not get seen.

Hello,

I'd like to report Czech language if you can take a look at that please. The Czech language is very often distorted and in many cases sounds like a kind of Klingon, with the AI completely making up words, although I specify a clear lyric to be processed in verse to fit into a half-minute chunk.

I've noticed that when generating one 30-second chunk, the Czech is normal and understandable, while when generating the second/competition chunk, the AI just makes up incomprehensible Klingon, even though the text is defined in proportion to the length of the 30 seconds. No matter on whether it's intro, section or outro. This leads me to conclude this is a technical problem rather than that Czech is naturally unable to be sung. Udio seems to know Czech, it just likes to make it up. So it's primarily a technical deficiency, not that it can't speak Czech. Do you think that something could be improved in this respect? Perhaps by defining the desired language for generating tracks in the user preferences? For now I just specify the requirement of the sung language in the prompts, alongside other themes and tags or instruments.

Thank you so much and I keep my fingers crossed for your excellent work.

Have a nice day

In the current Udio, when we try to get them to sing Japanese, they do not sing Japanese well and the words are not clear in parts, or they sing words that sound Korean or Chinese.

I would like to see this corrected properly, as trying to fix it with inpainting, etc. will only make it worse!

I would appreciate it if you could fix it properly.

She used to sing very beautifully before entering 1.5, but since entering 1.5, she has been getting stranger and stranger.

Translated with DeepL.com (free version)