Wow, thank you. I just learned more about writing songs than in all the videos I've ever seen. Thank you for sharing your knowledge. I just experienced something strange in Suno. I had a German text. It didn't really work. Then I did it in English. It was good. Then in French, Spanish and Italian. And with those three languages I almost always got a duet. Apparently the language also has a big influence on the result. I'll try to see if I can replicate that with Persona or Cover in English and German.
Hi Anadis, thank you for your kind remarks! Knowing the material I'm putting out helps other people means a lot and helps keep me going. I believe you are right about how the language used affects the quality of the output, as I've had Chinese speakers say they also have challenges with getting various things to work in Suno. I think it has something to do with the dataset/training data they use on the backend. I'd guess that the number of publicly available recordings of music for a certain language Suno has access to will impact the accuracy of the output. (But I'm not as well-read on the technical side of how AI works, it's just my musing.)
Thanks for the video, but it still did not work for a medieval style duet song. And when it FINALLY gave me a duet, it made a male voice where I gave the instructions to have a female vocalist and vice-versa... its frustrating. Maybe in a few months...
Hi John, thank you for the feedback. I will look into my options to improve the audio quality. I assume you're referring to the mic quality and volume differences between spoken vs. music sections, is that correct?
Wow, thank you. I just learned more about writing songs than in all the videos I've ever seen. Thank you for sharing your knowledge. I just experienced something strange in Suno. I had a German text. It didn't really work. Then I did it in English. It was good. Then in French, Spanish and Italian. And with those three languages I almost always got a duet. Apparently the language also has a big influence on the result. I'll try to see if I can replicate that with Persona or Cover in English and German.
Hi Anadis, thank you for your kind remarks! Knowing the material I'm putting out helps other people means a lot and helps keep me going.
I believe you are right about how the language used affects the quality of the output, as I've had Chinese speakers say they also have challenges with getting various things to work in Suno. I think it has something to do with the dataset/training data they use on the backend. I'd guess that the number of publicly available recordings of music for a certain language Suno has access to will impact the accuracy of the output. (But I'm not as well-read on the technical side of how AI works, it's just my musing.)
Big shout out back to you. Thank you for sharing your knowledge. I've learned so much from you.
No prob; I'm happy to!
Thanks for the video, but it still did not work for a medieval style duet song. And when it FINALLY gave me a duet, it made a male voice where I gave the instructions to have a female vocalist and vice-versa... its frustrating. Maybe in a few months...
Hi SAMMYTASTISCH, what style prompt did you use? I'd like to see it (or a couple of them).
Thank you very much for your vids. But please, can you improve your audio?
Why?
Hi John, thank you for the feedback. I will look into my options to improve the audio quality. I assume you're referring to the mic quality and volume differences between spoken vs. music sections, is that correct?