“AI”
in a time where “low budgets,” in music productions continue to dominate, “AI”, helps a lot in production, before involving real singers.
Personally, I have managed to make it a “dogma,” managing to find the right balance between various software and vocal synthesizers.
Recently, I took a new test by venturing into an unusual cover and in a language I do not know, “La Flaca.”
In fact, Spanish pronunciation, like any language has its phonemes and it is not easy to generate an A.I., without clashing with the inflections that the available software imposes, but I would say that in spite of everything, I am coming up with acceptable results.
These “efforts” of mine only want to show Dreamtonics, that maybe it would be time to open there mind and make use of “people like us”, to credibly enlarge the databases with the addition of Latin languages (Italian, Spanish, etc.), which in world music have substantial relevance.
I am a musician and because of this detail, I try to carefully curate, sound, groove, time and interpretation, because AI is cool, if “human” with musical skills, adds what a machine can never have.
Have a good listen and let me know what you think of this.
This is a good production, really good, I mean really really good! Your mix is fantastic and the vocals sit well into it.
To make it sound more human I’d try playing with the “expressiveness” sliders - quite a lot of the notes have that classic “autotune” robotic sound. You can also draw in notes with “br” as the lyric to create breath sounds at the start of phrases for added realness. But still, what you have done is great work.
I am from the Czech Republic and I can only speak Czech and very little English. I don’t know Spanish and Italian, but without understanding the text, I can listen to both languages. It seems to me that the singing is closer to Italian than to Spanish, but I don’t mind that at all. It is musically very nice and the singing seems to me to be from a live singer (excellent singing), although I have to watch that it is sung by A.I. singer and I’m confused because (probably on purpose) you didn’t mention who is used like A.I… Or I didn’t understand the Google translation of your post into Czech. But I am able and willing to play a song three times over and over, or even more times, as I have long been used to listening to a song that caught my attention in this way.
It’s not Kevin actually,
It’s a secret …
It’s actually one of the “eclipsedsounds” I used for the SynthV phase, while The voice I used, which you hear , derived from several files of a “real” voice I have available given to me by a recording studio.
I requested them precisely to try to derive “a voice” with Italian phonemes, precisely so that I would not have “those voices” imposed by the software, and with a lot of patience, I am still working on it.
Thank you @BenjiRage , I know most of the tricks you have listed in your feedback. I appreciate it anyway .
As for the “Autotune” thing, you know … , Autotune is now a component, an integral tool in music production…
Yes, it was me, and I’m glad I asked you for an example, as I see it is possible to have a good result It makes me want to experiment with other languages too
Well done! So, if I have understood correctly, actually this track is not fully made by SynthV but is a hybrid of SynthV and samples. Am I saying that right? Out of curiosity, please which eclipsedsound voice did you use for SynthV phase?
You are are right. All I can say is, we now have two Voicebanks by eclipsedsounds… so you can choose whatever of both for own songs, it doesn’t depend from the original timbre of them, but just from the outcome of their pronunciations and words articulation speed… Solaria is versatile while Asterian has his reason in other situations.
…
.
)… It was inspiring for me!
, I’m looking forward to listen your work