• We're currently having issues with our e-mail system. Anything requiring e-mail validation (2FA, forgotten passwords, etc.) requires to be changed manually at the moment. Please reach out via the Contact Us form if you require any assistance.
morrysillusion

morrysillusion

v flower enthusiast 25 From Socal
Messages
777
Reaction score
4,386
Medals
5
Joined
Last seen

Profile posts Latest activity Postings Awarded medals About

  • morrysillusion
    i feel like i gotta reiterate sometimes... i still love vocaloid 4 and older voices. like, AI voices are incredible with how great they sound. but old ones have their charm. theyre still their whole own thing. i love v4 flower, i love fukase, i love lola, i love kaai yuki, i love sweet ann, i love dex... i dont wanna forget how much i love every generation!
    morrysillusion
    theres some things synth v does with words that i find kind of annoying and prefer in vocaloid, but at the same time... some of what vocaloid does i like but want *slightly* tweaked. for example i do like sometimes how synth v does [enter multi syllable word] add [+] to continue it, because it does make it easier to change the source word and add [-] where you need. but also i liked the ease of vocaloid's choice of just splitting it up immediately lol. i can admit, im lazy and dont always like having to put the + myself! but there are obvious drawbacks from vocaloid's method too, probably more that make me say synth v feels better anyways

    also some of synth v's automatic phoneme choices also make no sense to me, esp annoying is "and" coming out as [ax n dx] which pretty much always sounds like "uhnd". i always have to make it [ae n dx]. that one always pisses me off.
    morrysillusion
    i know i really fell out of recording a new utau to release... it hasnt been my immediate thought. but i want to try and record a new vcv especially because i want to compare it to my voice pre-T! its been over 2 years, almost 3, since i started. so i think my voice is stable enough to give a good comparison.
    morrysillusion
    whoa i hadnt realized anri arcane came out? i have her old one so i was gonna redeem the new one. im just surprised i didnt get any kind of email telling me, as i had bought the old one...
    morrysillusion
    i plan to get gumi sv at some point simply bc i have never owned her in english in some form and id like to have that ability but, plan to wait a bit to see how she may update (if at all). in the mean time i have just gotten rose- i had little intent to hold back on her lol. she is a voice type i have been dying for and i look forward to using her
    morrysillusion
    anyone have any suggestions on artists i could look into to cover with saros? i am just not up to date on pop oriented stuff or really much of anything outside of my insanely niche bands lol
    • Like
    Reactions: MillyAqualine and Ryleigh Goldner
    Ryleigh Goldner
    Ryleigh Goldner
    Personally, if I could afford to have SAROS, I would use them to cover Okame-P's and Shoten Taro's songs. And maybe Nine Inch Nails, Poppy, and some Utsu-P?
    That's all I could think of for my use case
    I'm so sorry that I don't have any more (EDIT: that I could think of)
    MillyAqualine
    MillyAqualine
    Well, lately i've been spamming this to myself though it'sd a bit recent-old and minimalist

    Then maybe try some funk or jazz songs may they be Earth wind and fire or even this song by Samara joy and fellow ones she sang

    morrysillusion
    the one interesting thing about using ARA and telling it to find the phonetics is it really is 100% phonetic based and not *lyric* based and it actually is a cool learning tool to realizing how singers sing in ways that arent just strictly typing the words in and using the default. for example in synth v i enter the words [put]. so those notes will enter phonemes as [ay] [p uh t] because those are what i typed on those notes. however with ARA it will put it like this [ay p] [uh t] because it hears it that way which... makes a lot of sense and sounds really natural. a lot of instances like that occur, where a consonant at the start of one word will actually be at the end of the previous note to flow with the vowel before it.
    • Like
    Reactions: TheStarPalace, Twillby, wrong_thyme and 1 other person
    AmazingStrange39
    AmazingStrange39
    ace did the same thing iirc (i no longer have access to it since it went paid but i remember), nice to know synthv is doing that too

    i guess for vocaloid the closest equivalent would be vocalistener or mikumikuvoice? (last i checked you can't make vsqs in vocalshifter) there's also vocalo changer but that's a voice changer, not audio to midi, of course
    morrysillusion
    morrysillusion
    @AmazingStrange39 its been ages since i used vocalistener but iirc it just shows you the pitch bends it generates from the vocal source, but you still need to draw the timing of the notes yourself, and then also enter lyrics. so there is no generating of the lyrics/phonetics being done for you like synth v is now capable of
    morrysillusion
    been out of town etc and just busy so i havent worked on any solid covers w saros. but i am messing with ARA and really, it is quite good. i have tried to see how much the phoneme detection and pitch conversion is, it is far more sensitive on the ripped vocals i have. but the timing and generation of the midi notes is really good. especially the timing! im still impressed with it being beta.
    morrysillusion
    if you open up some of saros' demo files they came with you can see an unused/not final released vocal mode in the parameter tab having been used (and considered 'inactive' because its not one the voice has) (wouldnt assume anything crazy of this. its likely just a vocal mode that didnt have a name yet and i thought it was interesting)
    morrysillusion
    i didnt realize the ARA thing could also 1) detect phonemes/lyrics 2) convert pitch from the vocal source. thought it was just generating a base midi! this basically makes the realvoice plugin w praat no longer needed... which, ARA is much much more effecient. itll be a matter of seeing it improve, but it takes a LOT of pain and tedium out of the realvoice method. annotating a 3 minute song by hand? god. no more. praat's pitch detection on AI ripped vocals is probably as good as ARA right now anyways, since its unclean. im shocked to say the phoneme detection is pretty decent, this has worked very well with the AI ripped vocals i tested. it even detected vocal dry??? ex. i used an adelle song and boy, she does a lot of vocal fry. the detection is pretty literal, not trying to make words but rather sound, and with a source like adelle with her specific singing it actually used phonetics that matched her, not the words. i know its way better with a clean vocal but it still does a decent base for ripped vocals, just probably better to enter lyrics yourself if its ripped
    morrysillusion
    ive had this design in mind for Hayden i just havent had any drawing energy..... most i can say is i see him as a red head with curly hair. dont ask why it just came to me suddenly and im sticking with it.... i feel like id draw him in light colors complimenting that. dull warm colors (greens, yellows, browns etc). not entirely sure on fashion sense tho
    morrysillusion
    idk if its just me but a lot of synth v voices' vocal fry [cl] either doesnt work at all or just isnt apparent (even when changing note duration/strength) and its always annoyed me. however i am so pleased to hear hayden's vocal fry DOES work/sound obvious and the note edits for that phoneme make it even more blatant. so happy
    morrysillusion
    i got hayden and i love him but i have zero covers rn that i want to plug him into.... no idea what i wanna make with him yet
    • Like
    Reactions: WyndReed and Ryleigh Goldner
    Ryleigh Goldner
    Ryleigh Goldner
    In my case, if I ever get to purchase him, I'd use him to cover VOCALEAMO, VOCALOUD, and VOCAMETAL songs (like Okame-P's, Shoten Taro's, Yuyoyuppe's, and Utsu-P's discography). And maybe some of Okame-P's softer/soft songs (like ballads).
    His voice really reminds me of Oscar Wilde (not the writer, but Pascalle Lepas' character). (EDIT: Oscar has no voice acting since he's a character in a web comic)

    But I think it's up to you to decide what you want to use him for.


    Addendum:
    On a side note...I might cover PEGGY's (JPEGMAFIA) songs using him. SynthV's rap mode has never been to my taste, though.
    AmazingStrange39
    AmazingStrange39
    He kinda reminds me of John Mayer tbh

    But if I had him I'd probably use him in soft pop and soft rock type stuff, he'd probably sound good with EDM as well.
    morrysillusion
    morrysillusion
    agreed! i just plugged him into a KIRA and nostraightanswer song bc they felt potentially fitting, and he does sound really nice. ill have to explore what i may actually finish with him but that genre seems simple enough for him. but i do wanna try him with something heavier like rock...
  • Loading…
  • Loading…
  • Loading…
  • Loading…
Top Bottom