Here's my previous vocal synth project. https://scratch.mit.edu/projects/723570001/ This is a Japanese synthesizer because it does not understand the roman alphabet. It only understands 11 hiragana that I recorded. I may add some more in the future. It only includes these sounds. If you do not have a Japanese keyboard, you can copy-and-paste. あ/ア(a) え/エ (e) い/イ (i) お/オ (o) う/ウ (u) さ/サ (sa) せ/セ (se) し/シ (shi) せぃ/セィ (si) そ/ソ (so) す/ス (su) You can click the note and enter the hiragana, katakana, or romaji equivalent for any of these letters. Initially, the voice being played back will be set to Aitori (who is voiced by @HeliosTheSuperior), recorded specifically for this. If you press 2 on your keyboard, the voice will change to Akiko (who is voiced by @Moonwatcher127), whose recordings were taken from their UTAU voicebank. You can press 1 to turn the voice back to Aitori. Each pitch is not recorded individually, it is only recorded in one pitch and uses the pitch effect block. You can make the pitch higher or lower while pressing the up and down arrows while hovering your cursor over the note. It's recorded at F4 so anything too high or too low from that will sound off. If you want something with kanji characters, I recommend Nihongodera, as it can convert kanji to hiragana/katakana. How does this work? This works by saving each hiragana as a costume and sound and, when requested, it will play whichever hiragana its costume is. Things I want to add: Changing length Multiple notes Remix this by re-recording each sound (keep the same name and length though!!) with your own voice!
