Gruber
patreon
WIP - Ohh Yeah! Text-Speech in Pico-8
🕑 Added 2018-01-25 22:43:23 +0000 UTCHuge thanks to @castpixel and @enargy for including me in this project. We're all working hard on it! If you'd like to hear it, it's on my Twitter feed: https://twitter.com/gruber_music/status/956308869516414976
Though, the attached cart is more polished, and sounds a bit better.
AN EXPLANATION
When you say a vowel, there are certain prominent frequencies that give that vowel it’s characteristic sound. These are called formants. Each vowel has three (F1, F2, F3).
@Enargy, @castpixel and I were using this chart:
For 'Ohh,' - I divided it into two parts - ow + oo
Formant 1 for ‘ow’ is 570Hz. Approx D5. In Pico8 with the instrument editor (instrument 0), I used D1
Formant 1 for ‘oo’ is 300Hz. Approx D4. In Pico8 with the instrument editor (instrument 0), I used D0
(I set up the instrument editor like a macro control to make changing sound types, volumes and pitches a bit easier)
So D1 and D0 are my targets.
Starting at music pattern 00, SFX 8 and SFX 12 are representing F1 for ‘ow’ & ‘oo’
I thought that if F1, F2 and F3 all changed at the same time, at the beginning of pattern 01, it would sound too square and unnatural (Like at pattern 06). So it’d be better to change the formants more gradually over time.
If you look at SFX 8 to SFX 12, I gradually move D1 to D0 (instrument 0). This can be seen more easily in in pitch mode (TAB). Instrument 0 is red/green. Actually, SFX 8 gets to D0 before it hits SFX 12. This is because I don’t want my formant targets to be reached all at the same time. This was my approach for all the formants, for ‘ow-oo’ + ’il-er-eh’ (Oh-Yeah)
While the formants are changing like this, the two lower frequencies (instruments 1&2) are moving up by semitone (for “ohhhh”)…. and then down my semitones (for “yeahhhh”). This is also more easily visualized in pitch mode. I wanted to emulate the rising/lowering pitch of our voice when someone says it like this:
https://www.youtube.com/watch?v=FoDb1QCZyXw
Last thing I did was add a volume envelope, from 1-4 through the duration of the phrase.
Still a work in progress. Next step is to add consonants, make it less resource heavy, and use code to control it. We’ll see what happens!
