Question about the "pitch" scale

[AceLuke]

Hello! I'm wanting to explore the options of making SAM sing for a project. I'm wondering what the pitch system is based on. Is there any relative connection to the 12 TET system or is there any kind of chart that showcases the frequencies? Would love to learn more about it.

[andersjalevik]

I was searching for the same but didnt find anything so I hocked up my tuner and got these values (most a bit out of tune since pitch seems to be integers, would be nice with float support for better pitch precision)

const C0 = 204;
const Cs0 = 196; 
const D0 = 185;
const Ds0 = 175;
const E0 = 165;
const F0 = 156;
const Fs0 = 147;
const G0 = 139;
const Gs0 = 131;
const A0 = 124;
const As0 = 117;
const B0 = 110;
const C1 = 104;
const Cs1 = 98; 
const D1 = 93;
const Ds1 = 88; 
const E1 = 83;
const F1 = 79;
const Fs1 = 73; 
const G1 = 69;
const Gs1 = 66; 
const A1 = 62;
const As1 = 58;
const B1 = 55;
const C2 = 52;
const Cs2 = 49;

// shorthand for the scale I used
const D = D1;
const E = E1;
const Fs = Fs1;
const Gs = Gs1;
const A = A1;
const B = B1;
const Cs = Cs2;

Then I defined speed, pitch, delay and text (use phonetic for better control)

const line = [
    { speed: 1.8, pitch: Cs, phon: "D R AY" }, // Dri
    { speed: 1.8, pitch: B, phon: "V IH NG", delay: 0.05 }, // ving
    { speed: 1.5, pitch: A, phon: "IH N" },  // In
    { speed: 1.5, pitch: Gs, phon: "M AY" }, // My
    { speed: 2, pitch: Fs, phon: "K AA R" }, // Car
]

and did and did a crude looper

for (const { speed, pitch, text, phon, label, delay } of line) {
       const sam = new SamJs({
        phonetic: !!phon,
        singmode: true,
        pitch: pitch,
        speed: 50*speed,
        mouth: 120,
      });
      await sam.speak(phon || text);
      if (delay) {
        await new Promise((resolve) => setTimeout(resolve, delay * 1000));
      }
}