SpellSynth: Retro Vocal Synth

SpellSynth is a playable love-letter to the 1978 spelling toy that taught a generation to read — rebuilt for iPhone, iPad, and Mac, with four game modes, a crunchy VFD display, and three vintage voice characters across two decades of speech-synthesis history. Tap a letter and it speaks. Type ANY word — your name, this week's vocabulary list — and SpellSynth speaks it back. EVERY WORD SPEAKS The 500+ bundled words across 20+ themes use a hand-recorded retro voice. Anything else — a name, a foreign word, your kid's spelling list — falls back to the system engine through your chosen voice, cached so the second tap is instant. • FREE PLAY — type any word, a whole phrase, or a number and hear it spoken. Tap SPACE to build a sentence; it's read back word by word as the VFD cycles. Numbers ("2" becomes "two") and contractions ("don't") speak too. A GAP control tunes the pause between words from 50 to 500 milliseconds, and a RECENT row remembers everything you've played this session. • SPELL — SpellSynth says a word. You spell it. Three mistakes and you get the answer, then try a new one. Tap HINT to reveal a letter when you're stuck. • MYSTERY WORD — the classic "guess the hidden word" game. You get the category and the first letter free; pick letters until the word reveals itself. Six wrong guesses and it's game over. • SAY IT — listen-and-repeat mode for pronunciation practice. SpellSynth spells the word letter by letter, then says it whole. Tap REPEAT to hear it again, NEXT for a new word. • SHARE — render any word or phrase to a WAV and send it to Files, AirDrop, or your DAW. What you hear is exactly what you save. THREE VINTAGE VOICES Pick a voice character in Settings. Each one draws from a different era of speech-synthesis history, applied to every word you hear: • 1968 ODYSSEY — calm, deep-baritone monotone with a deliberately slow cadence. Fixed-pitch formant synthesis at ~65 Hz, time-stretched 1.6×. Aesthetic homage to the late-60s sci-fi AI voices. • 1978 LPC — the original Speak & Spell character. A 10-pole LPC vocoder driven by the actual TMS5220 chirp ROM, the same speech algorithm Texas Instruments shipped in the toy. Buzzy, granular, unmistakable. • 1986 DECTALK STYLE — fixed-pitch impulse train through formants extracted from your input. Modelled on the Klatt-style synthesis that defined Hawking's iconic computer voice. Deliberate, monotone, emphatically computery. • DEFAULT — clean system speech, no extra processing. For when you just want clear pronunciation without the retro stamp. All three work on every word and cache separately, so toggling between them is instant. Choice persists across launches. PLAY WITH A MIDI KEYBOARD Connect a USB, Bluetooth, or Network MIDI keyboard to your iPad and the keypad becomes playable from real keys. Notes C4–C#6 trigger letters A–Z, so any 25-key controller covers the whole alphabet. Pairs naturally with the LPC voice for a Speak & Spell instrument on your desk. BUNDLED AUDIO UNIT (AUv3) SpellSynth ships an AUv3 instrument extension, so the same voice engine loads inside GarageBand, Logic Pro for iPad, AUM, Loopy Pro, Cubasis, and any AUv3-compatible host. Type a word and it speaks live. Nineteen parameters cover pitch, speed, bit-crush, sample-rate reduction, filter, envelope, LFO, stutter, MIDI shaping, and voice character. Eight factory presets; all four voices play in your DAW. A serious MIDI surface for performers: sustain pedal (CC 64) for hold-and-release phrasing, Channel Pressure (aftertouch) routable to any parameter, three velocity curves (Soft / Linear / Hard), an adjustable pitch-bend range from ±0 to ±12 semitones, and 14 MIDI Learn targets covering every performance knob — bound CCs ride along with the preset. Stutter locks to the host tempo (1/4, 1/8, 1/16), and a Letter Map mode maps A–Z onto MIDI notes 60–85 so you can sequence spelling right in your DAW. Perfect for producers chasing that Kraftwerk / Stereolab / Broadcast sound.