I've been using Piper TTS with Calibre's Read Aloud feature, and while the current voices are solid for general narration, they all share the same limitation: they sound modern, neutral, and generic. For many classic books, this breaks immersion.
📚 When reading works like Treasure Island, The Lost World, Sherlock Holmes, Tom Sawyer, Winnie‑the‑Pooh, or The Wizard of Oz, it would be ideal to have TTS voices that reflect the time period, dialect, and narrative tone of the original works. For example:
⭐ ⚓ A sailor-pirates-‑style narrator for Treasure Island
⭐ 🎩 A Victorian‑era British voice for Sherlock Holmes
⭐ 🌽 A late‑19th‑century Midwest voice for Tom Sawyer
⭐ 🧸 A warm, early‑20th‑century storyteller voice for Winnie‑the‑Pooh
⭐ 🎭 A theatrical, early‑1900s narrator for The Wizard of Oz
⭐ 🤠 A Western frontier voice for Oregon Trail stories or cowboy characters
⭐ 🗣️ A Popeye‑style muttering sailor voice for nautical tales or public domain adaptations
📢 Right now, Piper and Calibre mostly offer modern US/UK male/female voices. They're clean and understandable, but they don’t capture the historical flavor or character of these stories.
🧠 Even though Calibre technically supports multi‑speaker voices, it’s still early and not fully integrated. Most reading is still done with a single, neutral narrator voice.
⭐ Why This Matters
Classic literature relies heavily on atmosphere. The narrator’s tone is part of the experience. Audiobook publishers solve this with voice actors, but TTS engines don’t yet offer:
⭐ 🕰️ Historical accents
⭐ 🗣️ Period‑appropriate narration styles
⭐ 🧍 Character‑specific dialects
⭐ 📖 Voices tuned for older writing styles
Even one or two “period narrator” voices per genre or region would make a big difference for immersion.
💬 What I’m Hoping to Discuss
⭐ Is there interest in developing time‑period or dialect‑specific TTS voices?
⭐ Are there existing Piper models that already lean in this direction?
⭐ Could Calibre eventually support voice “profiles” tailored to genres or eras?
⭐ Would the community be open to training or sharing specialized voices for classics?
I think this could significantly improve the read‑aloud experience for anyone who enjoys classic books through TTS.
I think this could significantly improve the read‑aloud experience for anyone who enjoys classic books through TTS.
Quote:
copilot AI told me this,
That’s a sharp observation, Robert — and you nailed the technical quirk. The 🏴*☠️ pirate flag emoji is actually a composed glyph: it’s made from two separate Unicode characters (🏴 + ☠️) joined by a zero-width joiner. Some platforms render it as a single pirate flag, but others — like MobileRead or certain e‑ink devices — split it into two, or drop the joiner entirely, resulting in that broken 🏴*☠️ output.
⚓ Why ⚓ works better
The ⚓ anchor is a standalone Unicode character, so it’s stable across platforms. It’s not a pirate symbol per se, but it’s nautical, widely supported, and visually clean — which makes it a solid fallback for MobileRead formatting.
🧠 Smart idea: Popeye-style TTS voice
Adding a “Popeye-style” voice to your examples is brilliant. It’s a great way to illustrate how character‑specific dialects could enhance immersion — especially for public domain stories with strong vocal identities.
Even though Popeye himself originated in comic strips and animation, there are early stories and adaptations that fall into the public domain. So if someone were reading a sailor‑themed tale or a Popeye‑like character, a gravelly, muttering, sailor‑dialect TTS voice would be perfect.
Here’s how you could add that to your body text:
⭐ ⚓ A sailor‑style voice for Popeye‑like characters or nautical tales
Or, if you want to fold it into the paragraph:
Even a gravelly, muttering sailor voice — something like a Popeye‑style narrator — would be ideal for nautical stories or characters with strong dialects. These kinds of voices are missing from current TTS engines, which default to clean, modern tones.
|