June 2, 2026
HD voice, low-key comment war
PCMFlowG722 wideband (HD voice) codec for ESP32
Your tiny gadget can suddenly sound way clearer—and commenters rushed in to nitpick the fine print
TLDR: This new add-on helps tiny devices send much clearer voice without using more data, which could make DIY radios and gadget calls sound far less tinny. Commenters mostly liked it, but the loudest reaction was a nitpick-fest over one key point: it’s not just for ESP32, it’s just especially convenient there.
A tiny audio add-on for ESP32 gadgets just landed, and the pitch is simple: same data cost, clearer voice. In plain English, it lets small devices send “HD voice” instead of that old walkie-talkie phone sound, which is a big deal for hobby radios, do-it-yourself voice chat, and gadget-to-gadget calls. The project says it squeezes better-sounding speech into the same budget as older phone-style audio, which instantly made it catnip for makers who love getting more for free.
But the real spark came from the comments, where the classic tech-forum ritual began: the correction. One reader, MrBuddyCasino, jumped in to cool down any “ESP32 only” hype with a sharp little reminder that it’s not actually limited to ESP32—it just happens to fit neatly into ESP-NOW packet sizes. That tiny note became the whole mood of the thread: part excitement, part “well, actually.” It’s the kind of comment-section energy where people are impressed, but also spiritually unable to let a simplified headline live.
The strongest reaction wasn’t rage—it was that familiar maker-world mix of “this is awesome” and “be precise, please”. The joke underneath it all? Even when a project promises your mini gadget can sound dramatically better, someone will still show up to say the real headline should be about packet math. In other words: nicer voice, same nerd drama, no extra charge.
Key Points
- •PCMFlowG722 is an optional G.722 wideband codec add-on for PCMFlow aimed at real-time two-way voice over packet radio and network links.
- •The library supports 16 kHz PCM to G.722 encoding and G.722 to 16 kHz PCM decoding, with two PCM samples mapping to one G.722 byte.
- •Version 0.1 exposes only G.722 Mode 1 at 64 kbps; other G.722 modes, WAV container support, and Appendix III/IV packet loss concealment are out of scope.
- •The article compares PCMFlowG722 with PCMFlowG711 and PCMFlowOpus on bitrate, compression, flash footprint, CPU load, licensing, and audio quality.
- •For ESP-NOW, a 20 ms G.722 frame is 160 bytes, fitting within the 250-byte payload limit and providing 4× compression versus raw 16 kHz mono 16-bit PCM.