Voice Restoration Through Quartz and Micro-Power Generation

Key Message: Embedding voice data directly into QR codes and engraving them on quartz glass to "preserve a voice for 1,000 years" — we are advancing technology development toward this dream and have verified that approximately 30 seconds of audio can be stored and played back from a QR code using Codec2 ultra-low bitrate encoding technology.

1. The Desire to Preserve Voice

Among what humans want to preserve, "voice" holds a special place.

Photos preserve appearance. Text preserves thought. But voice preserves one's "presence" itself. Tone, intonation, pauses for breath — voice contains information that text cannot convey.

History of Voice Recording Technology

In 1877, Thomas Edison invented the phonograph. This was humanity's first technology to "record voice." Less than 150 years later — we can record voices.

But "recording" and "preserving" are different.

Wax cylinder records — Degrade in decades
Magnetic tape — Magnetism weakens, degrades in decades
CDs and optical discs — Approximately decades to 100 years
Digital files — Risk of medium and format obsolescence

None of these technologies guarantee 1,000-year preservation.

Can We Preserve Voice for 1,000 Years?

TokiStorage's quartz glass has physical durability of over 1,000 years. However, current services record "text" and "URLs" in QR codes. Voice itself isn't recorded.

Why? Data capacity issues.

Standard QR codes can store maximum about 3KB of data. Meanwhile, audio files require several KB to tens of KB per second even when compressed. Direct voice recording in current QR codes is difficult.

— Or so it was thought. But we broke through this barrier.

2. Achieving Audio QR Codes

We tackled the challenge of "embedding voice directly into QR codes" and developed a working system.

The Limits of General-Purpose Codecs and Adoption of Codec2

General audio compression like MP3 and AAC considers 128kbps (16KB per second) as "acceptable quality." This won't fit in a QR code.

We first evaluated the Opus codec, but discovered through experimentation that a sharp cliff exists between 2-4kbps — a DTX transition where audio abruptly becomes silent. Within QR code capacity, Opus could only store approximately 3 seconds, proving impractical.

Based on this discovery, we adopted Codec2, an ultra-low bitrate speech codec originally developed for amateur radio. Codec2 maintains intelligible voice quality across a continuous range from 450 to 3200bps, even at 450bps where human speech remains comprehensible.

30 Seconds of Audio Storage Demonstrated

As a result, we have demonstrated that approximately 30 seconds of audio can be stored and played back within existing QR code specifications.

30 seconds — audio length achieved with proprietary voice encoding (demonstrated)

With 30 seconds, it's not just a single phrase. You can introduce yourself, share your feelings, and leave a message for the future. You can deliver a letter to your descendants — in your own voice.

Future Expansion Possibilities

Combining AI voice synthesis technology opens further possibilities. If "voice characteristic data" can be compressed to under a few KB, it could be recorded in a QR code, and AI 1,000 years from now could recreate any message in that person's voice.

3. Quartz's Piezoelectric Effect

Quartz has a property called "piezoelectric effect." This opens possibilities for playing audio without power.

What Is the Piezoelectric Effect?

The piezoelectric effect is a phenomenon where applying mechanical force (pressure) to certain crystals generates electricity, and conversely, applying voltage deforms the crystal.

Quartz (crystal) is the most representative piezoelectric material.

Direct effect — Pressure → electricity is generated
Inverse effect — Voltage → crystal vibrates

Crystal Oscillators

This property is already widely used. The crystal oscillator, the heart of quartz watches, vibrates at 32,768Hz when voltage is applied, keeping accurate time.

32,768 Hz — standard crystal oscillator frequency (2¹⁵)

Converting this vibration to audible range can produce sound. Indeed, piezoelectric buzzers and speakers work on this principle.

The Energy Source Problem

Producing sound via piezoelectric effect requires voltage. How to supply that voltage 1,000 years from now — this is the biggest challenge.

Batteries won't last 1,000 years. Depending on external power loses TokiStorage's "self-contained" nature.

But here lies the possibility of "micro-power generation."

4. Micro-Power Generation (Energy Harvesting)

Energy harvesting is technology that converts minute environmental energy into electricity.

Environmental Energy Sources

Our surroundings are filled with unused minute energy.

Vibration — Machine vibration, walking vibration, sound waves
Heat — Thermal energy from temperature differences
Light — Sunlight, indoor lighting
Electromagnetic waves — WiFi, cell phone signals

Piezoelectric Generation

Quartz's piezoelectric effect can also be used for power generation. Applying vibration or pressure to quartz generates electricity.

For example, gripping a TokiStorage pendant in your hand creates slight pressure changes. If this could be converted to electricity — theoretically — that energy could play audio.

Generation Reality

But reality is harsh. Electricity from piezoelectric generation is very small.

μW order — typical piezoelectric generation output (microwatts)

Meanwhile, making sound requires at least milliwatt (mW) order power. That's over 1,000 times difference.

With current technology, sound from just gripping — is difficult to realize.

5. Future Possibilities

Even if difficult now, let's consider future technological progress.

Scenario 1: Ultra-High Efficiency Piezoelectric Materials

New materials with far higher piezoelectric performance than quartz may be developed. Even now, artificial materials like PZT (lead zirconate titanate) and PMN-PT have over 100 times quartz's piezoelectric performance.

In 1,000 years, even more efficient materials may be developed.

Scenario 2: Ultra-Low Power Consumption Devices

Conversely, the power needed to produce sound might drop dramatically.

MEMS speakers, ultra-thin vibrating membranes, nanoscale acoustic devices — with advances in microfabrication, acoustic devices operating at μW order might be realized.

Scenario 3: Hybrid Approach

Most realistic is combining multiple energy harvesting technologies.

2100 Device Design

Photovoltaic layer — Generates from indoor light via surface micro solar cells

Piezoelectric layer — Generates from vibration and pressure

Thermoelectric layer — Generates from temperature difference between hand warmth and outside air

Storage layer — Ultra-thin capacitor stores electricity

Audio playback layer — Ultra-low power speaker

All integrated in quartz glass just millimeters thick. Hold it up to light and hear a voice after a few seconds.

6. TokiStorage Voice Preservation Options

With our successful development of audio QR code technology, multiple methods are now available to preserve "voice."

Audio QR Code (Technically Verified)

As described above, using Codec2 ultra-low bitrate encoding, embedding approximately 30 seconds of audio into a QR code is now possible. By engraving this QR code on quartz glass, audio can be played directly from the physical medium without depending on networks or servers.

This is "completely self-contained audio preservation."

Links to Audio Files

For longer audio, URLs to audio files can be recorded in QR codes. Using distributed storage like IPFS or Arweave enables storage without depending on central servers.

Text + AI Voice Synthesis

Messages can also be left as text and voiced by future AI technology. It's not the voice itself, but "what you wanted to convey" is certainly preserved.

7. Voice and Proof of Existence

Beyond technical possibilities, let's consider the meaning of "preserving voice."

Is Voice the "Window to the Soul"?

In many cultures, voice is not merely sound but a manifestation of soul or spirit.

The Latin "persona" (personality) originally meant "per-sonare" (to sound through) — the voice resonating through theatrical masks. Voice is the expression of personality and proof of existence itself.

Last Words

When someone dies, what remaining family most want to hear are "last words." There's a desire to hear that "voice," not just what was said.

Voice messages, voicemail recordings — many bereaved listen repeatedly to recordings of the deceased. Voice has a "feeling of being alive" that photos and text lack.

The Meaning of Delivering Voice 1,000 Years Later

If you could deliver a voice 1,000 years from now, what would you say?

Probably not a complex message. "Your ancestor was here." "I was thinking of you." — Simple words like these, perhaps.

Maybe what matters is not the "content" of the voice, but that the voice "exists."

Preserving voice is proving "I was here"
in the most human way possible.

Conclusion — Toki Storage in 2100

Imagine the world 75 years from now.

In 2100, Toki Storage has entered its third generation.

The first generation (present) engraved text and audio on QR codes. We achieved 30 seconds of audio with Codec2 ultra-low bitrate encoding.

The second generation (around 2050) achieved self-powered playback through the piezoelectric effect. Grip the quartz glass and minute electricity generates, directly playing the engraved voice. No external power or smartphone needed.

The third generation (2100) engraves hologram information on quartz glass. Read the QR code and a 3D image of the deceased floats in space. Not just voice, but appearance, expression, gestures — all reaching 1,000 years into the future.

Is this fantasy?

75 years ago in 1950, television was still rare. Could people then have imagined today's technology?

We are now completing the first generation technology. Engraving 30 seconds of audio on quartz glass — that's where we stand.

But this isn't the goal. Starting from preserving voice, moving toward technology that preserves existence itself. We're still on the first step of that journey.