Chapter 124: Voicetalk
As Seong-hwan finished his report and took his seat, Han Yu-ju stepped forward, holding a laptop.
“I’ll report on the progress of the Voicetalk update.”
The screen displayed the title, ‘Voicetalk Beta Structure.’
Han Yu-ju began her explanation, pointing to several lines on the screen with a pointer.
“Unlike traditional voice calls, Voicetalk must ensure stable packet transmission even in 3G environments.”
Han Yu-ju, standing in front of the screen, glanced around before continuing.
“Currently, UDP-based voice packet transmission and reception have stabilized, but… the issue is with 3G networks.”
“What’s the problem?” I asked.
“Under current network standards, the latency and packet loss rates are too high. Especially with low-cost overseas 3G networks, calls frequently drop or experience delays of over 300ms, making them unusable for real-world users.”
The meeting room fell silent.
“Can’t we lower the compression rate further?”
“It’s already down to 16kbps. Going lower would severely degrade call quality. Technically, it’s possible, but it’s not feasible for a commercial service.”
In short, providing a proper Voicetalk service on 3G networks was challenging, and an upgrade to the communication infrastructure itself was necessary.
After listening to Han Yu-ju’s report, I quietly tapped my fingertips.
‘Even before the regression, Voicetalk only became popular after LTE networks. Before that, call quality was terrible.’
But back then, it was because the technology wasn’t fully developed.
Now, things were different.
I looked up and met Han Yu-ju’s eyes.
“What method does the 16kbps voice codec use?”
“It’s an upgraded version of AMR-WB, an open-source codec, but… its transmission stability compared to audio quality isn’t particularly high.”
I stood up and walked over to the whiteboard.
“……What if we did this?”
I wrote a word on the board:
[Adaptive Queue Compensation Protocol]
AQC.
“Adaptive Queue Compensation Protocol?” Han Yu-ju asked, puzzled.
“Instead of preventing packet loss, it adjusts transmission priority and minimizes delay.”
I continued my explanation.
“The traditional method uses retransmission or predictive interpolation to prevent packet loss, which increases delay. But AQC is the opposite. Based on real-time flow, it quickly extracts only the ‘necessary information’ and then ‘accumulated correction’ of deficient data in subsequent packets.”
“Hmm… So, even if the front part is interrupted, it can be recovered from the back?”
“Precisely. From the listener’s perspective, there’s no interruption. The corrected content is automatically restored with a slight delay.”
Hong Jin-man chimed in.
“The data order might get a bit scrambled, but real-time audio quality can be maintained.”
“Right. But to do that, we’ll need a dedicated AQC module for Voicetalk.”
Manufacturing costs would rise, but
compared to existing communication costs, it would still be significantly cheaper.
After hearing the full explanation, Han Yu-ju seemed receptive.
“It’s worth a try. If it works as you say, launching the Voicetalk service might not be impossible.”
“If that happens… KTalk’s expansion will accelerate beyond comparison.”
And Quantum sales would likely increase as well…
But there was an immediate constraint: production capacity (CAPA).
‘Hmm…’
To win the platform competition against A-ple, Lunar needed to become mainstream.
That meant either producing and selling as many Quantums as possible
or getting smartphone manufacturers to adopt Lunar for their devices.
“Hyung… no, Representative. If we create this AQC module, will it only be for our products?”
Seong-hwan’s words sparked an idea in my mind.
“Not necessarily… Any device using AQC would have to install KTalk as a default app.”
The more it was sold, the more it benefited us.
And to use KTalk, either jOS or Lunar was essential.
Since jOS was exclusive to A-Phone, those companies would have no choice but to adopt Lunar as well.
I asked Seong-hwan,
“How’s NokiaTel’s smartphone development going?”
NokiaTel dominated the feature phone market with a 40% global market share.
“They recently launched the NokiaTel 6700, but the response hasn’t been great.”
NokiaTel 6700.
I knew that model.
‘Their insistence on Symbian, an OS with poor usability, was probably their downfall.’
True to NokiaTel’s style,
the design was sleek, but the operating speed was slow, and the touch response was frustrating.
The biggest issue was the near-nonexistent app ecosystem.
With developers even turning away from the platform, consumers had no choice but to reject it.
After a moment’s thought, I spoke again.
“Propose the AQC Voicetalk module to NokiaTel. The terms are simple: pre-install KTalk and adopt the Lunar OS.”
The meeting room fell silent.
Han Yu-ju’s eyes lit up as she asked,
“Do you think NokiaTel will accept? They still seem attached to their own OS…”
I returned to the whiteboard and jotted down a few ways to bring NokiaTel on board:
[NokiaTel Proposal]
- Provide AQC module (free)
- Support for capacitive touch screens and related patents and technology
- Pre-install KTalk & Voicetalk
- Adopt Lunar OS
- Joint distribution partnership in the European market
“I’m not saying to propose it right now.”
“Then should we wait longer?”
I nodded.
“It would be better to propose after even the model following the NokiaTel 6700 fails miserably.”
Only after consecutive smartphone failures would the prideful NokiaTel consider our proposal.
And implementing the AQC module to enable Voicetalk would take some time.
“Still, if we can get Lunar on NokiaTel’s smartphones, the scalability would be undeniable.”
Han Yu-ju’s voice rose with excitement.
With NokiaTel’s global market share of 40%, combined with their distribution network, brand trust, and hardware production capabilities,
Lunar becoming the market ‘standard’ was only a matter of time.
I added a few more words to the whiteboard:
[Network] + [Messenger] = Platform Dominance
A few weeks later.
Dojin Soft server room.
The update server was packaging the Lunar 1.6 beta build,
and Han Yu-ju was checking the final test logs.
‘Could this really… enable real-time voice calls in 3G environments?’
Perhaps her earnest hope was answered.
A small test terminal beeped with a Voicetalk incoming call notification.
[Voicetalk Incoming Call: Test Client 02]
Han Yu-ju pressed the call button,
and a voice came through the terminal.
-Can you hear me?
Han Yu-ju smiled at the representative’s uninterrupted voice.
“It’s coming through perfectly, no delay. Representative, it’s really… working!”
Her excited voice traveled through the receiver.
-How’s the packet loss and delay?
“Can’t say there’s none, but… compared to the traditional method, it’s much more stable. The AQC module’s accumulated correction is so effective that users barely notice any interruptions.”
Han Yu-ju sent the log data via KTalk,
then continued her explanation.
“The data I just sent represents the worst-case 3G network conditions. With the traditional method, packet loss rates exceeded 20%, and delays averaged over 200ms. But with the AQC module, loss rates dropped below 2%, and delays stayed under 50ms.”
The numbers alone showed results beyond expectations.
In terms of voice call quality, it was almost indistinguishable from regular calls.
‘As expected.’
I knew it wouldn’t fail,
but seeing it succeed was still overwhelmingly moving.
-You’ve all worked so hard. At this rate, we can prepare for service launch.
Han Yu-ju replied brightly,
“I’ll get everything ready to start the beta service next month.”
…
…
That same afternoon, Dojin Tech conference room.
Present were myself, Hong Chung-ki, Hwang Chilbong, and Oh Seong-hak.
On the table lay a small, square AQC module sample.
Hong Chung-ki picked up the module and said,
“I’ve heard the test results are good. So, we can start full-scale production of the AQC module now?”
“The structure is simple, so production shouldn’t be difficult.”
Turning to Oh Seong-hak,
I asked a follow-up question.
“Have you calculated the manufacturing cost?”
“Yes, I got a rough estimate from the procurement team. The production cost per module is about 5,000 won. Compared to communication chips, it’s not extremely cheap, but considering its utility, it’s practically free.”
I agreed.
Even domestic call charges at the time were significant.
With this, people overseas could make free calls anytime.
The impact would be beyond imagination.
“Once this is released, KTalk downloads will skyrocket.”
“And everyone will rush to buy Quantum.”
Just thinking about it felt good.
But unlike me, Hong Chung-ki looked uneasy as he stared at me.
“But… Representative, Voicetalk won’t work on already sold Quantums, right?”
He was concerned about potential complaints from existing users.
But I had a plan for that too.
“For previously sold units, an external AQC module can be connected to the earphone jack to enable Voicetalk without issues.”
I gestured,
and Chilbong pulled out an earphone-shaped external AQC module and placed it on the table.
“Oh, with this design, existing users can easily use Voicetalk.”
“Not just Quantum, but any smartphone with Lunar installed can use Voicetalk with these earphones.”
I took a sip of coffee from the table and continued.
“So, Director Hong, focus solely on securing the supply.”
It would sell like hotcakes once released,
so securing inventory was the most critical point.
“Quantum has sold about 7 million units so far, so we’ll need at least that many.”
“7 million isn’t enough. With the AQC module, we’ll need at least 100 million units.”
My words silenced the room.
“We’ll use some ourselves, but we’ll also need to supply NokiaTel for Lunar’s expansion. Depending on the situation, we might even sell them to Ahsung or A-ple at a premium.”