Recovery and restoration service:frontea online,corp.
‹ 2026/01/03 ›
Hello again.
This is Episode 9.
Last time I said “next episode will finally reach the main battlefield,” but…
I’m inserting one chapter before that.
Today, we’re doing the UAC answer check.
Back in the early episodes of this series, I wrote:
“PopoDAC will use UAC1.0 Feedback.”
Today, we verify that decision.
Some people may think:
- “In 2026, why would you NOT use UAC2? What’s the point of a DAC then?”
- “It’s not Hi‑Res.”
- “Total waste of effort.”
Hold on a moment.
UAC2 is high‑performance.
But the question is:
“Is UAC2 necessary for playback sound quality?”
Once you understand the actual playback route, another answer will naturally appear.
That is the topic for today.
UAC (USB Audio Class) is the audio protocol defined by the USB‑IF (USB Implementers Forum) for attaching external audio devices to a USB host.
There are two major versions:
At first glance, UAC2 looks like a simple “performance‑upgraded UAC1.”
From the USB specification perspective, it looks like all upside.
But I want you to pay attention to two things:
IT companies and audio companies appear, disappear, and switch sides.
You can almost smell the politics, power balance, and negotiations of the time.
It’s thick in the air.
According to Microsoft’s Windows USB Audio 2.0 Driver documentation, look at the driver architecture diagram.
You’ll notice:
UAC2.0 sits two layers farther from the Core than UAC1.0.
This means UAC2 is treated more like a plugin,
not a core, fundamental audio path.
Modern PCs almost universally include Intel HDA internally.
And HDA already supports output quality equivalent to UAC2‑class audio.
In other words:
High‑quality audio output is already built into the PC’s internal CoreAudio device.
HDA was proposed by Intel in 2004, with version 1.0 released the same year, and 1.0a in 2010.
Let’s return to the core question:
Ultimately:
UAC is a specification for adding an external audio device to a USB host.
And UAC has two major roles:
That’s it.
Let’s take another look at sound quality.
You may ask:
The answer is yes—and no.
More precisely:
They determine the quality of the recorded audio source, not the quality of playback.
So the answer is a mixture of both yes and no.
Let’s break down the mechanism in a simple way.
Both parameters are the numerical procedures required to convert sound—
a vibration, a wave—into digital form.
That’s all.
The answer is straightforward.
Human hearing, even for highly capable individuals, is roughly 20 Hz to 20 kHz.
Digital sampling, however, captures only points on the waveform.
A waveform is continuous, but sampling extracts only fragments of it.
Because frequency is a wave phenomenon, sampling inherently removes part of the wave’s original nature.
And when reconstructing the waveform, only half of the captured bandwidth can be restored.
(If you record a waveform as discrete points, the high‑frequency components between the points can “transform” into entirely different frequencies—aliasing.
I’ve experienced this firsthand while building detectors… quite the mysterious world.)
The classic CD sample rate, 44.1 kHz, has a Nyquist frequency of 22.05 kHz.
This comfortably covers the human audible range.
Thus, at the time, the industry settled on:
“44.1 kHz is sufficient to reproduce the audible spectrum effectively.”
And that was the consensus.
Let’s return to the fundamentals:
“What is the audible range?”
You might say:
“It’s the range of sounds humans can hear, right?”
That’s partially correct.
The full answer is:
“The audible range is the portion of sound waves that humans can perceive, within the broader range of sound that physically exists.”
This distinction is extremely important.
Please keep it in mind.
Some time ago, I built an ultrasonic detector because I wanted to hear the voices of birds and insects more clearly.
I took it into the mountains.
Here is what I observed.
When the forest became noisy with insects and birds, the detector picked up a strong peak around 27 kHz—ultrasound.
Ultrasound is famously used by bats and moths.
Dog whistles also operate in this region.
Many creatures can perceive ultrasonic frequencies far beyond human hearing.
Insects, birds, dolphins—
a surprising number of animals use ultrasonic bands for detection, communication, and even conversation.
When I built the detector, I tested it by striking two metal spoons together and observing the waveform.
To my ears, I heard the familiar “keen” metallic ring.
But the detector showed peaks at:
A dog would certainly hear all the way up to that final 32 kHz component.
The true sound of a spoon includes ultrasonic components that humans cannot hear.
The “real” spoon sound is the entire spectrum, not just the portion humans perceive.
Brass instruments, cymbals, strings—
every acoustic instrument produces harmonics and overtones that extend well beyond the human audible range.
Therefore:
“The integrity of the original sound extends into the ultrasonic domain.”
This is a crucial concept.
Professionals in the field already know this—and they work hard for it.
To convey the creator’s sound in its best possible form, they try to capture every part of the sound the creator produced, without losing anything.
When you do that, the effective range of the sound waves expands dramatically.
If you assume that capturing up to 60 kHz is necessary to preserve everything,
then the required sample rate becomes at least 192 kHz.
If you assume that capturing up to 48 kHz is sufficient, then 96 kHz is the minimum.
With that, you can say:
“This recording captures essentially the full sound of this performance.”
One way to view Hi‑Res is:
Hi‑Res expands the recorded bandwidth to match the effective range of the musical sound.
Bit depth represents the resolution—the restoring power—of each individual quantum (sample point).
Of course, deeper bit depth is better.
But because the information increases exponentially, there is a practical balance.
For comparison:
When you look at photos, 10‑bit displays show richer gradients.
But if the original photo is poor, even 10‑bit depth won’t make a visible difference.
In the same way:
Hi‑Res means “the recording captures a wider, more complete range of the original sound,”
but whether playback can enjoy Hi‑Res depends entirely on whether the playback route can faithfully reproduce that source.
Now we arrive at the key question:
“What are the quanta (particles) of sound that lie outside the audible range?”
And ultimately—
“Do we actually need them?”
The answer is yes.
We already established that ultrasonic components included during digitization are still part of the real sound.
Even if humans cannot hear them directly, they remain valid components of the sound wave.
And during playback:
Every quantum—audible or not—returns to the air as a physical sound wave.
It still reaches the listener.
It still vibrates the room.
It still interacts with the audible components.
All quanta, including ultrasonic ones, are part of the creator’s energy embedded in the music.
They may contribute to impact.
They may contribute to emotional intensity.
They are valuable energy—not something to discard lightly.
This is why mastering requires higher‑bandwidth recording, and why the playback side must secure a route capable of reproducing that energy.
And now, looking at the reality of playback in 2026, you may already see the conclusion:
For playback, the UAC2.0 specification is not required at all.
For listeners, 24‑bit/96 kHz, or at most 32‑bit/96 kHz, is entirely sufficient.
What matters far more is:
Whether the playback chain can output the source material at that quality without loss.
Securing that route is the true priority.
PopoDAC is being created precisely to provide that route.
Some readers may have already noticed:
“Wait… if PopoDAC outputs the full bandwidth, a normal amplifier won’t be able to reproduce it… right?”
Correct.
You are absolutely right.
To enjoy music from PopoDAC, you cannot pass it through a typical consumer amplifier.
You will need professional‑grade or dedicated hardware.
Specifically:
You need a front‑end and output stage that does not apply a low‑pass filter before 48 kHz, and intentionally allows output up to that range.
Many amplifiers include an input RC filter around 20 kHz.
I’ve built such designs myself in the past.
But with PopoDAC, that kind of amplifier cannot deliver its full potential.
(I realized this myself only later…)
And so I thought:
“Ah… I need to build a dedicated amplifier.”
(And yes, I laughed at myself.)
In the next chapter, we finally return to the operational core of PopoDAC.
Email
*Requests and feedback
