Thank you. I've consulted ChatGPT just for curiosity
:
The “measurements are everything” camp
Their core claim is this:
If a DAC has distortion and noise far below audibility, it cannot sound different in a meaningful way.
From a
psychoacoustic and
engineering standpoint, that is largely correct. Once you are at ~−110 to −120 dB SINAD, flat frequency response, low jitter, proper linearity, and no channel imbalance, you are well past the threshold where
known, repeatable audible artifacts exist for human hearing under controlled conditions.
In other words:
- You cannot reliably hear −120 dB distortion.
- You cannot hear jitter that is orders of magnitude below audibility.
- Two DACs that output the same analog signal within those limits will null almost perfectly.
So when someone says
“good DACs all sound the same”, what they really mean is:
They do not introduce audible errors that can be proven in blind testing.
That is an important distinction.
Where that position quietly overreaches
The leap of faith happens here:
“Therefore, there are no audible differences whatsoever.”
This is where engineering certainty quietly becomes ideology.
Why?
Because
measurements do not capture the entire playback chain, nor the entire human experience.
A few points that matter, even if they are uncomfortable for purists:
1. We measure what we know how to measure
DAC measurements focus on steady-state tones, sweeps, and noise floors. Music is dynamic, time-varying, complex, and interacts with downstream components.
Output stage behavior, impedance interactions, transient behavior under load, reconstruction filters, and even clock recovery strategies can differ while still producing “excellent” numbers on standard tests.
Do these differences
always matter? No.
Can they matter in
some systems? Yes.
2. R2R vs Delta-Sigma is not magic, but it is not nothing
An R2R DAC like the Cyan 2 behaves very differently internally than a modern delta-sigma DAC like the Luxman’s ROHM chips.
That does
not mean one is “more musical” in a mystical sense.
It means:
- Different noise spectra
- Different linearity error distribution
- Different filtering approaches
- Different output stages
All of these can be engineered to be inaudible, but they are not
identical.
3. Audibility is not the same as preference
Even if two DACs are audibly transparent in blind tests,
listener preference can still diverge.
Why?
- Slight tonal shifts at the margins
- Filter choices
- Interaction with amps and speakers
- Listening fatigue
- Long-term listening vs quick A/B tests
Blind tests answer the question:
“Can you reliably detect a difference?”
They do
not answer:
“Which one do you enjoy living with more?”
The opposite camp: “measurements don’t matter”
This side often falls into its own trap.
Claims like:
- “This DAC has better microdynamics”
- “More air”
- “Blacker background”
…are frequently made
without controls,
without level matching, and
without blind testing.
Human perception is easily nudged by:
- Price
- Brand
- Aesthetics
- Expectation
- Confirmation bias
That does not mean people are lying.
It means the brain is an active participant in the experience.
Dismissing measurements entirely leads to:
- Paying thousands for zero technical improvement
- Confusing coloration with quality
- Treating subjective impressions as universal truth
That road ends in audio mysticism.
My position, clearly stated
I sit between the camps, but closer to the measurement side
with conditions.
Here is my actual stance:
- There is little to no reason to spend more than ~$1k on a DAC for sound quality alone.
I fully agree with you here.
- Beyond a certain performance threshold, differences are small, system-dependent, and often subtle.
Not imaginary, but not night-and-day either.
- Measurements are necessary, but not sufficient.
They tell you:
- If a DAC is competently designed
- If it is transparent
- If it is objectively flawed
They do
not tell you:
- If you will enjoy it more
- If it integrates better into your system
- If it makes you want to listen longer
- Aesthetics, build quality, ergonomics, and emotional connection matter.
Anyone who dismisses this is pretending humans are oscilloscopes.
If someone loves a Luxman not because it “measures better”, but because:
- It looks beautiful
- It feels like a serious instrument
- It invites listening
That is a perfectly valid reason.
About your specific comparison
Cyan 2, Topping Centaurus, and Luxman DACs all sit firmly in the
transparent DAC tier.
None of them is “objectively superior” in a way that guarantees audibility.
Any preference between them will be:
- Subtle
- System-dependent
- Personal
The Luxman’s ROHM MUS-IC chips are not magical unicorns, but they are well-regarded, and Luxman’s analog stages and power supplies are typically excellent.
If you choose Luxman, you are paying for:
- Industrial design
- Brand philosophy
- Build quality
- Pride of ownership
Not for a revolution in sound.
And that is completely fine.
Final thought
The biggest irony in this debate is that
both camps are partially right and partially wrong, but each insists on being 100% correct.
Audio is engineering
and experience.
Ignore either side, and you miss the point
