Still caught up with work, so no time to talk to DUNU lately.
However, I did notice that DUNU posted an official screenshot of the DK-4001 frequency response on their Weibo. They did it in response to some third-party reviewer posting independent measurements of the DK-4001. I cannot find the third-party's post anywhere; I'm just not that adept at navigating around Chinese social media.
However, here's a cropped version of the FR that DUNU posted:
Keep in mind these things when you're examining this screenshot:
- Take a look at the horizontal and vertical axes; the vertical axis runs from 30 to 150 dB, while the horizontal axis runs from 20 Hz to 50 kHz
- DUNU measures its earphones on a Bruel & Kjaer Type 4157 ear simulator. While the 4157 is compliant with the IEC-60318-4 (former IEC-60711) industry standard, it should not be considered comparable to the newer ear simulators from GRAS that reduce the high-Q resonance in the upper frequencies.
- Additionally, because each person/company's calibration methods are different, cross-rig comparison should be kept at a minimum. At most, frequencies between 100 Hz and 8 kHz are more reliably comparable across different 711 type couplers. Non-711 type couplers should only be directly comparable to other measurements made by the same type of ear simulator (e.g. miniDSP EARS vs. miniDSP EARS, and Veritas vs. Veritas).
- There seems to be some degree of octave smoothing applied, but the amount is unknown. I would guess it's about 1/8th octave.
Because I thought the vertical axis was a bit wide, I wanted to take a closer look. Thus, I decided to re-render the FR from scratch, i.e. trace over the data points and plot them onto a scatter plot.
Here's how it looks with the axes unedited:
I tried to get the spectral plot to look as close to the original as possible without spending too much time splitting hairs. Under gross examination, it looks passable.
Then, I merely readjusted the vertical axes:
The peaks now look a lot more apparent, and it's easier to take some inferring remarks on the possible sound of the DK-4001.
I notice the following things:
- The 2-3 kHz canal resonance peak isn't gone, but buried behind a ~1.5 kHz midrange boost peak, which is between 3-6 dB in intensity above the canal resonance peak. It is likely this midrange "boost" is an effort to bring out the character of the Beryllium dynamic driver, but this is only a guess.
- The bass shelf is very reasonable ~5 dB above the 600-800 Hz nadir. It seems relatively flat down to 20 Hz.
- The broad treble peak rises ~ 4-6 dB above what I would consider a dead-neutral response.
- I have noticed from previous FRs that DUNU has released in the past, perhaps because of their testing methodology, almost all of their earphones tend to fall off precipitously after 8.5 kHz or so. I really don't know why this is the case. However, DUNU has allocated a dual super-tweeter for these upper frequencies, and I doubt they'd let it go to waste when they've had years of experience with building hybrid IEMs. It's tough to really assess the degree of treble extension unless we know exactly how DUNU inserts its earphones into its testing rig, and even the subsequent assessment of treble extension is not universally standardized.
- Because re-rendering the FR smooths things out even more, I cannot really speak to small subtleties. I am not really surprised by the way this FR appears, and overall it seems to be a moderately "W-shaped" sound signature, which is totally appropriate for a flagship-type earphone, and seems to be in keeping with what DUNU has been working towards. The main thing that was unexpected for me is the 1-2 kHz "boost", which almost surely has been done for musical tuning purposes. There are a number of IEM models that share a similar trait, including, but not limited to, the CFA Andromeda, the Sony EX1000, etc. Bear in mind that I am not stating that the DK-4001 sounds like these IEMs, because each of these models adjusts the relative level of the 1-2 kHz against the level of the 2-3 kHz area, as well as the 600-1000 Hz area, in different ways. In my experience, small changes as little as 1.5 dB in this extremely hearing-sensitive area can make very noticeable changes to the way a frequency response is perceived. Combined with differences in phase response, I find it very difficult to draw any solid conclusions regarding the sound of the DK-4001 beyond its moderately W-shaped response.
