Honing Your Craft…
How did you first develop your process?
I learned most of the basics from Unique Melody, but I also watched that video from 1964Ears – a very detailed tutorial into making custom IEMs. I watched it over and over and over again!
Image courtesy of 64Audio
They’re 3D-printing now!
Yes, but in those days, they used hand-poured UV resin, so it was still really helpful for me.
Do you think you’ll ever pick up 3D-printing?
Is it not clean enough (cosmetically), perhaps?
Yes! That’s one of the reasons. Actually, the KL-REF and Lakh did use 3D-printing for the housing of the DDs and the BAs. But, I don’t think I’ll make shells with 3D-printing, because of the reason you said.
Having examined the insides of – as you said – over 100 IEMs, how much did you learn from the process?
I think I learned most from Unique Melody and Ultimate Ears. Oh, and Heir Audio as well. The 10.a had a very unique tuning in its sound tubes and crossover network, which I learned a lot from. Those days, custom IEMs would only use two components: A capacitor for the tweeter and a resistor for the bass. I wanted to improve that system by adding more components. Taking the KL-Akara as an example, I attached a narrow tube onto the bass driver to cut high frequencies. I think that was the first time a narrow tube was used acoustically like this for custom IEMs. Of course, Shure’s SE846 used a very narrow tube as well, but the Akara may have been the first custom.
How have you then taken that experience and developed your own house sound?
Well, my favourite sound signatures were the UE TF10’s and the Shure SE530’s, and all I wanted to do was improve those sounds. I thought the TF10 would’ve been better with more treble, and the SE530 with more bass. These were my very first – primitive, even – sonic goals. These two were the start of Sanka, and so on.
Does the music you listen play a role in that as well?
When I was developing the NEXT 5 series, I listened to a lot of rock and J-Pop. The Meteo was created with hip-hop and EDM, as well as rock.
I see, that’s where all that impact comes from.
What do you prioritise most when tuning?
I think detail is the first quality that comes to mind. But, the first thing I always do when tuning is imagine the target frequency response in my head – whether it be through the BAs or the crossover network.
Ahh, that’s very interesting! So, you picture the graph in your head?
Yes, exactly! I don’t know if this is the way engineers usually work. An engineer friend of mine, for example, will listen to his one favourite song and tune by ear until it sounds good through the earphones. That’s usually his first step towards an IEM design, while mine starts with the frequency response. But of course, there are so many other ways out there too.
Were there moments where you’ve perfectly matched the graph in your head, but the outcome wasn’t what you expected?
Oh yes, always! (laughs) I have to do lots of trial-and-error – sometimes changing the design drastically between each attempt. Often, it’s tweaking the network or the acoustic filters, until I achieve the ideal sound.
Are there some things the graph can’t tell you?
I think one is THD (Total Harmonic Distortion). A frequency graph can’t read that. When I measure an IEM and the frequency response graph is perfect, but there’s a noisiness to the sound when I listen to it, it’s usually a problem with THD. When I have those problems, I’d change the balanced-armature drivers to reduce that distortion.
A Dynamic Shift…
After developing a couple in-ears with pure-BA designs, you transitioned to dynamic hybrids with the KL-REF and the Trio. What made you want to explore dynamic drivers?
Dynamic drivers have very low distortion in the bass frequencies, which was what first interested me in them. After using several high-quality dynamics in the market, I finally found FOSTEX’s driver units. This led to the KL-REF with two FOSTEX dynamic drivers and three BAs.
And, I heard you’re now working with FOSTER (FOSTEX’s parent company) as part of their Alliance program!
This means you – and any other company participating in the program – will gain access to FOSTER’s arsenal of proprietary dynamic drivers to R&D all-new products. What’s the experience been like so far?
Well, first I asked FOSTER what diameters they were willing to offer me. The sales person I talked to said they had 6.3mm, 9mm and 14mm. I selected the largest of the three.
Image courtesy of CNET Japan
When do you think that driver’s gonna make its way into a commercially-available product?
Still picturing the graph?
Yes! Still imagining the graph.
You had a prototype with the driver at a recent show, I believe?
Ah, yes. It was the Fujiya Avic show this past October. That one had the 14mm dynamic driver with 2 BAs.
A 14mm dynamic driver is rather rare, as far as I know. I often see 9mm, 10mm…
Yes, it is quite large. But, Sony’s Just Ear monitors also use a 13mm driver. Fender too, with six BAs.
The REF had two 9mm dynamic drivers in addition to the armatures. What are the main differences, do you think, between two 9mm drivers and a single 14mm one?
Well, they use totally different materials. The 14mm driver uses a paper diaphragm, while the 9mm uses a bio-cellulose diaphragm. But, I think the biggest difference between them is the volume of the bass. The 14mm driver has a huge, rich bass sound, which is what made me select it.
Ah, ‘sounds like it’ll be a departure from the NEXT 5 series; more Japanese.
Do you think there is such a thing as a Japanese sound? What’s the trend around here?
Hmmm… I think about six years ago, the Japanese market liked a fuller, bassy sound. But, this changed three years later; treble became very popular in Japan. And nowadays, I think that huge bass sound is coming back!
Always shifting seasons, I guess. But, that leads me to my next question: One of your mainstays is the ability to self-tune the bass of your IEMs – from the Trio, to the REF, to your brand new Focus. Why do you consider this feature so important?
I think preference with regards to bass volume can be very different from one person to another. Including a variable bass system – to me – is very good for the consumer.
But, not many companies choose to do so. Do you think there’s a challenge to it that not many are willing to tackle?
A variable system like this has a big defect risk – I think that’s the major reason. Even a big company like JHAudio, when they first introduced variable bass around three years ago, were forced to face these issues. But, I want our customers to enjoy our products as much as they can, so I was willing to take that risk.
Are there any technological challenges when designing an IEM with variable bass? Like crossover design, for example?
Well, applying a switch mechanism to crossover paths is very difficult, so we’ve never done it. For example, the NEXT 5 IEMs use a very unique network to achieve their bass. We used an electronic low-pass filter for the low driver, which is common in speakers, but quite unique in IEMs. Because of this, we couldn’t apply the variable bass system there.