Kali's latest speakers demonstrate some serious electro-acoustic know-how — and represent remarkable value for money.
Kali Audio are relative newcomers to the entry-level nearfield monitor market, and our esteemed Executive Editor Paul White reviewed the company's least expensive model, the LP‑6, back in the December 2018 issue. On that occasion, Paul found the LP‑6 to be a very capable performer, especially considering its genuinely affordable price, and gave it a definite Sound On Sound blessing.
So the subject of this review, Kali's new IN‑8, is one we've been anticipating with great interest, not least because, in contrast to the relatively safe and conventional nature of the LP‑6 (and its larger sibling, the LP‑8), the IN‑8 is significantly more ambitious, while still managing to hit a seriously competitive price.
I mention the LP‑8 because the IN‑8 is actually somewhat related. The architecture of the LP‑8 is a 200mm bass/mid driver integrated with a generously waveguide–loaded 25mm fabric dome tweeter. For the IN‑8, Kali have borrowed the low-frequency elements of the LP‑8 — a similar driver, similar enclosure dimensions and reflex port — but swapped out the waveguide and tweeter for a concentrically arranged compound mid driver and tweeter. It essentially turns the two-way LP‑8 into a three-way design — one that offers the significant benefits of dual-concentric, or rather, dual-coincident, drivers.
I'll describe the benefits of dual-coincident drivers in a few paragraphs time, but first there's a little more general IN‑8 description to run through. The IN‑8 is not a small monitor. I had read the dimensions from the spec sheet before the monitors arrived but I was still slightly surprised at just how substantial an IN‑8 turned out to be in the flesh. At 10.4kg each they're not particularly heavy, so there ought not to be any great issue with mounting arrangements.
The IN‑8's all-black finish comprises a vinyl wrap applied to the main carcass of the MDF enclosure, with a black–painted, plastic, injection–moulded front trim panel. The cabinet side panels sound a little live to the 'knuckle wrap' test, and the alignment of the plastic trim panel and the cabinet is slightly variable, but then at the remarkably inexpensive price of the IN‑8 it'd be unfair to expect the finish quality of a Rolls Royce.
The trim panel forms the mouth of the reflex port and blends the drivers gently into the front surface. Said port mouth has an interesting profile: half letterbox and half curved as it follows the lower profile of the bass/mid driver. The port mouth also carries a radius at its front edge and an internal profile that Kali write in the IN‑8 brochure is specifically designed to, "ensure that all of the air leaves the port tube at the same velocity". In reality, laminar airflow requires that velocity near the outside of a tube is lower than the velocity at the centre (much like the water near the bank of a smoothly flowing river moves more slowly than the water mid-stream), so ensuring that the velocity is equal seems to me to be slightly contrary. Perhaps I've fallen into the trap of assuming sales copy has watertight technical validity, but having said that, it's good to see a monitor manufacturer acknowledging that port airflow is important, and taking steps to ensure it stays laminar to as high a volume level as possible.
The bass driver is a nominally 200mm (8–inch) unit with a polymer–coated paper diaphragm. This style of diaphragm construction is becoming more common, and I suspect the appeal is that it offers the advantages of paper's light weight and rigidity combined with the damping and more 'engineered' look of plastic. The dust cap of the IN‑8 bass driver is slightly unusual in that it is fitted in concave orientation rather than the more usual convex. Probably of more significance in terms of the performance of the driver, however, is that the dust cap is attached to the end of the voice coil former rather than, as is more often the case, to the diaphragm. All other things being equal it's a subtle little manufacturing detail that I like for its purity and simplicity.
Having said that the bass driver is 200mm in diameter, I should mention that its diaphragm is actually only 160mm. This very common anomaly, one that is not by any means exclusive to Kali, comes about because the standard driver diameters (5–inch, 8–inch, 10–inch, 12–inch, etc, which I believe date back to an ancient British standard) are traditionally based on chassis size rather than diaphragm size. However, as the bass driver chassis on the IN‑8 is hidden behind a trim (as with the majority of monitors these days) we never really get to see the justification for that 200mm specification (don't ever let anybody tell you the world of monitors isn't truly riveting).
Whereas the bass driver is relatively conventional in appearance and construction, the dual-coincident compound mid and tweeter driver is of course much less run of the mill. I've reviewed a number of monitors in these pages over the last few years that have featured a dual-concentric or dual-coincident (where the tweeter is located right at the mid diaphragm apex) arrangement of mid driver and tweeter, but aside from the KEF LS50 Wireless, the driver fitted to the IN‑8 is perhaps the purest expression of the idea to date — so much so that I believe it would almost certainly have fallen foul of the original KEF dual-coincident patent. The IN‑8's 25mm fabric dome tweeter is located such that the mid driver diaphragm, along with playing its primary role, effectively acts as a tweeter waveguide and helps ensure that there's no sharp discontinuity of dispersion characteristics through the mid–range/tweeter crossover. Along with the time-domain benefits and off-axis response consistency that dual-coincident drivers provide, the lack of a crossover dispersion discontinuity is an equally significant benefit.
Something else somewhat unusual about the compound driver is that the mid–range diaphragm has a very small (almost non-existent) surround component. There's two reasons for this, both of which speak of the advantages of a three-way format over a two-way format. The first reason is that, with no need to play any bass, the diaphragm movement required of the mid–range driver is minimal — a couple of millimetres at most — so the large half-roll surround that's typical of bass and bass/mid drivers, and enables generous diaphragm movement, is simply unnecessary. This is of particular benefit in dual-coincident drivers that employ the mid–range diaphragm as a tweeter waveguide, because conventional roll surrounds can result in a significant and unhelpful discontinuity in the waveguide profile. The second reason why a minimal surround approach is a Good Thing is that the profile and materials that work best for roll surrounds in bass drivers tend not to work very well at mid–range frequencies. This often results in what speaker engineers know as, "the surround problem". Look at the frequency response and time-domain waterfall curves for almost any conventional 5- to 8-inch bass/mid driver and you'll often see a characteristic resonance problem between, say, 750Hz and 1.5kHz. It's caused by surround 'mis-termination'. Rather than absorbing the energy in the mid–range diaphragm as it speeds outward from the voice-coil, the surround reflects the energy back, which results in a resonance. One solution to this in drivers that aren't required to play any bass is to optimise the surround for mid–range frequencies, and in some cases that means the surround is all but non-existent. The first surroundless mid–range driver I was aware of was developed by Mike Deadman at British hi-fi speaker company Mordaunt-Short in the early 1980s. Moving to contemporary times, many current Bowers & Wilkins mid–range drivers have, effectively, no traditional surround.
The IN‑8 is truly exceptional in that it offers genuinely extended low-frequency bandwidth, midfield volume level abilities, and mid-range clarity and imaging to die for. And it does all that at a seriously competitive price.
There's one last aspect of the mid–range driver to cover. I mentioned that in three-way systems, the mid driver is typically only required to move a millimetre or two. Along with generally making the mid–range surround design easier, this is good news in the context of dual-coincident drivers because it means such minimal movement of the diaphragm won't modulate the tweeter output to any significant degree. The mid–range diaphragm is, after all, acting as a waveguide for the tweeter, so it obviously wouldn't be a good thing to have it pumping backwards and forwards by a significant amount. In two-way dual-coincident systems, movement of the bass/mid diaphragm at low frequencies can indeed modulate the tweeter frequency response.
Having pretty much exhausted the opportunities for discourse on the drivers and front panel, I'll turn the IN‑8 around and describe what's to be found on its rear panel. First, the connections: there's an IEC mains socket, an unbalanced RCA phono input, a balanced TRS jack input and a balanced XLR input. From that you'll gather that the IN‑8's internal signal chain is entirely analogue. Then there's a mains switch, an input sensitivity control and some EQ options. The input sensitivity control offers - (ie. silent) to +6dB adjustment, with a detent at 0dB. My traditional nit-picking grouse applies here, because unless both monitors are set to either 0dB or +6dB it's not easy to be sure they are level matched. The IN‑8 user manual suggests that different levels for each monitor of a pair might be appropriate in asymmetric listening installations, but you can't compensate for asymmetric monitor positioning with relative level adjustment alone, you have to adjust the relative delays too, so I'd argue strongly for fixing the asymmetry. Before I get on to the EQ options I'll mention that the internal Class–D amplification is rated at 60 Watts and 40 Watts for the bass driver and mid–range/tweeter respectively, and that the crossover frequencies are 330Hz and 3kHz.
The most interesting aspect of the IN‑8's rear panel is the EQ facilities it carries. A DIP switch module enables 12 different EQ options to be selected. Eight of the options are specific profiles designed to adapt the IN‑8 to different installation positions, while the final four engage 2dB of LF or HF boost or cut. The eight position–specific EQ profiles were developed from acoustic measurements made at Village Studios in LA and are usefully illustrated on the IN‑8 back panel. I had the monitors installed on my usual über-sturdy wall brackets either side of my DAW screen, but I spent a bit of time trying some alternative EQ settings before finally settling on their recommended wall-mount option. I also experimented with both landscape and upside–down orientation of the IN‑8s to see if either would help integration with my room acoustics. Such versatility is more workable on dual-coincident monitors than it is with conventional monitors, because dispersion character is much less variable with orientation. For what it's worth, I actually ended up preferring the IN‑8s in landscape orientation.
I'll finish off a few paragraphs down with some FuzzMeasure response curves, but before going there, I'll deal with the small matter of what I heard when I fired up the IN‑8s and listened. The first thing I ought to mention, and perhaps get out of the way, is that the electronics are a little noisy. Nearfield monitors have a tough gig on the noise front because, when listening relatively close, almost any residual noise from the electronics is hard to miss. When the IN‑8s are playing some busy material, the residual noise is very quickly masked, but when asked to play some unaccompanied piano, for example, the hiss is audible in the gaps. It's perhaps unreasonable to expect the impeccable engineering of serious high-end audio to be a divine right at an entry-level price though, so maybe a little hiss is acceptable. A second slight niggle was that the IN‑8's auto-standby feature occasionally triggered at too high a level. I tend to have review nearfield monitors playing along quietly most of the time while I'm working on other things, and on a few occasions I had to raise the volume a little higher than ideal to stop one or other of the speakers spontaneously switching to standby. Perhaps the signal sensing threshold could be set a little lower.
I've only described how the IN‑8s behave when they're not playing anything, so I'd better put that right. When actually listening to music, my first impressions were of a warm tonal balance and of seriously extended low-frequency bandwidth. In fact the bass was too much in my room until I configured the EQ DIP switches for the wall–mounted setting, and I also engaged the -2dB LF shelf. Those two tweaks helped significantly to bring the bass under more control, but while the IN‑8s still displayed a very generous level of low-frequency bandwidth, they also seemed to me to add a degree of extra heft to kick drums, floor toms and bass guitar fundamentals that I was never entirely comfortable with. So, while the IN‑8's subjective low-frequency performance is impressive with its extended bandwidth, and probably more than competitive in those terms with other similarly sized ported monitors, I'd have preferred a trade-off between low-frequency bandwidth and time-domain accuracy that favoured the latter more (see box). Your mileage may of course vary, and regardless of how accurate the low-frequency time-domain performance, it is sometimes extremely valuable to be able to hear if anything is going on down there. The bass alignment didn't quite work for me, but you may love it.
Even with the wall–mount setting and -2dB LF EQ engaged the overall balance of the IN‑8 still seemed a little warm, so I also engaged the +2dB HF shelf, which again helped bring the balance back into more familiar territory. And moving away from the bass end of the bandwidth, the IN‑8's mid– and high–frequency performance was extremely impressive, and provided in abundance all the characteristics of 'hear-through' clarity and pin–sharp stereo imaging that you'd hope a dual-coincident driver would offer. It's a startling level of performance that renders a genuinely high level of mix detail easily audible without resorting to an over-emphasised mid balance. It's all the more remarkable in the context of the IN‑8's competitive price. Also impressive for the price is the IN‑8's ability to play at genuinely high levels without significantly changing tonal character or sounding strained. In terms of low-frequency bandwidth and volume capabilities, the IN‑8 approaches midfield monitoring territory. The only minor fly in the ointment was that my feelings about slightly live cabinet panels were I think confirmed by a mild over-warm and boxy subjective character, most noticeable on exposed male vocals. It's something you'd be most likely to notice if your work majors on voices and acoustic instruments, but if it doesn't, you'd perhaps wonder what my fuss was about. I guess the bottom line is that while the IN‑8 doesn't reach BBC monitor levels of natural clarity on voice recordings, it probably doesn't need to for the roles it's most likely to fulfil.
So, to the IN‑8 FuzzMeasure analysis I mentioned a few paragraphs ago. One thing I discovered straight away with the IN‑8 is that it produces slightly flatter frequency response curves when measured with the microphone slightly off axis (about 15 degrees). This 'flatter off-axis' phenomenon is not uncommon and is simply a result of the response discontinuities caused by diffraction from symmetrical cabinet features being averaged out by asymmetry. An example is the ridge around the IN‑8's mid–range driver. When measured with the microphone on the central axis the ridge appears to cause a discontinuity from around 8 to 12 kHz. But measured slightly off axis, asymmetry averages things out and suppresses the discontinuity.
Diagram 1 shows the IN‑8's basic full–band frequency response 15 degrees off axis from 20Hz to 20kHz, along with second- and third–harmonic distortion. The response is suitably flat and well controlled (with no sign of any mid–range surround resonance), and confirms Kali's impressive low-frequency bandwidth spec of -3dB at 45Hz. The suck-out features at 5.5 and 12 kHz are likely a result of how the tweeter radiation integrates with the mid–range driver and the physical structures nearby. Both suck-out features are sensitive to the exact measuring microphone position so are likely to be relatively innocuous in subjective terms. The harmonic distortion performance is very good with, as is usual for moving-coil speakers, an increase at low frequencies. To put some numbers on "very good", peak second harmonic, referenced to 85dB at 1m, is around 30dB down on the fundamental, which equates to about 1.4 percent. Above around 200Hz, the distortion falls typically to well below 0.5 percent.
Diagram 2, just for interest, shows the output from the IN‑8 port captured by a microphone placed a centimetre or two inside. Even though the port data typically captured by this type of measurement is unavoidably 'contaminated' by the bass driver output (especially when the bass driver is so close) it can still reveal if there's any undesirable 'organ-pipe' resonances going on in the port. As revealed in the curve though, the IN‑8 port looks completely clean. A very good, and surprisingly unusual, result.
Reading back, I fear I might have been a bit too tough on the IN‑8 and given the impression that it's flawed. Well it is, but then so, in one way or another, is almost every monitor. The IN‑8's flaws are typical of many, many monitors and in those terms it's in very good company. However, along with being a fairly typical monitor, the IN‑8 is truly exceptional in that it offers genuinely extended low-frequency bandwidth, midfield volume level abilities, and mid–range clarity and imaging to die for. And it does all that at a seriously competitive price. In those terms it's a remarkable monitor and a genuinely impressive example of electro-acoustic design. Although I didn't entirely get on with the way it plays bass, its character could probably be learned in much the same way that we all learn to accommodate the quirks of our regular monitors. The IN‑8 is in many ways, but especially for its dual-coincident driver, an exceptionally capable monitor and, once again, it's remarkably inexpensive.
Genuine three-way monitors around the price of the IN‑8 are extremely rare. The closest is probably the KRK Rokit RP10-3. And inexpensive three-way monitors with a dual-coincident driver? Well, they previously ranked with unicorns. There are, however, some very capable conventional two-way monitors at around the same price. Models from PreSonus, ADAM and KRK, for example, would be worth a listen.
I referred in the main text to the trade-off between low-frequency bandwidth and time-domain accuracy, and it's something that effectively every speaker design has to negotiate. Given a fixed driver diameter, magnet power and cabinet volume you can design a system with extended LF bandwidth and higher levels of group delay and overhang (the system ringing after the signal has stopped), or you can be more conservative on bandwidth in the interests of lower group delay and overhang. To demonstrate this I've modelled a low-frequency system of similar enclosure and driver size to the IN‑8, and generated some curves illustrating a couple of bandwidth extension and time-domain options, which you can see on the next page.
In the first option, which I've called 'Extended', the bandwidth reaches 50Hz at -3dB, the group delay peaks at 12ms (again at 50Hz) and the step response shows the system ringing for one and a half cycles. The port is tuned to 50Hz in this system.
In the second option, which I've called 'Conservative', the port is tuned to 30Hz, which results in an increase in the -3dB point to about 70Hz and a drop in 50Hz group delay to about 5ms (yes, the group delay increases significantly at lower frequencies but they are not so musically significant). The system also rings only for a cycle.
Of course modelling such as this doesn't reflect the real world, where things are somewhat more messy and complex (for example, getting a 30Hz port to work well in a relatively small enclosure is trickier than for a 50Hz port), but hopefully it goes some way to explaining the electro-acoustic issues.
- Exceptional mid–range and high–frequency performance from the dual-coincident driver.
- Very extended low-frequency bandwidth.
- Capable of high volume levels.
- Hyper-competitive price.
- Slightly noisy electronics.
- Bass somewhat overblown and slightly sluggish.
- Mild lower mid–range coloration.
Though not without its quirks, there's no denying the IN‑8 is a huge slice of monitoring bandwidth for the money, and its dual‑coincident driver is exceptional. It's very much worth hearing.