Today’s review focuses on the MV6, a speaker produced by Karak Electronics, a manufacturer better known under its official brand name, Soundis, in Korea.
I had previously reviewed this product through another community, and I recall pointing out several performance shortcomings at that time.
Not long after, Karak Electronics reached out to me directly.
Since I have rarely encountered manufacturers who respond positively to criticism of their products, I was quite nervous. However, Karak Electronics surprised me by actively working to improve on the feedback I provided and asked me to remeasure and review the updated product.
I don’t know how others might feel, but I was personally quite moved. It’s rare for a manufacturer to embrace criticism and attempt to address the shortcomings highlighted by a reviewer.
In this review, I will focus on the product’s characteristics while also highlighting the changes and improvements made since my initial assessment.
---
Frequency Response
There’s a dip of about 2-3dB across an octave in the upper treble range, but aside from that, the overall tonal balance has significantly improved. The bass extends down to approximately 55.7 Hz (-6dB). I didn’t provide a trendline due to the irregular slope of the bass roll-off.
Let’s compare the tonal balance improvements with the earlier data.
The red graph shows the on-axis frequency response data of the initial sample I measured previously. Compared to the earlier version, which had pronounced peaks and dips across a wide range, this newer version is much flatter.
While it may not have been possible to change the physical structure or the drivers, the tonal balance was meticulously refined through EQ adjustments. This was clearly no coincidence and appears to have been set based on measurements. There is a small ripple between 100 and 200 Hz, which seems to be a minor DSP tuning error. (I disassembled the unit with the manufacturer’s permission for measurement, and no anomalies were found in the individual drivers.)
---
Nearfield Measurement
The response has been properly "normalized."
Let’s compare it with the previous sample data.
The earlier tuning seemed to focus more on the response of individual drivers rather than the overall harmony of the complete speaker. In reality, the design must take into account not only the driver integration but also the acoustic interactions caused by the speaker’s physical design. Flat response in near-field measurements doesn’t directly translate to good real-world performance. However, it’s clear that these factors have been addressed in the improved version.
This improvement in the near-field response impacts not just the on-axis frequency response but also the speaker’s overall performance, including directivity and distortion. Let’s explore these further!
---
Directivity
From around 2kHz to 6kHz, the speaker maintains quite wide directivity. The overall high-frequency response is both broad and smooth, indicating notable improvement compared to the previous version.
Here’s a contour plot of the initial sample’s horizontal directivity. In the earlier version, there was a distinct bulge in the mid-high range, indicating an exaggerated narrow band response. This issue seems to have been corrected, and the mid-high frequency response, which affects imaging, has clearly been smoothed out.
---
Vertical Directivity
Comparing the vertical directivity with the earlier version reveals that null points have been reduced and the overall response is much smoother.
Beamwidth
The speaker maintains a dispersion angle of nearly 80 degrees in the mid-high frequencies. For a studio monitor, this is quite wide, which can result in a more colorful sound when reflections are involved at longer listening distances. While I personally prefer speakers with narrower directivity, this is ultimately a matter of taste.
As mentioned earlier, the vertical directivity is notably smoother compared to the initial sample, with a tolerance of +/- 20 degrees at the crossover point.
Polar Plot
This plot provides a more intuitive understanding of the speaker’s wide dispersion. Even at a 60-degree angle from the speaker’s front (0 degrees), there is only about a -3dB reduction up to 5kHz, which is impressively wide!
THD
The THD was already quite clean in the earlier sample, and the new sample delivers similarly clean performance. Let’s briefly look at the earlier data.
In the previous sample, there was noticeable air leakage around the rear amp section at the port’s resonant frequency of 50Hz, which resulted in significant HOHD (Higher Order Harmonic Distortion), as seen in the green graph. HOHD represents harmonic distortion components of 10th order and above. Such distortion can arise from various factors but often appears when there is extraneous noise that doesn’t directly contribute to the timbre.
In comparison, the updated version is significantly cleaner.
In relative terms, harmonic distortion across the frequency range is below 1% above 100Hz. However, the spike around 1.5kHz in the third harmonic distortion suggests the absence of a shorting ring in the woofer driver. I’ll cover the role of shorting rings in more detail in a separate post.
While the absence of a shorting ring can impact the overall performance during music playback with multiple frequencies, I don’t consider it a major flaw given the speaker’s price point. Of course, if this were happening in a speaker costing several million won, it would be quite disappointing (and it does happen occasionally).
The performance remains quite stable, even at 96dB SPL@1m.
Multitone Test
Multitone distortion results are on par with the initial sample, with no notable distortion spikes observed in the raw data plot, so I’ll leave it at that. This often boils down to the balance between [woofer driver quality vs. cost], a challenge that’s hard to address in lower-cost products.
Overall, the multitone distortion baseline isn’t super clean, but it remains stable without significant increases as the output level rises. In low-cost products, it’s common for the 96dB SPL distortion line to overlap or exceed the 76dB SPL line, but the MV6 performs quite stably.
Compression Test
Interestingly, compression is more pronounced in the midrange, while the performance in the low-mid and tweeter ranges is surprisingly good, which was a pleasant surprise.
Deviation between 2 samples
This is perhaps the most impressive aspect, alongside the improved on-axis frequency response. Like many budget speakers, this model uses a master-slave configuration with the main speaker and a satellite. Excluding the high frequencies above 10kHz (where measurement errors can exceed sample variance), the variance is within +/- 0.3dB across almost the entire range. While it’s not enough to draw definitive conclusions based on just one pair of samples, it’s still an excellent result for a budget speaker.
---
In summary, the MV6 presents a notably flatter tonal balance and more refined overall performance. However, its directivity control and high-output handling leave a bit to be desired. The wide range of input options, though, will appeal to casual users.
Finally, as mentioned at the beginning, this review was a particularly emotional experience for me. It’s rare for a manufacturer to take constructive criticism so positively and work on improving their product in response. Karak Electronics did not respond emotionally but instead focused on addressing key issues like frequency response tuning and rear amplifier leakage, leading to tangible improvements.
While the MV6 still isn’t on par with professional brands like Genelec or Neumann in terms of overall performance, and it may not rival the price-to-performance tuning of Kali Audio, I must applaud the sincere effort Karak Electronics has made to improve their product within the constraints they face.
