Vivid Kaya 45 loudspeaker Measurements

Sidebar 3: Measurements

Using DRA Labs' MLSSA system, I measured the Vivid Kaya 45's farfield behavior with a calibrated DPA 4006 microphone and its nearfield responses with an Earthworks QTC-40 mike. The Kaya 45's specified sensitivity is 87dB/2.83V/m. To my surprise, my estimated sensitivity was higher, at 90dB(B)/2.83V/m. (I say "surprise" because my experience is that measured sensitivities are often lower than the specified figures, but they're rarely significantly higher.) The Kaya 45's nominal impedance is specified as 6 ohms, with a minimum magnitude of 2.8 ohms. The solid trace in fig.1 shows that the impedance drops below 6 ohms throughout the upper bass and lower midrange, with a minimum value of 2.4 ohms at 122Hz. The electrical phase angle (dashed trace) is often high, and the combinations of –51° and 4.3 ohms at 30Hz and +41° and 3.65 ohms at 195Hz will be particularly demanding on amplifiers. The Kaya 45 should be partnered with amplifiers that are comfortable with 2 ohm loads.


Fig.1 Vivid Kaya 45, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).

The gracefully curved enclosure seemed lively when I rapped it with my knuckles. When I investigated the enclosure's vibrational behavior with a plastic-tape accelerometer, I found resonant modes at 375Hz, 414Hz, and 777Hz on the front baffle (fig.2) and the sidewalls. This behavior is not too high in level, however, and, in combination with the modes' relatively high Q (Quality Factor), might not lead to audible coloration.


Fig.2 Vivid Kaya 45, cumulative spectral-decay plot calculated from output of accelerometer fastened to center of front baffle midway between the midrange unit and the woofers (MLS driving voltage to speaker, 4V; measurement bandwidth, 2kHz).

The ports on the Vivid's sidewalls are tuned to a low 35Hz, this indicated by the fact that the impedance-magnitude plot has a saddle centered on that frequency. The blue trace in fig.3 shows the nearfield response of the two woofers, and the minimum-motion notch, which is the frequency at which the back pressure from the port resonance holds the cones stationary, occurs as expected at 35Hz. The nearfield response of the ports (green trace) peaks between 25Hz and 60Hz, and its upper-frequency rolloff is clean. The woofers cross over to the midrange unit (red trace) at the specified 300Hz with symmetrical steep slopes. The black trace below 300Hz in fig.3 shows the sum of the Kaya 45's nearfield woofer and port outputs, taking into account acoustic phase and the different distance of each radiator from a nominal farfield microphone position. The usual peak in the upper bass in the output, which will be due to the nearfield measurement technique, is absent. This suggests the Vivid's bass alignment is tuned for articulation rather than absolute extension, the speaker relying on the usual room gain at low frequencies to give sufficient bass weight.


Fig.3 Vivid Kaya 45, anechoic response on tweeter axis at 50", averaged across 30° horizontal window and corrected for microphone response (black), with the nearfield responses of the midrange unit (red), woofers (blue), ports (green), and their complex sum (black) respectively plotted below 500Hz, 800Hz, 500Hz, and 300Hz.

The Kaya 45's farfield response, averaged across a 30° horizontal window centered on the tweeter axis, is shown as the black trace above 300Hz in fig.3. The balance is superbly even up to the presence region, where there is an excess of energy reaching +3dB at 4kHz. (This might have affected my sensitivity estimate.) However, the plot of the Kaya 45's horizontal dispersion, referenced to the response on the tweeter axis (fig.4), indicates that there is a lack of energy in this same region to the speaker's sides. In medium-sized and large rooms, this will tend to balance the on-axis excess, though the speaker might sound a touch bright in small rooms. The contour lines in this graph are otherwise even throughout the midrange and treble, implying stable stereo imaging. The apparent off-axis peaks above 20kHz in this graph are actually due to the tweeter's output rolling off faster on-axis than it does to the sides. This is shown in fig.5, which shows the actual off-axis responses rather than normalizing them to the on-axis output.


Fig.4 Vivid Kaya 45, lateral response family at 50", normalized to response on tweeter axis, from back to front: differences in response 90–5° off axis, reference response, differences in response 5–90° off axis.


Fig.4 Vivid Kaya 45, lateral response family at 50", from back to front: responses 90–5° off axis, response on tweeter axis, responses 5–90° off axis.

In the vertical plane (fig.6), a suckout develops in the crossover region 15° above the tweeter axis. The tweeter is 39" from the floor, and this suckout won't affect the Vivid's tonal balance for listeners with their ears level with or just below the tweeter.


Fig.6 Vivid Kaya 45, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 15–5° above axis, reference response, differences in response 5–10° below axis.

In the time domain, the Kaya 45's step response on the tweeter axis (fig.7) indicates that the tweeter, midrange unit, and woofers are all connected in positive acoustic polarity. The decay of the tweeter's step, which arrives first at the microphone, blends smoothly with the start of the midrange unit's step. The slight discontinuity at 4.4ms suggests that the optimal blend between the midrange unit's output and that of the twin woofers occurs just below the tweeter axis.


Fig.7 Vivid Kaya 45, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).

The Vivid's cumulative spectral-decay plot (fig.8) is very clean overall. While two ridges of decayed energy can be seen in the mid-treble, which are presumably due to breakup modes in the midrange unit's metal cone, these are suppressed by the crossover. (As always with my CSD plots, ignore the small ridge just below 17kHz, which is due to interference from the computer monitor's line-scan frequency.)


Fig.8 Vivid Kaya 45, cumulative spectral-decay plot on tweeter axis at 50" (0.15ms risetime).

Overall, the Vivid Kaya 45 offers excellent measured performance.—John Atkinson

Vivid Audio
US distributor: Vivid Audio LLC
1815 SE 40th Street
Portland, OR 97214
(650) 996-2295

Ortofan's picture

... the similarly priced Revel Ultima2 Studios?
Do the Kayas capture the "flag" that the Revel Performa F228Bes "planted in the ground at the $10,000/pair meridian?"
Are the Kayas more of a "pure delight" than the Dutch & Dutch 8cs?
If the GoldenEar Triton One.R "does it all, and at a fair price", then what, if anything, does it lack?
To paraphrase the amp review, if one of these speakers is right, then the others must be wrong.

If the "the Kayas sounded good from the moment they hit the carpet" and "never failed to please, with any musical source at any reasonable listening level", then why is KR not buying them (instead of returning them and missing them)?

Kal Rubinson's picture

Upon re-reading this comment, I feel the need to say that I do not change my equipment willy-nilly even when greatly impressed with a new product. One must admit that there are a lot of great speakers in the same general price range, some already mentioned in this discussion, and one can make a good case for any one of them.

I rely a great deal on my intimate familiarity with a product, especially a loudspeaker, as it works in my system and my room and that takes a lot of time and effort. To change, one must begin again and that's after surmounting the necessary physical and financial costs. OTOH, it is always on my mind....................

Ortofan's picture

On this site, and elsewhere, it's not uncommon to read reviews that conclude with words to the effect that the reviewer especially enjoyed the sound quality of a product, hated to see it go and will really miss it.

The reader is often left wondering if the product was truly that good then why didn't the reviewer buy it?

Perhaps such reviews should conclude with some rationale for the reviewer's decision not to buy that particular product.

As you suggested, below, the outcome might conceivably have been different with the Kaya 90.

Kal Rubinson's picture

The reader is often left wondering if the product was truly that good then why didn't the reviewer buy it?

There are so many factors in such a decision, many of which are quite personal and not relevant to others. For example, color and style. Readers can judge these for themselves and, unless I can say something quite positive about the product in this regard, why introduce it? (Quality of finish and construction is a different issue.)
Another is spouse acceptance.

So what I try to do is compare and contrast the performance of the DUT with similar ones that I know and let the reader consider that in the context of his/her own personal preferences.

Bogolu Haranath's picture

May be KR could also review the new Polk Audio Legend L800 floor-standing speakers with SDA-PRO technology, $6,000/pair ........ L800 were favorably mentioned by RS in a recent dealer demo ...... L800 were also favorably reviewed by S&V magazine :-) ........

Kal Rubinson's picture

I prefer the overall performance of the Studio2s and, in particular, their bass weight and extension. It has bigger woofers. Perhaps the bigger Kayas would suit me more.

Anton's picture

Thanks for the added consideration!

Bogolu Haranath's picture

The top of the Kaya line Kaya 90 are priced at $26,000/pair ........ May be worth reviewing by Stereophile :-) .......

Bogolu Haranath's picture

The Revel Performa F328Be ($15,000/pair) may equal the bass extension and performance of Kaya 90, for less cost ...... F328Be are even less expensive than Kaya 45 :-) .........

Kal Rubinson's picture

I am well aware of that. :-)

AJ's picture

I find non normalized off axis data far easier to read/interpret, thanks.
Hopefully this is the new "norm" ;-)

Oilman's picture

A non-normalized off-axis view gets us closer to “spin-o-rama” information, which is well established to predict speaker audio quality. JA’s normalization of off-axis measurements are oft-criticized by Dr Floyd Toole. JA, please ditch the normalized curves once and for all and give us non-normalized!

Bogolu Haranath's picture

You may already know this ....... KR wrote an article about 'blind listening tests' at Harman facilities and about 'Spinorama' graphs :-) ........

Bogolu Haranath's picture

If you see Figures 4-6 in measurements section of Kaya 45, you can see a little bit of 'Fletcher-Munson' type of 'bowl' in the upper midrange and treble region (although, it is not that much pronounced) :-) ........

Bogolu Haranath's picture

Wanna see some 'ripples' this same region? ........ Take a look at any Wilson speaker measurements (reviewed by Stereophile) in similar section of measurements :-) .......

Bogolu Haranath's picture

Lot of B&W speakers also show 'bowls and ripples' in their measurements in this area :-) .......

John Atkinson's picture
Oilman wrote:
JA, please ditch the normalized curves once and for all and give us non-normalized!

Space in the print magazine is limited, so I decided many years ago that the primary dispersion measurements I would publish would be normalized to the tweeter (or other recommended axis) response. It was the changes in the response off-axis that I felt more revealing. Only occasionally, as the case of the Vivid Kaya 45, have I also published the actual responses.

Perhaps for the website reprints of the magazine's loudspeaker reviews I could publish both normalized and actual dispersion graphs.

John Atkinson
Technical Editor, Stereophile

AJ's picture

It was the changes in the response off-axis that I felt more revealing.

This is generally true, since spectrally *dissimilar* reflections are perceptually harder to ignore, which normalized will emphasize.
But in the case of the subsequent Magico review, the normalized makes the >10k response look very wide, which in fact it is not, since it is actually correspondingly down in level with the on axis. I can see it this way being used to viewing normalized, but the casual reader eye might mistake it for otherwise.
I'm with Dr Toole on this one. The "raw" response impinging the surroundings is more informative, IMHO.

Soundfield Audio

Bogolu Haranath's picture

If, you look at the Audio Physic Virgo III Stereophile measurements, the JA1's listening room FR measurements (Fig.7) look very much similar to Magico M2 :-) .........

AJ's picture

The active Dutch & Dutch 8cs must be placed close to the front wall

Hmmm, a cardioid/controlled directivity type speaker *can* be placed closer to the front wall, less detrimentally than a non-cardioid, due to the rear output nulling and controlled off axis.
However, that shouldn't be mandatory, especially with an adjustable (such as an active) speaker.
With proper adjustment, should function just fine away from wall. Subsequent spatial rendering should be commensurate.


Soundfield Audio

Kal Rubinson's picture

True but then their appearance would offend my wife even more. I told Martijn Mensink that when he last visited me.

AJ's picture

Aha, acoustics puzzle solved. Occams razor.

Anton's picture

They look like the Shmoo from L'il Abner.

Kal Rubinson's picture

Yeah. I did say that, didn't I?

Bogolu Haranath's picture

Kaya 25 kinda looks like Jar Jar Binks :-) ........

Bogolu Haranath's picture

Kaya 25 pair could work well with a couple of powered subwoofers :-) ........

MikeP's picture

They just come out with the much smaller and even better $6K Vivid Audio Kaya S12's !