Audio Physic Step Plus loudspeaker Measurements

Sidebar 3: Measurements

I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the Audio Physic Step Plus's frequency response in the farfield, and an Earthworks QTC-40 for the nearfield responses. (I didn't use the grilles.) My estimate of the Audio Physic's voltage sensitivity was 84.2dB/2.83V/m, significantly lower than the specified 87dB/2.83V/m. While the specified nominal impedance is 8 ohms, the solid trace in fig.1 shows that the impedance magnitude drops slightly below 8 ohms in the midrange and to 6 ohms at 20kHz. The electrical phase angle (fig.1, dashed trace) never gets extreme, meaning that this speaker is easy for an amplifier to drive.

119APStepfig1.jpg

Fig.1 Audio Physic Step Plus, electrical impedance (solid) and phase (dashed) (2 ohms/vertical div.).

Discontinuities in the fig.1 traces at 420 and 650Hz imply the presence of cabinet-wall resonances at those frequencies. I did find a mode quite high in level at 650Hz on the sidewalls (fig.2), and another at 420Hz on the top panel. These modes are high enough in Q and frequency that they might not be excited by music.

119APStepfig2.jpg

Fig.2 Audio Physic Step Plus, cumulative spectral-decay plot calculated from output of accelerometer fastened to center of sidewall (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz).

The saddle centered on 50Hz in the impedance-magnitude trace in fig.1 suggests that this is the tuning frequency of the reflex port on the rear panel. However, as shown by the blue trace in fig.3, the woofer's minimum-motion notch, which is when the cone is held stationary by the back pressure from the port resonance, occurs a little lower in frequency, at 44Hz. The port's output (red trace) peaks a little more broadly than is usually the case, and its high-frequency rolloff is disturbed by low-level peaks at the same frequencies as the panel resonances noted earlier.

119APStepfig3.jpg

Fig.3 Audio Physic Step Plus, anechoic response averaged across 30° horizontal window on tweeter axis at 50" (black), averaged across 30° horizontal window centered 10° above tweeter axis (green), both corrected for microphone response, with nearfield responses of port (red), woofer (blue), and their complex sum, respectively plotted below 925, 310, 310Hz.

The complex sum of the nearfield woofer and port responses is shown as the black trace below 300Hz in fig.3. The usual nearfield bump in output in the upper bass is minimal, which, in conjunction with the depressed level of the farfield response (black trace above 300Hz), suggests that the Step Plus is intended to be used close to the wall behind it, in order for its balance to gain the benefit of some boundary reinforcement. However, as Ken Micallef noted, this will diminish the speakers' imaging performance. I must admit to some puzzlement at KM's finding that the Step Plus offered "first-rate bass-frequency reproduction," as the port tuning suggests that it won't be able to fully deliver the lowest note of the four-string double bass and bass guitar at the correct balance with the upper harmonics.

Of greater concern in fig.3 is the major suckout in the crossover region in the farfield response on the tweeter axis (black trace). This suckout negatively affected my sensitivity estimate and will make the speaker sound lacking in life if the ear interprets the midrange level as being correct, or midrange-forward if the presence region is taken as the reference. This will depend on the music being played. This suckout does tend to fill in to the speaker's sides (fig.4)—I note that KM ended up with the Steps wide apart and with their sidewalls barely visible—but, more important, it disappears when a listener sits with his ears well above the tweeter, as can be seen both from the graph of the speaker's vertical dispersion (fig.5) and from the green trace in fig.3. If the Steps are placed on stands low enough that the tweeter is around 8" below the listener's ears, the treble balance will then be even and flat. But with the speakers on stands that place the tweeters higher than the listener's ears, the suckout will be even deeper.

119APStepfig4.jpg

Fig.4 Audio Physic Step Plus, 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.

119APStepfig5.jpg

Fig.5 Audio Physic Step Plus, vertical response family at 50", normalized to response on tweeter axis, from back to front: differences in response 45–5° above axis, reference response, differences in response 5–45° below axis.

In the time domain, the Audio Physic's step response on the tweeter axis (fig.6) indicates the cause of this suckout: the drive-units are out of phase in the crossover region. While both drive-units are connected in positive acoustic polarity and, as usual, the tweeter's output arrives at the microphone before the woofer's, the decay of the tweeter's step overlaps the start of the woofer's step but is in opposite polarity. This could have been solved by inverting the polarity either of the tweeter or of the woofer, so that one unit's output smoothly blended with the other's. Alternatively, by moving the microphone significantly above the tweeter axis, the woofer's output is pushed back in time until its step smoothly blends with the decay of the tweeter's step. As a result, the outputs of the two drivers, instead of canceling, now correctly sum in the crossover region—again, compare the green with the black trace in fig.3. The cumulative spectral-decay plot on the tweeter axis (fig.7) is pretty clean, however.

119APStepfig6.jpg

Fig.6 Audio Physic Step Plus, step response on tweeter axis at 50" (5ms time window, 30kHz bandwidth).

119APStepfig7.jpg

Fig.7 Audio Physic Step Plus, cumulative spectral-decay plot on tweeter axis at 50" (0.15ms risetime).

I note that KM enjoyed his time with the Audio Physic Step Plus. However, its measured performance suggests it must be used on a low stand. I will be investigating this aspect of the Step Plus's behavior in a future issue.—John Atkinson

COMPANY INFO
Audio Physic GmbH
US distributor: VANA Ltd.
2845 Middle Country Road
Lake Grove, NY 11755
(631) 246-4412
ARTICLE CONTENTS

COMMENTS
tonykaz's picture

are the foundation of the Audiophile existence because of their superb Driver technology, design and build quality.

A nice representative Small European Loudspeaker will cost out around $2,500. Even the phenomenal Genelec 8020 Active Loudspeakers, made in Europe, will cost under $2,500!

But Audio Physic pretends to be German while sourcing in China.

Too bad for the replacement driver assurance program.

Yet another disappointment.

Tony in Michigan

georgehifi's picture

All Audio Physic's even going way back when they started that I've heard, do this disappearing act really well, whether stand or floor mount. What is the measurement or construction parameter that is important that makes this happen. Seems like Audio Physic has whatever it is nailed down.

Cheers George

ok's picture

..Audio Physic (mostly Step and Tempo 25/Plus) iterations of this particular midrange/tweeter/cabinet configuration over the years and I can subjectively confirm JA’s prediction that their top surface should see slightly below the listener’s ears for optimal HF performance. This peculiarity is mainly due to a somewhat upward-shooting tweeter combined with a deliberately poor off-axis response (no more than 30 degrees flat thanks to cone driver and steep foam ring) that is further highlighted by the grills which tend to create some 3db drop at about 8KHz. They are excellent for near field and almost immune to HF room reflections for far field auditioning; however they do need considerable rear estate for enhanced soundstage accuracy and deep bass response as KM has already pointed out. By the way woofer's aluminum phase plug can dissipate an awfully lot of heat in case of amplifier clipping (not recommended!) so no worries about chinese: my neighbours would dismayingly attest that these drivers are virtually indestructible :-}

Bogolu Haranath's picture

Elac Navis active, self-powered bookshelf speakers ($2,000/pair) were very favorably reviewed in the latest issue of TAS ....... They don't need a power-amp ....... The reviewer said that, he would recommend them for product of the year award :-) .........

doak's picture

Haven’t heard these speakers, but really want to... and possibly own a pair.
Thanks for the heads up on the TAS review.

Bogolu Haranath's picture

In the on-line Audiogon discussion forum, one of the commentators says he compared the sound of the Elac Navis bookshelf with the sound of KEF LS-50 wireless and, he preferred the sound of Navis bookshelf (both models are active self-powered speakers) :-) .........

AaronGarrett's picture

The Audio Physic/ Pass Aleph combo at Singer in 1996 made me an audiophile! EDIT (Actually I misremembered -- it was Stereo Exchange. I am now an old audiophile with a faulty memory!)

Bogolu Haranath's picture

Welcome to the 'audiophiles with faulty memory' club :-) ..........

jimtavegia's picture

I would expect better performance in terms of pure value. Andrew Jones and many others do a great job for under a grand. In the right room these might float someone's boat.

I don't mean to be critical, but at this price there must be excellent performance.

mumulaha's picture

I knew the measurement have it's flaw before purchase. But after I heard it in my room with correct placement, I am sold. 6moon's Srajan Ebaen bought Audio Physic Codex as his new reference speaker. I used to own Triangle Titus with same electronic.

Jason P Jackson's picture

Wavecor are a large, well known loudspeaker driver manufacturer with many of it's designs done by the engineers of another large, well known loudspeaker driver manufacturer.

Jason. P. Jackson

SpeakerScott's picture

Mr. Atkinson,

Would you ever consider adding baffle step correction to your nearfield measurements...or at the very least measure speakers using the ground plane method to eliminate this measurement error?

I understand that taking some of the massive mega-buck speakers to a parking lot or tennis court isn't possible, but there are several ways to mostly eliminate the emphasis shown in the measurements with modified technique.

Scott

John Atkinson's picture
SpeakerScott wrote:
Would you ever consider adding baffle step correction to your nearfield measurements...or at the very least measure speakers using the ground plane method to eliminate this measurement error?

Some other reviewers have done this, but the "correction" is arbitrary, given that every room will modify the "bass bump" to a different extent.

There are 2 ways of assessing a speaker's low-frequency output: 1) in an anechoic chamber, which represnts one extreme, and 2) using a nearfield measurement, which represents the other extreme. As using a true anechoic chamber is financially out of reach for an organization of Stereophile's size, I prefer to stick with the other extreme. This also has the advantage of being consistent for the past 30 years of speaker measurements published in Stereophile

The problem with ground-plane measurements is that you need a relatively large flat paved area in which to perform them, which is problematic given where I live. I have experimented, but ended up sticking with how I currently perform low-frequency speaker measurements, in the nearfield.

John Atkinson
Editor, Stereophile

SpeakerScott's picture

An anechoic chamber is well out of reach for me to! (Plus ones that get that low in frequency are rare.) I could only dream of having one, which is why I suggested the baffle step correction method. As you well know, when you measure nearfield you eliminate the reinforcement (or lack thereof depending on frequency) as the driver transitions from 4pi to 2pi space.

The process, if done correctly isn't arbitrary. I have compared the results for large towers, book shelves and even large pro-sound PA speakers to measurements I have taken using tennis courts or parking lots. In the case of the ground plane measurements I have actually taken the time to reverse the double height baffle step from the mirror image speaker and the results have been spectacularly accurate.

You could plot the more accurate method along with the old method for comparison It only takes a few minutes to do, it doesn't add significantly to the process of your entire measurement set but improves accuracy.

(I do know how long a full measurement set takes to make...I've done it more times than I can count.)

dial's picture

All these little boxes tend to sound extremely similar and it's difficult to say that really one is better than others, I even have one pair I build myself and compared to Triangle Titus, they were poorer in the bass but a little clearer in the midrange. But I liked both. And they cost 10% of these 'Audio Physic' !

MiklD's picture

I auditioned Step 25 (the previous iteration without the ceramic foam and sundry tweaks) and was similarly impressed. Took home Sitara 25 (floorstanding 2.5-way version with an additional HHCM providing sub-500 Hz reinforcement) which provides a good half-octave of bass extension, while maintaining the sonic virtues of the bookshelf sibling. So this review provides some gratifying insight.

Interesting to see the wrinkles in the measured performance. Hard to put floorstanders on shorter stands though, maybe a raised platform for the listening chairs, and/or some DSP? I suppose to incorporate soundstage in the more objective realm some sort of standardised routine around one of those Chesky clap and footstep test recordings could be employed? That aside the range of individualised timbres delivered from diffent instruments (real and synthetic) is a complementary element to the legendary spatial qualities and ‘disappearing act’ and just as impressive.

I started the Sitaras with my venerable Krell KAV-300i and was very happy. I described the baby Krell’s untimely end in a comment pleading for Herb’s promised Micromega M-100 review (cat awoke from Class A/B warmed torpor and threw up into the venilation slots, no more input boards from Krell for that model, alas, and shameless pitch for Sydney’s Len Wallis Audio who tried valiently to find parts). A loaner Musical Fidelity M3i wasn’t bad at all, but I settled on the Micromega for more modern features and design and nice sound. Too nice though (soundstage didn’t quite open up and the bass lacked weight/authority/timbral richness and yes that may not be a real word) but remedied for now by running the M-100’s balanced pre-ouput to a Krell KAV-2250 from Ebay. Cobbling together satisfying sound on a tight budget has its moments, I guess (bid unseccessfuly for an Evo, them’s the breaks). With the right amp, the bass sounds deeper than it measures, so I’m not too surprised by Ken’s observations.

So no, as most readers here surely know (but not all commenters, it seems) speakers don‘t sound the same, amps don’t sound the same, and viva la difference (or the German equivalent for the Audio Physics).

Thanks also for the Irrisari, a nice discovery.

MiklD's picture

... and imaging. Looking at the Step’s cumulative spectral decay plot, it is unusually clean and fast to decay below 10 kHz (it’s less clean above) even compared to a range of very good-sounding offerings reviewed here. Everything is pretty much done and dusted within a millisecond, while others take 2-3 times that. Wondering if this characteric is a key reason for the soundstage, imaging and tonal differentiation I can also hear in the similar Sitara (and can our ears and brains resolve that difference)?

X