What is even more important to me, however, is that it also produces major sonic benefits even in a good room and a good location. The bass is much tighter, and transients are far better defined. You can hear the full range of bass without dominant peaks and fewer apparent suck-outs. Higher-level dynamics are cleaner, particularly in the bass. The Aeris does not have all of the power and bass detail of the Legacy V, but it can overdrive my room at every bass frequency that is musically relevant. Adding the Bohmer room correction means that the overall sound is much cleaner at higher volumes. There are fewer room-boundary problems, where higher bass levels mask the rest of the music to some degree or are too sustained to sound realistic. Room correction not only provides great bass detail, it does so more evenly.
The critical transition from the deep bass to the midbass is cleaner and more musically natural, as is the transition from upper bass to the lower midrange. This allows the Aeris to do a better job of cleanly reproducing the natural warmth of music that is present in good recordings and doing so more accurately. The middle and upper midrange and the treble become clearer when the hills and valleys in the bass response, and excess room resonances, are reduced. This is something I’ve also noted in really good speakers without room correction and that measure exceptionally smoothly in the bass in a given listening room. Getting the bass right is critical to getting the best in midrange and treble sound.
Soundstage detail and depth become cleaner and more detailed, and imaging becomes notably more precise and natural in many recordings. The Aeris’ soundstage is very good even without room correction, but the speaker seems to act more like a point source with room correction engaged.
The combination of the Aeris and Wavelet provide some of the most musically realistic sound I’ve ever encountered. They take digital processing and room correction a vital step forward, and show they can reach levels that are competitive with even the best purist speakers.
The ability to make firmware upgrades, as the interview attached to this review with Bill Dudleston (the chief designer of the Aeris) indicates, will lead Legacy and Bohmer to make steady improvements in processing, operating, and set-up features, and don’t forget, as you look at the price, the Wavelet is also a really good analog and digital preamp and DAC as well.
If I now have a new wish, it’s to hear what the Bohmer level of correction can do when applied to other brands of speakers. As the interview indicates, this is another wish that may end up being granted.
Specs & Pricing
Type: Frequency- and time-domain-optimized 4.5-way loudspeaker with directivity controlled array
Tweeter: 1" AMT neodymium ribbon
Upper midrange: 4" AMT neodymium ribbon
Midrange: 8" cast-frame, titanium-encrusted diaphragm, dipolar configuration (open baffle)
Midbass: 10" cast-frame, carbon-fiber/pulp diaphragm, dipolar configuration (open baffle)
Bass: Dual 12" aluminum diaphragm, Aura neodymium motor, sealed enclosure
Frequency response: 18Hz-30kHz +/-2dB
Impedance: 4 ohms
Sensitivity: 95.4 dB @ 2.83 volts1m in-room
Recommended amplification: Bass section is powered internally with dual 500-watt ICEpower amplifiers; 30 watts or greater required for upper section
Crossover: 80Hz, 2.8kHz, 8kHz
Inputs: 1 pair of external binding posts, 1 XLR balanced inputs
Dimensions: 14.5" x 58" x 16"
Weight: 171 lbs.
AHC Talks with Legacy’s Bill Dudleston
Let’s talk about the future of both your efforts in room correction and plans for the Wavelet. Any plans for a universal version of the Wavelet that could handle any speaker, including ones with a single input?
Yes. An example is already in the works. We are introducing a 750-watt flex-powered version of our Focus speaker, which can be driven three different ways; mono-amplified internally with a single input, bi-amplified with internal crossover, and bi-amplified with the Wavelet crossover. In all three variations the Focus XD can employ the room correction. To correct a generic speaker with a single input, we will offer a basic menu of target function choices, including excursion limit protection at low frequencies, and request the user input the best fit of the speaker’s radiation pattern (e.g., omni full-range, omni bass with cardioid upper range, and dipolar). Another parameter to be input will involve the number of subwoofers in use if any.
If we can switch back to the Legacy Aeris and Wavelet, what adaptations from the room correction for the Legacy V did you have to make to suit the Aeris?
The room correction process remains the same with a very similar target function as the V. However, building the V system from the ground up using the Wavelet revealed several areas where we could get more performance from the Aeris in the crossover region and in driver correction. Aeris is more coherent with the Wavelet in place before the room correction is even applied.
How is your approach to room compensation evolving? Have there been changes since the Legacy V review and what changes are you exploring? Will they all be possible through software changes?
Most of the recent improvements to the software have been made to improve setup and user control, such as the polarity check and level adjustments. We are now looking at the upper range of the reverberant field more closely. Here the density of reflection, spectral balance, and relevant temporal information are being studied along with the psychoacoustic weighting of this information. All improvements will be available through software updates downloaded to the USB stick and inserting in the Wavelet port.
The remote control features and software download commands are now accessed via Wi-Fi. Do you have plans for a wired network connection to realize software updates?
The prototype Wavelet originally hosted its own network, but this prevented the continual improvement to the remote interface and functions. The current method provides many advantages include control of multiple units simultaneously. Software updates will eventually be offered from the Legacy website, regardless of what port is used. Users will be allowed to subscribe to updates for a modest fee to keep the programming current.
Any potential for Wavelet to adopt MQA or higher sampling rates?
That really is a DSP question. The Wavelet will presently accept PCM files rates at high as 352.8kHz and higher. But remember Wavelet is not just converting a single sample of data per unit time, but correcting a 40msec window. This is equivalent to 14,080 samples to applying complex computations upon in real time.
A sampling rate of 96kHz is more efficient, consuming less processing power and sonically equivalent in the end. Think of a digital photo. A sharply focused image at 150 dpi will provide more real detail than a slightly out of focus image at 300 or 600 dpi. That is why it is the role of the Wavelet to sharpen focus in the time domain at 56 bits and then use apodizing to remove digital artifacts.
As you know I am a strong believer in the workings of Bob Stuart’s MQA. I think if audiophiles experienced it, even if they didn’t comprehend the genius of the solution it offers, they would realize that sonic improvements are not to be had by merely increasing sample rates. I hope the press gives MQA the attention it deserves. The consumer must demand it for the format to gain acceptance.
We would then most certainly consider a license. While I will personally always record live in WAV, MQA is the best solution yet proposed to deliver music to the audiophile. No compromise in dynamics, noise floor, or audible bandwidth, yet file size is similar to a 16-bit WAV file. It is certainly a great replacement for FLAC, DXD, DSD, SACD. Today, music should not even be distributed in MP3, AAC, WMA. These formats should be used in talking appliances.
Your literature describes the way the Wavelet reduces room reflections, but does not describe frequency correction. How does the Wavelet do this?
First the loaded algorithm corrects the loudspeaker frequency anomalies in each channel of output, independent of the room. Let’s say we have a shallow dip at 1800Hz, for example, but next to it is a sharper rise in response at 2200Hz. Previous methods would apply a broad boost at 1800Hz, and a sharp cut at 2200Hz using filters that introduce phase shift. While this can make the frequency response appear smoother at a single mike position, the ear is aware something is still wrong in the time domain and the power response. Time domain measurements substantiate this.
The Wavelet applies a totally different approach. First of all, it will not force the frequency response flat at the expense of transient behavior. It will address the cause of the problem and make a psychoacoustic correction weighing the time domain heavily. The Wavelet’s software will identify domain errors introduced by stored energy in the diaphragms, which are ultimately the actual cause of the peaks and dips. It will undergo iterative calculations to determine the most optimal solution with regard to phase to preserve transient response. It may apply a gentle lift if energy is lacking, or remove a resonance in the diaphragm material but the time domain will always be improved in the process.
The room correction continues in the same manner, with emphasis on treating errors introduced by boundary interaction. It does not merely notch response due to room resonances. It instead works to prevent these resonances from forming by looking for late arrival of redundant information in the measurement process. Resonances take quite a while to build up. Even a simple floor to ceiling axial resonances requires at least 16msec to form. The problematic buildup from the wall behind the speaker usually requires less than 8msec in comparison.
This old information is predicatively and literally fed forward in time, canceling its own presence. It is not accomplished relative to a position or multiple averaged positions in the room but relative to the launch from the speaker itself. It is unique in this manner. The process really should be described as automated loudspeaker adaptation instead of room correction. We didn’t change the room a bit!