Bidjan Tschaitschian, The FIR-BoxX Project

Published: 19/07/2026

Author: Karsten Hein

Category: Explorations

Tag(s): Active HiFi Systems

Go to comments

The Hamburg-based IT expert Bidjan Tschaitschian uses digital Finite Impulse Response (FIR) technology, based on the “FIR-BoxX” multi-channel amplifier developed by Andreas Heggendorn in Switzerland, to ensure ultra-linear and phase-correct performance from each loudspeaker driver and improve crossover management. The eight-channel FIR-BoxX facilitates phase-, time- and amplitude correction, as well as room adaptation, and has the potential to bring personalised, high-end listening experiences into ordinary homes. — Does that sound complex?

From Vibration to Sound

Let's start from the beginning. Sound waves originate from physical vibrations that set the surrounding air in motion. Such vibrations are produced by our vocal cords when we speak, sing, or shout. In this process, air is pressed through them from our lungs, travels upwards through the vocal apparatus in the throat and out of the mouth into the open space. The resulting sound is influenced by each body’s unique physique. Some people sound loud, coarse, or boisterous, while others sound soft, quiet, unenergetic, or timid.

Similarly, musical instruments produce sound waves originating from vibrating reeds, strings and drum skins, or from the metal of triangles or gongs. Much like human vocals, these vibrations are shaped and amplified by the unique form, resonating chamber, and size of each instrument. The materials used also play a role, as they contribute their own resonance frequencies to the resulting sound. This is why each instrument has its own distinctive character and is valued more or less highly by the musician playing it.

Sound waves leave each resonating body in a specific pattern, providing the listener with clues about the type of instrument or person, about their playing or speaking style, and their distance and location in relation to the listener. If there is more than one active instrument or person present, our ears try to determine the distances between them and create a mental map of the scene or stage. Phase and time coherence are especially important in this context, as they enable us to estimate location. Higher frequencies are more revealing and easier to trace than bass frequencies.

Every vibration has a defined beginning, duration, and end. When time is factored in, a specific vibration creates a four-dimensional energy event traveling through space. Instrument builders look for pleasant harmonics and try to eliminate undesirable resonances from the instrument's output. Human beings also place an emphasis on sounding pleasant for the sake of social acceptance. Consequently, we all grow up in a world in which each utterance has purpose and design, and each sound has coherence. Any deviation from this is intentional and carries meaning. For example: Dissonance often signals distress. And a sudden compression of frequencies indicates that an object is moving towards us at high speed.

Recording: From Large to Small

Although instruments can range in size from tiny and compact to large and bulky, it has been found that faithful recordings of their acoustic output can be made using a single microphone diaphragm. In such a setup, the diaphragm is much smaller than the instrument and made from materials that mostly remain neutral across the musical spectrum in terms of their own resonance. The microphone's role is to capture the acoustic event without introducing too many of its own characteristics. While such microphones can be expensive, high-quality recordings are possible across a wide frequency spectrum once a good mic has been found. At this stage, the original phase-time relations are captured by a single diaphragm and remain intact, much as they do when frequencies hit the human ear.

While a small microphone diaphragm is sufficient to capture the vibrations of large instruments and life-size vocalists, playback tools in the studio or our living room need to again be large and life-like to recreate an authentic impression of the event. The air needs to vibrate in the same way as it did on the day of the recording, even though the original instruments and vocalists are no longer around. In other words, there are no metals, woods, strings or drums present. And it is this large amount of air movement that leads to most of the challenges of loudspeaker building.

Playback: From Small to Large

High-frequencies require a rigid but ultra-light diaphragm to facilitate nimble acceleration. Midrange frequencies need to be especially true to tone, because humans are ultra-sensitive to tonal shifts in the frequency spectrum of the human voice; and bass frequencies need to be powerful and at the same time layered and nuanced. While the transducers should do all this, they must not add any resonance frequencies of their own. The traditional approach has therefore been to work with two or three dedicated drivers, with each one perfectly designed to do its job within a narrow frequency range.

Recreating the original sound via separate drivers poses some challenges, of course. Because the materials, diaphragm masses, and projection planes are not the same between them. Each driver has its own characteristics in terms of acceleration and lag, material resonances, electrical and physical resistance, power sensitivity, diaphragm travel, etc. No matter how well the characteristics are adjusted and matched via custom-built crossovers, the resulting sonic image is never fully intact. While phase and time shifts also occur in natural instruments and in the recording room, the inaccuracies produced during playback contain no useful information, are never desirable, and always take away from the integrity of the overall experience.

Single vs. Multiple Drivers

Manufacturers have dealt with this issue in different ways: Saul Marantz’s Dahlquist DQ10 loudspeakers, for example, are meticulously matched 5-way speaker phase-arrays. In his design, Marantz brought all five drivers onto a plane so that time-smear between them would be reduced to a minimum. This design worked so well that the DQ10 are considered exceptionally musical performers to this day. Marantz’s idea stemmed from an earlier design by the Quad company from England, who had designed a loudspeaker of similar dimensions to the Dahlquist that was based on a single diaphragm. The Quad ESL 57 is famous for producing phase-correct music that surpassed most of the loudspeakers of its time. Saul Marantz wanted to marry the bass performance of an electrostatic speaker with the output volume of a dynamic system.

My Martin Logan SL-3 loudspeakers aim for a similar result by using an electrostatic ‘Mylar’ diaphragm for the treble and midrange and pairing this with a dynamic bass driver for more natural bass weight and dynamics. For electrostatic speakers to cover the full frequency spectrum, however, they would need to be very large. The SL-3’s dynamic woofer only comes in at frequencies where the human ear is no longer as sensitive, which makes the dual driver setup more agreeable. However, electrostatic speakers are bi-polar by nature and require careful placement, lots of space behind them, and plenty of experience in setting up systems to sound at their best. Magneto-static speakers, from Magnepan and others, offer similar benefits and are easier to set up. They are also better suited for smaller rooms. All of the mentioned designs are unique approaches to solving the single driver issue. However, the most successful applications have remained small in number, because they require an excellent music source, thoughtful cabling, a powerful amplifier, and a treated room.

In my experience, far fewer people are willing to invest the money, time, and energy to learn about properly setting up a High End system (along with a properly treated room) than are interested in the experience of listening to and owning one. In fact, so much can go wrong in the setting up of a HiFi-system that I devoted some 30 chapters to this subject, all featured in the “High Fidelity” section of this blog. While it cannot hurt to have in depth knowledge and to lay down the basics for your HiFi or High End system, FIR technology now offers all-new possibilities that provide a similar effect at lower cost and with much less fuss to those who only want to enjoy the pleasure with less of the pain.

Maximising Driver Control

When setting out on your FIR adventure, it is important to understand that each of the amplifiers inside the FIR BoxX needs to be directly connected to one of the drivers in the loudspeaker. Usually, this means that the existing loudspeaker crossover is removed or bypassed and a new multi-channel terminal is installed at the back of the loudspeaker. Speakon connectors—originally developed by Neutrik for professional audio applications—are useful in this context, as their 8-pole version is able to run four amplifier channels through a single plug.

The eight amplifiers inside the FIR BoxX mean that one such unit is able to power one pair of 4-way speakers. The Dahlquist DQ10’s unusual 5-way design would mean that two FIR BoxX amplifiers are necessary (perhaps not yet feasible in clock-perfect fashion at this time), whereas a 3-way speaker would have two channels free for the addition of subwoofers, if these are desired. Each FIR BoxX amplifier produces up to 30 watts RMS of Class-D amplification, and—while this may not sound like lots of power—the effective output cannot be compared to that of a classic amplifier from looking at this number alone. Here is why:

The classic amplifier sends all frequencies to the loudspeaker where passive crossover components divide the signal using capacitors, coils, and resistors. This means, the amplifier sees a relatively complex load, and the passive components take away from its power output. In the FIR BoxX, on the other hand, the signal is treated, before it reaches the amplifier, so that each amplifier only needs to process the signals intended for the one driver it is directly connected to. Of course, in the case of my 2-way Martin Logan speakers, considering the complex load of an electrostatic panel, 30 watts might still be insufficient, but then electrostatic speakers already operate in a phase-coherent fashion and are not the primary focus of FIR technology.

The FIR benefit is much greater for loudspeakers with multiple dynamic drivers. During the setup (that currently still needs to be performed by someone with experience in this field, such as the Hamburg based FIR-developer Bidjan Tschaitschian) the output of each loudspeaker driver is measured, linearised on its own and then harmonised with the others. The prompts might be: “Send frequencies above 800 Hertz to the midrange.” “Reduce a particular resonance by 4 dB.” “Delay the tweeter by 0.17 ms.” “Change the phase relationship between the drivers.” “Compensate for the specific driver enclosure.” Or: “Apply a specific equalisation curve.” — In doing so, the FIR BoxX is able to align the acoustic centres of the drivers and to compensate for electrical and mechanical inaccuracies in the original driver and loudspeaker design.

As I learnt, there are often surprises during these early measurements. It is not rare, for instance, for one of the driver’s polarity to have been inverted during the original loudspeaker assembly. Once all drivers have been measured and set individually and as an ensemble, a final measurement is conducted of the complete stereo setup, in which the loudspeaker response is once again calibrated to suit the specific characteristics of the listening room when heard from the position of the listener. I was a witness to this type of setup in my audiophile companion Henrik’s listening room. His room had already been acoustically treated, we had set up the speakers according to the room’s specific resonance profile, and we had previously optimised the coupling of his speakers towards the floor. Given these preparations, Henrik’s passive setup already sounded great.

Henrik’s Dieter Achenbach Trinity speakers have power-sensitive chassis of high quality, and the original crossover was well-conceived using high-quality components. Looking over Bidjan Tschaitschian’s shoulder during the setup, I could see the driver resonances and the peaks and their output curves sloping. Bidjan carefully adjusted the amplifier’s output to flatten the curve, a process he diligently performed for each driver and then for the drivers in combination. I took some photos of the process and had to leave before the setup was completed and only came back for listening some days later when everything had already been put in place.

The FIR BoxX can handle audio signals from both analogue and digital sources, but it does not offer RIAA correction for phono applications. There is a total of four audio inputs available: 2x line stereo and 2x optical. (True for the tested FIR BoxX DIY. Some older versions have only one of each.) Henrik’s analogue frontend consisted of a massive Transrotor Zet3 turntable that was connected to a Phasemation t550 step-up transformer. RIAA correction was via a Symphonic Line “Die Erleuchtung” preamplifier. From there the signal could be routed either to the FIR-BoxX for digital processing or take the classic analogue route to a Musical Fidelity M6 PRX power amplifier. Henrik had left the Achenbach’s original crossovers intact and only bypassed them for FIR activation. This way, A/B comparisons between both technologies were still possible.

Henrik’s digital frontend consisted of an affordable FIIO SR11 digital transport that fed its data directly into the FIR-BoxX using the optical input. At the time, Henrik did not have a digital frontend for the analogue amplifier. Therefore, the only immediate comparison we could make was of the turntable signal routed to the analogue Musical Fidelity M6 PRX power amplifier and, alternatively, to the internal DAC of the FIR-BoxX using the analogue input. The next step will be to upgrade the digital frontend to include a CD-player, a streamer, and a DAC, which will make digital comparisons from multiple sources possible. However, this will have to wait for the next article.

From our time spent listening to the FIR-BoxX playing music from both analogue and digital sources, I can say that it does not draw much attention to itself as an electronic device and instead sets the stage for the music to unfold. This is what all great high-end systems should do, but the FIR-Box has the additional advantage that it is not so much dependent on perfect speakers and surroundings. It can compensate for imperfections where necessary. Henrik’s Achenbach loudspeakers are highly responsive but do require some control from the amplifier due to their large woofers. The FIR-BoxX rose to the challenge and did not show any audible weakness in terms of bass performance. In fact, it so much maintained the musical integrity of each recording that discussions of individual aspects seem a bit misplaced.

The music output is comparable to my electrostatic speakers and does sound somehow more solid and PA-like on the Achenbach due to their directional nature. Where the Martin Logan electrostats create an intimate small club event, in which even the tiniest of sounds are give some attention, the Achenbach Trinity tell the story of a large stage event with more drive and ferocity. Both playing styles have their charm. If you have read my reports on the inflationary use of DSPs at trade fairs and the degrading effects this has on music, I can assure you that the FIR-Box is a completely different beast. Nothing about its rendering of music seems dead, stale, or contrived. It can very well keep up with other High-End setups in terms of agility, tonality, and soundstage. Each sound has its space and dimension, bass sounds as natural as it should.

If I had to define a bottle neck, it would be the FIR-BoxX’s internal analogue to digital converter, simply because switching from the Transrotor to the FIIO SR11 digital transport brought about a clear improvement in terms of performance. Everything sounded bigger and also more convincing. From the analogue source the FIR-BoxX could sound a little thinner and more restrained than going through the Musical Fidelity M6 PRX power amplifier. This may have had to do with the FIR-BoxX compensating massive room modes at at 25 Hz and 50-60Hz, which the Musical Fidelity amplifier simply ignored, but then the same correction was in place when the music was streamed via the FIIO SR11, so the correction was surely not the sole reason. More tests are needed to determine if this effect had more to do with the setup or with the FIR BoxX itself.

Much to our surprise, the FIR-BoxX managed to maintain its perfect command of the speakers up to very high volumes. Each amplifier maintained perfect control of the driver and both timing and detail were very impressive. The sound is highly immersive, and although the sweet spot is similar to the passive, analogue system, walking through the room reveals a more even and tonally balanced impression. This makes listening more enjoyable for sharing a music event with friends and family.

Andreas Heggendorn and Bidjan Tschaitschian are currently working on the automation of the set-up procedure to make the technology more readily accessible to households. The one obstacle that remains is the need for the loudspeaker’s original crossover to be bypassed. (At least this is true for any conventional loudspeaker that is to be FIR-activated, see below.) On the other hand, the FIR-BoxX’s ability to mitigate all kinds of flaws in drivers, speaker cabinets, and listening rooms, makes its application highly attractive. Hobbyist loudspeaker builders, owners of older and lesser loudspeakers, HiFi enthusiasts with less than ideal listening spaces, all of them will benefit from the FIR-BoxX’s existence.

I look forward to hearing more of it.

Note: At the time of writing this, there is an ongoing FIR-ready loudspeaker development project by Daniel Gattig of Lautsprechershop. And Markus Grelka of Rose Handwerk already offers FIR-ready loudspeakers for the larger (100 WPC) MUSICWONDER ONE.