The basis of good design is – paradoxically – the purpose of the design itself. When designing a floor-standing loudspeaker, for example, it is necessary to take into account the way it will be used (its purpose). A floor-standing speaker is likely to be used in positions more or less distant from the rear wall, in medium to large rooms; the optimal listening position shall be expanded horizontally.

Considering these assumptions as redundant or even limiting, denotes the classic attitude of those who believe that audio is a field in which it is enough to aspire to perfection or to an absolutist ideal. Not considering the importance of reconciling the needs of the real listener with the indissoluble laws of physics. There is only one certainty: the design of a loudspeaker is the result of undeniable compromises and the skill of the designer lies in optimising their coexistence.

The starting point of a bookshelf loudspeaker – as its name suggests – must take its final positioning into account: more or less close to the rear wall, in rooms of very variable size; the listening position must be as unrelated as possible to the reciprocal speakers’ and listeners’ positioning. Last but not least: the size of a bookshelf speaker should not exceed what its name suggests.

Starting points of Atomica Strato project:

1. Speakers close to the rear wall: bass emphasis and reduced soundstage depth.

2. Variable listening room dimensions: high power handling is needed.

3. Variable listening positions: optimal horizontal and vertical dispersion is needed.

4. Maximum speakers’ dimensions must not exceed 32 cm in both depth and height.

In the classic bookshelf loudspeakers, point 4 is considered the main design limit, since limited dimensions stand for limited internal volume, and therefore lack of bass; various tricks are adopted, such as reduced woofer’s dimensions or extended chassis’ depth. These solutions cannot fulfil the other points because: 4-inch (12 cm) woofers cannot handle adequate power (point 2); obtaining the highest possible bass gain is not the best way to deal with point 1; creating a box with a depth close to 30 cm does not take into due consideration point 4: many users would have to put the speakers against the rear wall.

Point 4 is Atomica Strato’s starting point and strength. It is possible to stay within these dimensions, and it can be even advantageous. After the first prototypes, the dimensions of the project have been fixed in a height of 31.5 cm, width of 20 cm, and depth not exceeding 25 cm including binding posts; this is the maximum possible depth to not interfere with rear reflex port.

The resulting internal volume allows for low register roll-off which compensate for bass emphasis, fulfilling point 1 (the prototype has been tested on 8 different positions, more or less close to the rear wall – both listening sessions and measurements were performed). Such internal volume also allows for a reflex tuned exactly to the fs of the driver, thus drastically reducing cone excursion and increasing power handling and dynamics according to point 2 (not to mention the benefits in terms of clearness and intermodulation distortion).

Point 3 takes advantage of a reduced front wall, because baffle diffraction is optimised across the relevant frequency range. Polar dispersion has been further enhanced with the unmistakable Strato’s smoothed edge. Such uniform dispersion can also compensate for the soundstage depth reduction described in the second part of point 1, as the creation of a very diffuse sound field and the presence of uniform first reflections make up for the above-mentioned lack.

Oblique polar dispersion at 0° – 30° – 45°. 10 dB scale with 1/3 oct. smoothing.

Great linearity independent of listening position is also due to the special crossover network, which required countless listening and testing sessions.

The special shape of the Atomica Strato loudspeakers allows for the creation of incredibly rigid and therefore energy-efficient and dynamic chassis. Atomica Strato has an efficiency of 88 dB 1W/1m, which is truly unique for a bookshelf speaker and makes it suitable for even low-wattage amps. A special self-levelling aluminium support enhances stability and dampening.

This section presents the theories behind the choice of construction materials for the Atomica Floor and Book 2 loudspeakers: speakers whose cabinets are based on the contrast between okoumè plywood and glass.

We now approach the coupling of two materials whose characteristics are – to say the least – opposite. But before talking about this, we need a small introduction to the physical phenomenon of resonance:

1- We speak of mechanical resonance when a mechanical system is excited at a frequency compatible with the characteristic oscillation of the system itself, and it amplifies it. This phenomenon causes an accumulation of energy within the system, which is counteracted by the damping coefficient.

2- Acoustically speaking, a range of vibrations are generated inside a resonator (the cabinet) which are characterised by its geometric and mechanical properties. The resonance phenomenon involves both the resonator and the vibrating element (the loudspeaker).

3- A vibration that is triggered in a system made up of differentiated geometries and sections is called parcellized. These geometries and sections vibrate simultaneously and separately giving rise to a complex motion (multiple harmonic motions).

It is now clear why glass alone cannot optimally contribute to the composition of a cabinet: although it is an exceptionally hard material, its high specific weight and low damping cause considerable energy accumulation in the system (see 1).

As for the wooden frame, which will house the loudspeakers, a considerable effort has been made in the Atomica loudspeakers to minimize energy exchanges between the parts, by using carefully studied geometries and shapes (see 2). For further information, please refer to the article “Elements for cabinet design”.

The above-mentioned frame is therefore characterised by high stiffness and very low elasticity and mass. A perfect element, were it not for its poor damping properties. And this is why glass is incorporated into the frame.

The combination of these two materials is achieved by interposing a special semi-expanded seal, through which the glass is bolted against the wood at a specific – electronically measured – pressure. Besides giving the glass the necessary damping, this procedure allows the wooden frame to gain the necessary damping mass.

The resulting cabinet takes due account of points 1-2-3. Parcellation is best exploited because two materials that are completely incompatible in terms of hardness, elasticity, and mass are coupled in such a way as to mutually dampen the resonance motions and delocalise them into the frequency domain.

The following charts show the harmonic distortion detected on Atomica Book 2 and Atomica Book Fraxinus respectively. The latter differs only in the side walls, which are made of solid ash instead of glass.

Both charts show a preponderance of second and then third harmonics, highlighting a pleasant characterization. The total average (upper line) is very close in both models. Note, however, its trend and also that of the distinct harmonics: in the glass model, we can appreciate an exceptionally constant and linear trend, free of perturbations, and this shows that this type of cabinet provides superior performance, better damping, and greater naturalness compared to its counterpart built with traditional materials.

REFERENCES

• Ahnert W, Steffen F, “Sound Reinforcement Engineering”, 1999.
• Arnol’d V, “Mathematical Methods of Classical Mechanics”, 1999.

The spaciousness of soundstage is not only determined by the features of a system (or orchesta, for example). Spaciousness is also related to the subjective perception of space created by the room acoustics. Therefore, also speakers positioning and paradoxically the room parameters themselves cause the illusion of being in a larger room.

For a long time, it was common belief that spaciousness was a direct function of an even distribution of the sound field within a listening room. Only since the late 1960s real progress has been made in finding the cause of spaciousness: diffuse field is not the primary cause.

The element that most emphasizes the spaciousness is the lateral arrival of the first reflections. Therefore, a wider spatial impression can be created wirh few delayed early reflections reaching the listener’s head from lateral directions.

The diagram above shows the path difference between direct sound (blue) and early reflections (violet); “a” is the arrival angle. 

Early reflections are independent of room’s reverberation. Only reflections with a delay time with respect to direct sound in the range from 5 to 80 milliseconds ( from 6 feet in to 90 feet in  additional path) contribute to spaciousness. Frequencies from 500 to 3000 Hz contribute most to the perception of spaciousness, and their arrival angle is proportional to their effectiveness. For example, a zero arrival angle (i.e. early reflections from the room’s bottom) will result in no spatial enhancement due to the lack of sound pressure difference between the listener’s ears.

Therefore, is it possible to improve spaciousness with better speakers’ or listener’s positioning. It is also possible, when desired, to modify room’s acoustics in order to adjust early reflection to the listener’s seat.

First reflections’ arrival angle is the factor which most contributes to spaciousness, but not the only one. For example, spaciousness increases proportionally to the listening volume.

REFERENCES

• Barron M, Marshall AH, “Spatial Impression Due to Early Lateral Reflections in Concert Halls. 1981.
• Burgtorf W, “The Percectibility of Delayed Sound Signals. 1964.
• Neu G, Mommerz E, Schmitz A, “Investigation on the True Directional Sound Reproduction by Playing Head-related Recordings Over Two Loudspeakers”. 1992.

Issues generated by speakers positioning mainly concern low frequency response. Electrodynamic speakers are small in comparison with such frequencies wavelength, therefore the sound pressure is proportional to the radiation load: in other words, the closer the speaker is to the intersection of two or more walls, the higher the pressure will be.

If the sound is generated in angular conditions comparable to a wall-embedded sound source, the radiated power will be twice as high as if it were generated in free space. On the other hand if radiation is confined by two intersecting walls, the power output will be four times greater than in free space. If the source is placed in a corner formed by three mutually perpendicular boundaries, instead, its power output will be doubled once more. These conditions are true as long as the source and the distance between these walls are sufficiently smaller than the relevant wavelengths.

Such considerations show how low frequencies are critical during domestic-listening systems set up. This is true even in acoustically equipped rooms: low frequencies are distributed according to rooms geometry and speakers positioning, because they characterize nondiffuse resonance modes and sound reinforcement caused by radiation load.

Here is a quotation from Groh’s studies on low-frequencies acoustics: “The ultimate determinant of fidelity to an original source is the sound field at the listener’s position. Even if a loudspeaker system is capable of delivering uniform power to a room, the energy is redistributed by the room nondiffuse resonance modes, and the listener’s location with respect to these standing waves is not knowable. Nevertheless, if loudspeaker systems are designed with due regard for those boundary effects, it will certainly reduce the average deviation from the ideal sound field”.

After these words, Groh described a method of placement equalization. This useful technique is rather distruptive of conventional listening rooms’ arrangements.

The assumption of those considerations has enabled our company to find a solution to low-frequencies reproduction issues: Atomica Audio’s products are equipped with the Atomica Reflex. The Atomica Reflex is a patented resonator which has been created to be easily set by the user according to their tastes and their listening room. Such device solves the issues caused by the listening room or listener’s and speakers’ positioning.

The Atomica Reflex features a resonator whose extension can be adjusted by means of a knob positioned on each speaker side. The lowest resonator extension guarantees maximum bass emphasis and power, while the pipe extension causes a graduated low frequencies damping. Maximum extension causes pipe blocking and no low-frequencies emphasis. In other words, the system can work both like a reflex and a closed box passing through middle positions, allowing great adaptability.

The following graph shows a prototype’s response. It is a small box fitted with a 6” woofer equipped with the Atomica Reflex (red line: max pipe extension; green line: min pipe extension; violet and yellow: two middle positions).

To better understand the possible applications of this device, the following graphs show simulated responses of a classic reflex floorstanding speaker, in different positions.

Speaker distant from side and rear walls

Speaker at 20 inches from side wall, 3 feet from rear wall

Speaker at 3 feet from the side wall, close to the rear wall

Graphs 2 and 3 are now shown again but with the addition of the resulting response after setting up the Atomica Reflex (green line). Speaker and position are the same.

Thanks to the Atomica Reflex, it is possible to abtain the best calibration not only in the situation shown in graph 2 – possible even in the best equipped rooms – but also in challenging situations like the one shown in graph 3. The user can set it up easily every time adjustments or positioning changes are required, without any electronic artefacts.

REFERENCES

• Allison RF, Berkovitz R, “The Sound Field in Home Listening Rooms”, 1972.
• Allison RF, “The Influence of Room Boundaries on Loudspeaker Power Output”, 1974.
• Davis D, “Further Comments in Directivity Factor”, 1973.
• Groh AR, “High-Fidelity Sound System Equalization by Analysis of Standing Waves”, 1974.
• Robber RJ, “Approximations to the Directivity Index”, 1974.

As a general rule, a cabinet can be described like a structure having two purposes: to be fitted with one or more loudspeakers, and to prevent loudspeakers rear radiation from mixing with direct radiation.
Speakers walls tend to be excited by both internal pressure and by the energy transmitted from the loudspeakers structure.

A list follows containing many design solutions, starting from classic and ending up with up-to-date techniques.

Sound absorbing and deadening

Nowadays many designers agree that absorbing materials cannot be useful in cabinet damping. Some even argue that absorbing is detrimental for transient response and efficiency. It can be useful to prevent the speakers’ rear radiation from bouncing back to them; however, that issue can be solved beforehand by studying internal cabinet geometry.
Much better are damping materials like hi-density PVC or sheet tar, provided they perfectly adhere to the wall. They are useful even if they don’t cover the entire wall. The best position is on the wall centre, covering at least one third of the wall.

Bracing

A classic of good cabinet design is bracing, which consists in dividing a wall into two sections by adding a strengthening element parallel to the longer side. In this way a higher fundamental resonant frequency is obtainable without varying panel dimensions.
It is possible to joint opposite cabinet sides so that they cannot vibrate according to internal pressure. This kind of stiffening is more efficient when used to connect two or more bracing mounted parallel to the longer side of the walls.
Such bracing can increase the resonant frequency by up to four times.

As a general rule, it can be said that high resonant frequency is desirable for three reasons:
1-Every musical track average energy is in reverse relation with frequency.
2-High frequency resonant mode in a cabinet wall requires that different wall sections vibrate in opposite directions. In this way is very hard to obtain high sound pressure inside a cabinet. Outside there will be not much sound propagation due to cancellations and reduced moving areas.
3-Due to the reduced moving areas, a wall excited by high frequencies will tend to radiate perpendicularly to its surface. This is desirable for every cabinet wall except for the one facing the listening position, as the output energy will not overlap with the direct sound.

Walls’ shape

Resonant frequency isn’t independent of the cabinet walls’ shape. Thickness and area being equal, a long and narrow wall will present a much higher resonant frequency than a square wall. Theoretically speaking, resonant frequency is equal to the reciprocal of the square root of the shorter dimension; whenever high accuracy is necessary, experimemtation is essential.
Another rule to raise resonant frequency in a structure is to employ high stiffness to weight ratio, keeping in mind that good damping requires a certain amount of mass.

Materials mixing

Different materials are characterized by different resonant frequencies due to their specific gravity, their fibre, or composition: they tend to react in different ways to mechanical stress. Those differences can be used to evaluate the best choice for a given project.
The best way to take advantage of those differences is by mixing different materials. This can be accomplished not only by coupling different layers, but also by assembling a cabinet using a different material for each wall. For this reason, Atomica speakers’ cabinets are built using up to four different materials. For example Floor and Book 2 speakers are made from two different types of solid plywood, tempered glass, and AISI 304 stainless steel. They work together as an extremely inert structure, which is free of every kind of interference. This kind of structure remains inert also at high listening levels, not only because of their excellent materials quality, but also because those materials present such different mechanical and resonance parameters that they reciprocally repress resonant forces.

Walls’ curvature

Resonant frequency depends on walls’ curvature. Up-to-date machining processes make possible to build curved surfaces, made of wood or other materials, giving the opportunity to enormously increase the stiffness and therefore the resonant frequency of a cabinet. Atomica speakers are made using the best numerical control machines, making it possible to obtain every kind of shape, creating loudspeakers that are structurally unrivalled and at the same tim

e incredibly pleasing to the eye.
The use of curved shapes can be taken to its extreme by making a spherical cabinet. We have achieved this goal with the Atomica Giove speakers, using a specific proprietary milling program that allows the creation of a perfect sphere made of very hard solid wood.

Materials’ direction

There are materials characterized by a directional structure, or by a fibre (natural wood). Those materials can be glued or fixed togetrer to form a multi-layer panel. This method is the one used by plywood manufacturer. For instance, the most common type of 30 millimetre-thick okoumè plywood consists of twelve layers, which are alternately oriented, glued and pressed. Thanks to torsional and bending compensation, this process allows to have at least four times better characteristics than single layer okoumè wood.
Atomica Floor, Book 2 and Bookshelf XL speakers employ an extreme concept of fibre oriented compensation. Their cabinet are made using multiple layers of plywood, glued and pressed using hydraulic press at several tons of pressure. We obtain a very special cabinet, whose wooden part consists of  dozens of alternately oriented wooden layers. Layer structures also makes it possible to totally incorporate bracing, with shapes and efficiency not even remotely imaginable with conventional methods.

Gluing, fitting and screwing

Whatever the procedure to assemble the parts of a cabinet is, i.e., by gluing, fitting, and/or screwing together, an experienced manufacturer will be faced with a doubt: will I get a pair of cabinets with similar mechanical characteristics? Given the importance of equality between two loudspeakers required by stereophony, it is normal for a conscious designer to have an eye on how the cabinet parts are assembled. This phase can create stereophonic dissimilarity far more severe than those usually avoided by, for example, selecting and matching the components of a crossover.
In Atomica speakers the problem is tackled thanks to innovative machinery and techniques, never before used in the audio world.
For instance, the parts of a cabinet – whether made of wood, glass, or metal – are selected and milled by numerical control as well as any screw, bolt, or bonding dowels. At this point the frames are coupled and glued together in the same press, at the same pressure, and for an equivalent time. All clamping components are screwed with very precise dynamometric tools. This results in cabinet sets whose accuracy is excellent to say the least.

REFERENCES

• Ahnert W, Steffen F, “Sound Reinforcement Engineering”, 1999.
• Bailey AR, “Non-resonant Loudspeakers Enclosure”, 1965.
• Benson JE, “Theory and Design of Loudspeaker Enclosures”, 1972.
• Olson HF, “Direct Radiator Loudspeaker Enclosures”, 1972.
• Small RH, “Closed-Box Loudspeaker Systems”, 1972.
• Small RH, “Direct-radiator Loudspeaker System Analysis”, 1972.
• Zwikker C, Kosten CW, “Sound Absorbing Materials”, 1992.