Gradient design principles

Sounds good in any room

Throughout its history, Gradient has designed speakers that work in the best possible way in a normal living space. What seems like an obvious thing has required a lot of research to lay the foundation for the design. The unique solutions that emerged have set Gradient apart from its competitors. Gradient's slogan, since the 1980s, has been 'Sounds good in any room.'

So-called high-quality hi-fi speakers are evaluated by measuring their sound curves in an anechoic chamber. In other words, in laboratory conditions where there is no reverberation, standing waves, flutter echo — no so-called room acoustics at all. The sound comes only from the sound source. Sufficiently perfect results have been achieved in such measurements for decades. Based on this, there has been hardly any real progress.

In their research in the early 1980s, Gradient’s product development team discovered that measurements in an anechoic chamber tell only a small part of the speaker's true performance. When a speaker that performed excellently in laboratory conditions was placed in a normal living room, it was not necessarily any better than a product that fared worse in the lab — often the opposite. Why is this? Gradient set out to find an answer to this question, and the result was a feature common to all Gradient products. We call this 'controlled directivity.' To understand this phenomenon, we must first start with the basics: What is stereo really about?

Gradient's founder, Jorma Salmi, presents the prototype of a controlled directivity speaker he designed for AAD in 1982.

What is stereo really about?

From time to time, even experts benefit from revisiting the fundamentals of sound reproduction. For instance, what is stereo really about?

In stereophonic sound reproduction, we have two channels. What does this mean in practice? During recording, multiple microphones might be used, but the final mix is always reduced to two channels: the right and left. Essentially, we have two signals, two waveforms, which are precise acoustic replicas of the original sound. These sound waves represent the sound pressure variations captured during recording, which are then converted into electrical waveforms and stored as digital recordings.

Now, we have two digital signals. These signals differ and continuously change, containing all the information needed to reproduce the width, depth, and even height of the performance and its acoustic environment. All the necessary information is embedded in these two signals. If we pause these signals momentarily, each one would have a single value, representing voltage or pressure. At any given moment, we have just two figures.

The challenge lies in reproducing these signals without quality loss. While electronic equipment like amplifiers minimally alters the signal, the real issue is how to deliver these signals to the listener's ears in typical rooms without degrading the audio information.

Standardized chain

Ensuring the unmodified transmission of a signal from the control room to a listener’s living room requires that all equipment used is compatible worldwide. Program sources and electronics are manufactured according to international standards, allowing HiFi enthusiasts to build systems using different brands. It’s even possible to mix tube and solid-state equipment. The biggest variations are found between power amplifiers and loudspeakers; some speakers with very low impedance may only work with certain amplifiers. However, in most cases, everything works well together.

Listening rooms are varying acoustic loads

Just as an amplifier interfaces with a loudspeaker, a listening room acts as an acoustic load on the loudspeaker. In recording studios, monitor speakers and control rooms are typically designed together to achieve optimal sound quality. However, this is not the case in typical living rooms. When building a house, there are no standards for acoustic load, leading to significant variations between rooms. As a result, most rooms are more likely to be acoustically poor rather than good for sound reproduction.

Given the unpredictable speaker-room interaction, it’s challenging to find a loudspeaker that works well in all rooms. Bass response alone can vary greatly: in a stone or concrete room, you may get excessive bass, while in a wooden house, the bass might be too weak.

Why do loudspeakers sound different in different rooms?

This problem has been studied to some extent by various loudspeaker designers and other experts. For example, the renowned Danish company Bruel & Kjaer, a manufacturer of audio measurement equipment (Møller, 1974).

The B&K team tested five different loudspeakers in three different rooms, with sound quality evaluated by five listeners in a blindfold listening test. The results were surprising: it was impossible to predict which loudspeaker would perform best in any given room. None of the loudspeakers sounded the best in all three rooms. This is understandable because each room creates a different acoustic load for the loudspeaker. Changing the load changes the results, much like how amplifiers perform differently depending on the load.

What advice did the B&K research team offer to customers looking to buy a loudspeaker system? They recommended taking various loudspeakers home and measuring them using pink noise.* Finally, select the pair that provides the smoothest response at the listening position within the frequency range of 60 Hz to 6000 Hz. Typically, this loudspeaker also sounds the best.

Before founding Gradient Ltd., Jorma Salmi, along with his colleague Anders Weckström, decided to study the room/loudspeaker interface problem (Salmi and Weckström, 1982). Instead of testing and measuring loudspeakers in ordinary rooms, they utilized an anechoic chamber. See their findings in "The Absolute Listening Test".

*This method is still in use today. For instance, John Atkinson of Stereophile uses a computerized system that averages 60 measurements for a single loudspeaker, totaling 120 measurements for a stereo pair. This type of measurement correlates well with subjective listening results. As seen in Stereophile (March 1997), the in-room measurement of the Gradient Revolution was an impressive 32 Hz - 10 kHz +/- 1.3 dB (see fig. 5: Stereophile Review).

The absolute listening test

At first, Salmi and Weckström conducted a listening test in a large anechoic chamber. What is an anechoic chamber? Simply put, it is a room without reflections—no reverberation, standing waves, flutter echo, etc. Essentially, it’s a space with no acoustics at all.

They listened to various loudspeakers at the spot where the measured frequency responses were as flat as possible (usually found on the so-called design axis). When listening to high-quality recordings through high-quality speakers, they were surprised by how good they sounded. Here are their main findings:

The sound was very clean and pure.
Imaging was excellent.
The acoustics of the recording venue were reproduced excellently.
Spatial information was reproduced much better than in ordinary rooms.
Sound quality was much better than in ordinary rooms.

After listening to loudspeakers in an anechoic chamber, they conducted the so-called Absolute Listening Test (ALT). The test was conducted as follows:

The loudspeaker under test was placed inside an anechoic chamber.
A high-quality measuring microphone (B&K 4133) was placed on the design axis of the speaker.
Music was played through the speaker from a source (record player, microphone, or tape recorder).
Monitoring took place outside the anechoic chamber.
By switching, it was possible to detect the difference between the speaker/microphone combination and the source (a straight wire).
Both headphones and loudspeakers were used for monitoring.

This test produced the following surprising results:

When the frequency response curve of the test speaker, as measured by the microphone, was flat, listeners could not distinguish between direct sounds and those reproduced by the speaker.
Non-linear and phase distortion, delayed resonances, and similar phenomena did not appear to affect the sound quality of high-quality speakers.
Flat frequency response was the most important factor.

Listening to good-quality recordings in an anechoic room sounded excellent. The difference between direct sound and that passing through the loudspeaker and microphone was minimal. At that point, they decided to investigate what happens in an ordinary room—or how your listening room kills sound quality.

The room-loudspeaker interface: How your listening room kills sound quality?

The fundamental question was: Why do loudspeakers with similar frequency response curves sound different in ordinary rooms?

In a typical reverberant space like a living room, only 20% of the dominant sound energy is transmitted directly to the listener, while 80% consists of reflections from the floor, walls, and ceiling.

Since an anechoic chamber has no acoustic reflections, the solution had to lie in the reflected sounds of the listening room. This proved to be the case. The core issue was how these reflected sounds deteriorate the frequency response. Research into human hearing provided insight. The key was the first reflected sounds reaching the ear, which travel a slightly longer distance than direct sounds (3 to 70 cm longer, or 0.1 to 2.0 ms in time). The human ear and brain perceive these early reflections as contiguous with the direct sounds, creating a "comb filter" effect, where the frequency response curve dips and rises alternately. As a result, even a good speaker's response worsens. The extent of this deterioration depends on the speaker's directivity and its distance from reflective surfaces, particularly the floor.

What are the subjective effects of early reflections on sound?

Imaging suffers.
Lack of spatial information.
Reduced three-dimensionality.
The sound seems to come from the loudspeakers rather than "out of nowhere" (the loudspeakers don't disappear).

The first Gradient was born

An absolute listening test provided Jorma Salmi with new insights into speaker design. He was so convinced by the results that he decided to establish a company around his inventions. This led to the creation of Gradient and soon after, the first product, the Gradient 1.0. (Read more about the operating principle of the Gradient 1.0 here.)

The primary design principles were minimizing early reflections, maximizing direct sound at the listening position, and achieving even radiation patterns—a technique known as Controlled Directivity.

Towards the Revolution

Research related to controlled directivity continued. The next focus was on reproducing low frequencies in a room and the significant disturbance caused by so-called standing waves. Standing waves are a common problem, especially in small to medium-sized rooms. Rooms with large parallel surfaces easily generate strong standing waves. These standing waves emphasize certain low frequencies, causing them to dominate the sound. Bass starts to boom, and the lower register of instruments becomes distorted.

Based on research conducted by Gradient, an effective way to minimize this phenomenon was to use a so-called dipole bass, which radiates in only one direction, forward and backward in opposite phases (Salmi and Weckström 1982; Salmi, 1992). The first bass speaker utilizing the dipole principle was launched in 1991: the Gradient SW-63 was developed as the bass section for the electrostatic Quad ESL-63 speaker.

Soon, the Gradient Revolution was created, in which the dipole bass section could be directed in three alternative orientations. During installation, the orientation was chosen to minimize the excitation of standing waves in the room. A special acoustic resistance enclosure was developed for the speaker's midrange, which provided a cardioid directivity pattern towards the front. Early reflections from behind and to the sides of the speaker were practically eliminated. You can read more about the development of the Gradient Revolution here.

Over its 40-year history, numerous innovative speaker solutions have been created, often deviating from conventional designs. A common feature of all products has been the minimization of the detrimental effects of the room. The speaker, and above all its behavior in the room, has by far the greatest impact on the audio reproduction chain.

References

Møller, H, 1974: Relevant loudspeaker tests in studios, in hi-fi dealers' demo rooms, in the home etc., using 1/3 octave, pink weighted, random noise. AES Convention 47, Copenhagen.

Salmi, J. and A. Weckström, 1982: Listening room influence on loudspeaker sound quality and ways of minimizing It. AES Convention 71, Montreux.

Salmi, J. and A. Weckström, 1983: A new, psychoacoustically more correct way of measuring loudspeaker frequency responses. AES Convention 73, Eindhoven.

Salmi, J, 1992: Dipole source placement in a room. AES Convention 92, Vienna.

Pictures

Jorma Salmi and AAD-prototype, The editorial team of HIFI-lehti.

"A Gradient Bear", Heli Ahonen

"Gradient Revolution", Hannu Hyrske

Other pictures, Gradient

Menu

Settings