This review page is supported in part by the sponsors whose ad banners are displayed below

Lack of cabinet resonance can be heard in the non-colored nature and purity of the sound. This is undoubtedly also the case with the Credo 4. Yet I am inclined to think that Credo 4's rigid and non-resonant enclosure (made of some special density MDF if I understood correctly) contributes to the sound's excellent spatial and imaging properties only partially. The other ingredient is the woofer crossover. The center image and the sound at non-omni-directional frequencies in general was very stable and localization of sound sources on the horizontal axis exemplary. In music samples where the instruments are mixed on the sides to easily get stuck to the cabinet of an ordinary speaker, they remained clearly separate from the Credo 4. With Vivaldi's Concerto for two mandolins, most speakers are able to position the two instruments where they play but with the Credo 4, their location was extremely accurate. When the room itself is dead, this means pin-point imaging accuracy.

The sound was not only generally spacious and airy but also well stratified. There was nothing noisy or vague about it, no uncertainty about presenting spatial cues on the record. In this respect, the Credo 4 reminded me of some of the smaller Mark & Daniel models. The sound's intensity, conciseness, integrity, strength and focus were of a high level. The sound had sovereignty and tranquility, which increased responsiveness towards the music fed through the speaker.

In designing the cabinet, Esa thus made use of the necessary fact that the front baffle would be notably tilted back. Why necessary? To coordinate the phase behavior between woofer and mid/bass unit: "... most design engineers do not question the fact that the only relatively universal test of a speaker's tonal equilibrium is to measure the speaker on the principal axis. This is the image of the sound running directly from the speaker to the listener. If we are unable to determine the invariable principles of interaction between the speaker and the room, we must completely disregard those phenomena and completely eliminate their effect, i.e. any reflection, to standardize the measurement method. Measuring in a dead room or using an impulse system may not be an ideal solution, but (in practice) they are the only reasonable conditions for comparisons between different speaker sets, allowing one to establish whether at least on the principal axis the loudspeaker is operating correctly."

I reproduced the above quotation to suggest that Esa has taken the spoon in the right hand and simply made a virtue of necessity. As they wisely proclaim, even if we could determine what the sound characteristics from of each driver are on any axis and at any frequency, it is impossible to clearly and unambiguously define
theoretically what those characteristics should be. And if this is so, then the optimization must focus on a factor whose impact on the sound can be predicted and compensated by design: the floor bounce.

With a floorstanding speaker, the distance between drivers and floor is a known fixed quantity whereas the distance from other reflective surfaces such as ceiling, walls, windows and furniture varies unpredictably from room to room. The same holds true for one aspect of the listening position - the distance between ears and floor is more or less constant. Reflections themselves aren't evil (they add to the naturalness of the sound) but when close to the sound source, they are. Floor reflections will always be delayed compared to direct sound to distort the location of apparent sound sources and cause frequency response squiggles depending on the phase differences between direct and reflected sound as measured on the zero axis. At which frequencies these peaks and troughs occur depends on the distance between loudspeaker and listening seat.

At two meters and with the listener's ear and the mid/woofer 80cm off the floor as in Esa's example, phase-induced attenuation occurs around 300Hz, at 5 meters around 650Hz. No speaker radiates those frequencies solely to the listening position and there is no way to absorb
these frequencies with a simple throw rug. In two-way speakers this cannot be corrected but in three-way speakers, the frequency range affected by the phase difference can be by lowering the woofer closer to the floor or raising it higher where the effect is less harmful and can be more effectively damped.

Following Esa's reasoning in an idealized case, the crossover frequency should be set well beyond the practical operating range of the woofer, say above 650Hz but even a more moderate parallel solution helps alleviate the floor bounce problem. It is for this reason that the two 22cm aluminum cone ScanSpeak Revelators of the Credo 4 are positioned closer to the floor while the tweeter and mid/woofer are at ear height. The crossover frequency for the woofer is not abnormally high but not as low as is common. But why is the front baffle tilted? If it were upright and the woofer 30cm above the floor, the distance between the listener's ear at 80cm and the woofer would be longer than the distance of the mid/woofer, resulting again in phase errors (more problematic in 2.5-ways). The problem can be partly remedied with the filter but according to Esa, it is more beneficial to use low-order filtering for the bass and tilt the front baffle. The reason for the slanting front baffle is thus the need to match the sound between mid/woofer and woofer.