Optimal 4pi-acoustic-power vs. frequency response - Forum

Hi Olli,

While the human hearing system does not change to fast, the consensus on how loudspeakers should work seems to change much faster. There are some widely accepted recommendations on the listening conditions enabling the listener to perceive small details in an audio recording. One of the more widely accepted of these is Recommendation ITU-R BS.1116-1. There, it is recommended that the frequency response of the loudspeaker system at the listening position is flat, i.e. the same level of audio is received at all audible frequencies. Naturally, one must also look at the decay characteristics of the room at the same time. The capacity of the room to store audio will change our perception of the sound color. The same recommendation has it that the decay time of the room should be equal at all frequencies.

The power response is the power radiated by the loudspeaker as a function of frequency. The acoustical axis of a loudspeaker is the direction of listening to the loudspeaker where the frequency response has been designed to be optimal. Usually the frequency response is designed to be flat on the acoustical listening axis in anechoic conditions (free field). If the power response of a loudspeaker is flat at the same time the frequency response is flat on axis in anechoic conditions, the loudspeaker radiates the audio energy in approximately the same space angle at all frequencies. When the loudspeaker is placed in a room, the audio radiated to directions other than the listening direction becomes audible because this sound energy remains in the reverberation of the room and decays over time. This energy will modify the listener’s perception of the sound color.

In conventional loudspeaker designs, the radiator diameter (size) causes the radiator to become more directive when the frequency increases. If the sound level on axis remains the same, this translates to outputting less power for higher frequencies. This then leads to designs where the power response can slope down for higher frequencies.

You mention an “optimal” power response curve slope. Such figures can exist for certain types of designs. It is dangerous to say that this slope would be universally true. Monotonically down-sloping and smoothly changing power response is usually advantageous, however, compared to power responses that suddenly change when frequency changes. Sudden changes imply that there can be large differences in the amount of power put into the reverberant field of the room when moving from one frequency to the next. This can become audible as disturbing colorations.

The Directivity Control Waveguides (DCW) that Genelec uses can maintain controlled directivity and improve the neutrality of audio compared to designs that do not use them. The DCW increases the sensitivity and the acoustic output of a driver towards low frequencies compared to a system that does not use a DCW. All Genelec drivers are electronically equalized for flat frequency response. The combination of the DCW and the electronic equalization can create a nearly perfect radiator with controlled directivity combined with a very flat on-axis response.

For any living room, including Finnish living rooms, the response calibration should result in the experience of a flat frequency response at the listening location. The reverberation in the living room will significantly affect this. All Genelec products contain the room response calibration settings to compensate for the colorations affected by the reverberant field of the room. These can be set by listening or by using a measurement tool.

The sensation of the “brightness” of audio can be very personal. There are also differences in if listeners want to achieve very neutral reproduction or if they prefer a certain type of sound balance. Sometimes this leads to responses that indeed are sloping down toward the high frequencies.

The picture you are referring to (“tweeterlens.jpg”) is not a picture of Yamaha NS-10 but of DynaudioAcoustics BM5A MKII. It does not have an acoustic lens on the tweeter. There is no DCW structure either. This type of design has directivity properties essentially determined by the size of the front baffle, and the diameter of the tweeter. These directivity properties are affected by the possible breakups of the dome structure of the tweeter, causing the dome to no longer have pistonic movement. Loudspeakers with this basic design can be made flat on the acoustical axis but usually show relatively strong variations of the power response. Usually the reverberation characteristics of the listening room will have stronger effect on the perceived sound color of products with this design approach compared to designs that use the DCW structure.

Dr. Aki Mäkivirta
R&D Director