WHAT IS ROOM CORRECTION AND WHY DO YOU NEED IT?
All manufacturers of high-end audio products face the same problem: audio that is measured in the laboratory and tested by experienced listeners is far from what most future system owners will enjoy at home. This problem stems from the fact that speakers are developed in sonically-ideal listening rooms with optimized dimensions and acoustic treatments; even if manufacturers used a normal room, the result in an individual’s room will always be much different.
At Lyngdorf, we recognized that just focusing on creating processors and amplifiers with a linearity of +/-0.2 dB makes no sense when the music is being played in a room with a linearity of perhaps +/-10 dB.
Our products are designed to make speakers perform perfectly in rooms full of furniture, curtains, and bookshelves, because we create products for use in everyday life.
How Our Venture Into Room Correction Began
Our first processors and amplifiers had digital signal processing (DSP) which makes filters relatively easy to create, so our initial approach to room correction was an obvious one. We took the following steps:
1. Measure the frequency response at the listening position
2. Add the target curve in the form of a perfect frequency response
3. Create filters that cancel out all imperfections and upload the filters to the DSP
There was only one problem with this approach: it did not work!
Because we didn’t have the microphone in the exact same position each time, new measurements would detect other peaks and dips in the frequency response, and a calibration could take months of fine-tuning and experimenting. The idea of room correction became a toy instead of a functional tool.
We identified basic errors in this approach:
1. You don´t get sufficient data through a single-point measurement, and it is very hard to evaluate which problems can be corrected and which cannot. The acoustical problems of a room are in all three dimensions.
2. If a measurement and resulting calibration was used for removing a single problem in the bass, the process might work, but the outcome with poor data at hand was most often a disaster.
3. The perfect one-size-fits-all target curve does not exist. We realized we should not have all speakers calibrated to the same target curve, because doing this means all speakers would sound nearly identical.
The objective in developing a room correction technology, therefore, was that:
1. It must work in the full frequency range
2. It must function with all types of speaker setups
3. It must adapt to the speaker system in action (not reduce the qualities of the speakers)
4. It must allow for the most dynamic music reproduction (optimize impulse response)
5. It must be simple to calibrate
Loudspeakers and Acoustics
All loudspeakers sound different, and you choose a given model because you love its performance and looks. The type of loudspeaker, its design, and the choice of components will create a specific and desirable product performance.
Reflections in the room are natural for our experience of music; in fact, it feels quite unnatural if reflections are mostly or completely removed. At the listening position, the energy of all reflections is actually higher than the energy of sound directly from your speakers. However, your brain will still identify the position of the speakers, as the direct sound from the speakers reaches your ears first (this is known as the Haas effect). Major problems occur in the first reflections, where the various frequencies bounce off the walls, floor, and ceiling, thus arriving at the listening position with a delay related to the distance travelled.
The traditional rules of thumb for positioning speakers in the room take this into consideration, because having the speaker a meter away from the front wall and about 1½ meters from the side walls will result in the reflections arriving at the listening position out of phase with each other, so no frequency is intensified. This “out-of-phase” arrival of reflections will muddle the sound, which is why positioning speakers in this manner is the worst possible solution if you want optimal impulse response.
If you do not have a 100% symmetrical speaker setup, the left and right speakers will not have identical acoustical coupling to the room, and with different tonality from left and right speakers, the stereo perspective will be reduced.
However, the biggest problem in your room is room modes, which occur for all frequencies and in all three dimensions: front to rear, side to side, and floor to ceiling. As a frequency is bouncing off opposing walls, it will cancel out the energy when it is out of phase. This will result in a negative phenomenon: at every point in the room, you will get a dip on specific frequencies.
With a traditional positioning of speakers, the room modes of the selected position will result in the inability of the speaker to transmit sound on specific frequencies. Thus, the optimal position of traditional box speakers is against the wall, where there are practically no modes. This position also eliminates the big first reflection from the wall behind the speaker.
For speakers not designed for this position, the audible result will be a tighter, more dynamic bass reproduction—but will also contain too much bass. We have a cure for that: it is called RoomPerfect™.
Our studies of the various problems of room dynamics resulted in the development of the patented room correction system called RoomPerfect™, a sophisticated technology with a completely user-friendly setup.
The RoomPerfect™calibration process is fully guided through the display of the processor/amplifier, and all equipment required for calibration is included with the product. (No computers required.)
RoomPerfect’s™ unique calibration process contains two steps:
1. A single measurement is taken at the listening position to give RoomPerfect™information about:
High frequency roll-off
Characteristics of low frequency roll-off
Left and right speaker differences (when speakers are not positioned in perfect symmetry)
Frequency response 20 – 20,000 Hz
2. A number of random measurements of the power response are taken throughout the room. These measurements provide information to the system about the causes of the problems identified at the listening position. The system then determines what to correct:
Power response throughout the room
From the room measurements, RoomPerfect™ knows how your speakers sound, and it is now this tonal balance which will be cleaned up individually for each speaker.
In your RoomPerfect™ enabled product, you can find a Room Correction index, which is a simple expression of the audible changes of the filters added. Basically, this index is a function of the complexity and importance of the filters and frequencies. Our best advice is to trust your ears—even the smallest correction can be of utmost importance for you if it cleans the reproduction of your favorite singer.
If you have been able to set up your speakers with a nearly perfect frequency response, RoomPerfect™ will not change much. This is proof of the RoomPerfect™ concept. In this case, we do recommend that you try optimizing the impulse response by positioning your speakers closer to the front wall; then do a new RoomPerfect™ calibration and you will enjoy the best possible music reproduction.
We love dynamics—that is what brings music alive. Impulse response refers to the perceived experience when a speaker is replaying a piece of dynamic music. It is a common problem that a bass frequency is enhanced in your room—and that alone will mask the general experience of dynamics. The fact that you have multiple reflections of the sound coming to your listening position delayed due to the distance travelled will mask the perceived dynamics even further. With correct positioning of speakers and a RoomPerfect™ calibration, you have the optimal dynamic performance!
Test Signal and Measuring Techniques
In order to get the necessary room knowledge for our ambitious goals, we had to invent new ways of obtaining data.
Using traditional test signals (pink noise, for example) was not possible, as we wanted a long time window to allow for analysis with high resolution, including the effect of the reflections. With all kinds of noise influencing the measurement, pink noise can only be used in a very narrow analysis time window, thus focusing on the direct sound from the speaker.
We found a way to use multiple pure tones, which allowed for a long analysis time window with both high frequency resolution and excellent suppression of noise. For this purpose, we designed a dedicated microphone, which has the optimal construction for these measurements.
For both the focus measurement and the room measurements, the system generates 50 low tones in the left channel, providing data from 20 Hz to 350 Hz, then 71 tones to provide information for the frequencies up to 20 kHz. Then low and high tones are played through the right channel.
The tones are evaluated individually, and when the system knows that the measurement is exact, it will store it for subsequent processing. After each measurement, the system evaluates the information received and decides if it needs more information to continue.
For the calibration to function, the room measurements must be made at random positions throughout the room in order to obtain information about the energy response in the room. The illustrations below, taken from the RoomPerfect™ manual, indicate the space from which you should gather information. If you do symmetrical measurements, or if you try to “calibrate” for specific objects in the room, you will not provide the system with the data it requires, and the calibration will not be optimal.
RoomPerfect’s™ information gathering, evaluation of information, and subsequent calibration make up the largest patent in the field of audio processing, holding a total of 55 claims under the RoomPerfect™ patent.
The beauty of the RoomPerfect™ calibration is that it knows whether a specific problem in the frequency response can be corrected—and by how much—without stressing the speaker or the amplifier in use. Below you will find a typical example of the frequency response before and after a calibration.