Aim of this STF project was to find a close-to-original recording and reproduction system which fulfills the following requirements:

• The background noise recording technique should be
  – Easy to use
  – Easy to calibrate
  – Capable of wideband recording
  – Available at reasonable costs
  – Mostly compatible with existing standards and procedures used in telecommunications testing
  
  

• The background noise simulation arrangement should be:
  – easy to set up
  – usable in typical office type room environments without special acoustical treatment
  – provide a mostly realistic background noise simulation for all typical background noises occurring in telecommunication applications
  – easy to calibrate
  – mostly insensitive to the positioning of (test) objects in the simulated sound field
  – applicable to all typical terminals used in telecommunication
  – available at reasonable costs

A variety of methods for close-to-original sound recording and reproduction have been reviewed. The 4.1 loudspeaker arrangement was selected from the methods reviewed as the most suitable to fulfill the requirements. It provides background noise simulation in standard office rooms as well as in cars.

Simulation of sound fields in office room

The setup for simulating a sound field in office rooms is shown in figure 1. The system consists of 4 loudspeakers and one subwoofer which are arranged as shown in figure 1.	Figure 1
Before a background noise can be simulated via this arrangement, the system has to be calibrated and equalized. Figure 2 shows a block diagram of the calibration and equalization setup.	Figure 2 (Click on image to enlarge)

The equalization is done in several steps:

1. Separate equalization for each of the four loudspeakers
2. Separate level adjustment for each loudspeaker
3. Equalization for the two left-hand and the two right-hand loudspeakers
4. Equalization and level adjustment for the subwoofer
5. Delay compensation
6. Equalization for all loudspeaker (verification and reduction of cross-talk)

A standard office room was equalized as described above. To determine the capability of the 4.1 loudspeaker arrangement the following test was conducted: In addition to the 4.1 loudspeaker setup three sound sources were placed in the office room: 2 additional loudspeakers and a computer which produces the typical fan noise. A part of this setup is shown in figure 3. The sound sources which produce the “original” sound field are marked green. The speakers belonging to the simulation arrangement are marked red. Note that the rear speakers of the 4.1 loudspeaker arrangement are not visible on the photo.

Figure 3 (Click on image to enlarge)

Via the additional loudspeakers a binaural recording (recorded in a lounge during the lunch break) was played back. This “original” sound field plus the computer’s fan noise was binaurally recorded with the artificial head in the centre of the 4.1 arrangement (see figure 3). In the second step this binaural recording of the original sound field was played back via the 4.1 simulation system and again recorded with the artificial head. Figure 4 compares the power density spectra of the original and the simulated sound field recorded at the artificial head’s left respectively right ear.

Figure 4 (Click on image to enlarge)
Left ear (original sound field: green curve, simulated sound field: red curve)	Right ear (original sound field: green curve, simulated sound field: red curve)

Two listening examples are available for subjective comparison of the original and the simulated sound field.

 

	Artificial Head(binaural recording)
Original sound field	AH_original.wav*
Simulated sound field	AH_simulated.wav*

Some terminal test Examples

Similar recordings (original and simulated sound field as described above) are conducted with several types of terminals. As an example figure 5 and 6 show the power density spectra of the original and the simulated sound field transmitted via the GSM net with two mobile phones.

Figure 5 (Click on image to enlarge): Mobile phone 1 (with background noise signal processing): original sound field: green curve, simulated sound field: red curve	Figure 6 (Click on image to enlarge): Mobile phone 2 (without background noise signal processing): original sound field: green curve, simulated sound field: red curve

For both phones the spectra of original and simulated sound field match quite well. In addition to the average power density spectra figure 7 and 8 show the 3D power density spectra (versus time and versus frequency) of the original and the simulated sound field for mobile phone 1 (with background noise signal processing).

For subjective comparison, listening examples of original and simulated sound field are available for both mobile phones: