Hardware

Loudspeakers

Name Serial No. AAU No.
Brüel & Kjær OmniPower 4296 2251009 33950
Brüel & Kjær OmniSource 4295 2705455 86831
Custom-made 3'' directional loudspeaker 47 N/A

You can download an estimate of the on-axis impulse response of the 3'' directional loudspeaker. Note that the impulse response has has been measured in a 2pi/infinite baffle and has neither been time-aligned nor high-pass filtered. The impulse response is measured by Martin Olsen from Bang & Olufsen A/S.

Microphones

Up to 23 microphones are used. These microphones and microphone pre-amplifiers are all of the type

Instead of a microphone, one of the microphones pre-amplifiers is mounted with a dummy microphone of the type

  • Bruel & Kjaer JJ-2617 Coaxial input adapter (51 pF).

In the table below, the mappings between the microphones, the microphone pre-amplifiers, and the ADC input channels are listed. The connection column denotes the electrical connection between the microphones (which are in the listening room) and the ADCs (which are in the control room). The ADC channel notation X.Y should be read as ADC X ch. Y.  Finally, the label in the parenthesis after the ADC channel refers to the microphone name. For example, O1X refers to the first microphone in the x-direction of the orthogonal array whereas ULA 5B refers to the fifth microphone on the second uniform linear array. The label 'loop back' refers to the loop back connection from the loudspeaker input channel to an ADC input channel.

Connection Mic. Serial No. Pre-amp. Serial No.

ADC ch. Conf. XX0X

ADC Ch. conf. XX1X ADC Ch. conf. XX2X
1A 100230  78045  1.1 (O1X)  1.7 (ULA 1A)  1.1 (UCA 1A)
2A 100231  78170  1.2 (O2X)  1.6 (ULA 2A)  1.2 (UCA 2A)
3A 100248  78165  1.3 (O3X)  1.5 (ULA 3A)  1.3 (UCA 3A)
4A 100243  78168  1.4 (O4X)  1.4 (ULA 4A)  1.4 (UCA 4A)
5A 100244  78164  1.5 (O5X)  1.3 (ULA 5A)  1.5 (UCA 5A)
6A 100242  78049  1.6 (O6X)  1.2 (ULA 6A)  1.6 (UCA 6A)
7A 100236  78031  1.7 (O7X)  1.1 (ULA 7A)  N/A
8A 100246  78136  2.1 (O1Y)  2.1 (ULA 1B)  2.1 (UCA 1B)
9A 100240  78029  2.2 (O2Y)  2.2 (ULA 2B)  2.2 (UCA 2B)
10A 100229  78177  2.3 (O3Y)  2.3 (ULA 3B)  2.3 (UCA 3B)
11A 100232  78180  2.4 (O4Y)  2.4 (ULA 4B)  2.4 (UCA 4B)
12A 100233  78181  2.5 (O5Y)  2.5 (ULA 5B)  2.5 (UCA 5B)
1B 100261  78041  3.1 (O1Z)  2.6 (ULA 6B)  2.6 (UCA 6B)
2B 100265  78161  3.2 (O2Z)  2.7 (ULA 7B)  N/A
3B 100235  78163  3.3 (O3Z)  3.7 (ULA 1C)  3.7 (ULA 1C)
4B 100245  78171  3.4 (O4Z)  3.6 (ULA 2C)  3.6 (ULA 2C)
5B 100247  78169  3.5 (O5Z)  3.5 (ULA 3C)  3.5 (ULA 3C)
6B 100259  78162  3.6 (O6Z)  3.4 (ULA 4C)  3.4 (ULA 4C)
7B 100260  78174  3.7 (O7Z)  3.3 (ULA 5C)  3.3 (ULA 5C)
8B 100266  78186  2.6 (O6Y)  3.2 (ULA 6C)  3.2 (ULA 6C)
9B 100263  78172  2.7 (O7Y)  3.1 (ULA 7C)  3.1 (ULA 7C)
10B N/A  78167  2.8 (Dummy)  2.8 (Dummy)  2.8 (Dummy)
11B  100252  78183  3.8 (Single)  N/A  N/A
N/A N/A N/A 1.8 (Loop back)  1.8 (Loop back) 1.8 (Loop back)

Microphone Arrays

Four microphone arrays were used in the recordings.

  1. The simplest array was just a single microphone (Fig. 7).
  2. Three uniform linear arrays (ULA) were used (Fig. 5). Each ULA consisted of seven microphones which were separated by 5 cm. The microphones were indexed from 1 to 7 with the microphone farthest to the right (when viewed from the front) having index number one.
  3. Two uniform circular arrays (UCA) were used (Fig 6). The radii of these two UCAs were 4 cm and 6 cm, respectively. The microphones were indexed from one to six in the the counter-clock wise direction with the microphone at the three o'clock position (when viewed from the front) having index number one.
  4. An orthogonal array consisting of the three ULAs pointing in the x-, y-, and z-direction (Fig. 4).

Besides these microphone array, a dummy microphone was used (Fig. 8).

In the table below, it is depicted how the columns of the data matrix (.dataMatrix) of the struct returned by the loadSmarData.m function are mapped to different microphone and loop back signals:

Column no.

Mic. in conf. XX0X

Mic. in conf. XX1X Mic. in conf. XX2X
 1  O1X  ULA 1A  UCA 1A
 2  O2X  ULA 2A  UCA 2A
 3  O3X  ULA 3A  UCA 3A
 4  O4X  ULA 4A  UCA 4A
 5  O5X  ULA 5A  UCA 5A
 6  O6X  ULA 6A  UCA 6A
 7  O7X  ULA 7A  N/A
 8  Loop back  Loop back  Loop back)
 9  O1Y  ULA 1B  UCA 1B
 10  O2Y  ULA 2B  UCA 2B
 11  O3Y  ULA 3B  UCA 3B
 12  O4Y  ULA 4B  UCA 4B
 13  O5Y  ULA 5B  UCA 5B
 14  O6Y  ULA 6B  UCA 6B
 15  O7Y  ULA 7B  N/A
 16  Dummy  Dummy  Dummy
 17  O1Z  ULA 1C  ULA 1C
 18  O2Z  ULA 2C  ULA 2C
 19  O3Z  ULA 3C  ULA 3C
 20  O4Z  ULA 4C  ULA 4C
 21  O5Z  ULA 5C  ULA 5C
 22  O6Z  ULA 6C  ULA 6C
 23  O7Z  ULA 7C  ULA 7C
 24  Single  N/A  N/A

 

 

A/D Converters

The three A/D converters (ADC) used were all of the type

The first ADC in the table above has the AAU no. 56543, the second ADC has the AAU no. 56545, and the third ADC has the AAU no. 61403. The three ADC units were all isolated electrically via the B&O Power Supply RT12 with AAU number 08691.

D/A Converters

Besides four dual-channel ADCs, the Behringer Ultragain Pro-8 Digital ADA8000 also contain eight dual channel D/A converters (DAC). From the unit with AAU no. 56543, the first DAC channel was connected to the power amplifier whereas the second channel was connected to ADC channel 8 (the loop back signal in the tabel above).

Power Amplifier

Rotel RB-976 with AAU no. 33973. After a loudspeaker had been changed in the configuration setup, the volume was adjusted so that the sound pressure level (SPL) was approximately 80 dBA 1 m from the loudspeaker (on axis). The measured SPL is stored in the configuration files in the database and was measured using a Brüel & Kjær 2238 Mediator.

Sound Card

The sound card was of the type RME Digi 9652 Project Hammerfall with expansion board (no AAU no.) and connected to the ADCs via the optical ADAT interface. The sampling frequency of ADCs were controlled via the sound card's word clock.

 

Software

For making the recordings, we used the following software.

Omnisource 4296Fig 1. The B&K Omnisource 4296

OmniSource 4295Fig. 2. The B&K OmniSource 4295

Fig. 3. The 3'' directional loudspeaker

Orthogonal arrayFig. 4. The orthogonal array without foam. The measurements were made with foam (see the ULA in fig. 5)

Uniform linear array

Fig. 5. A uniform linear array.

Uniform circular array

Fig. 6. The uniform circular arrays

Single mic

Fig. 7. The single microphone

Dummy mic

Fig. 8. The dummy microphone