Sound Utilities 2.0: A Software System for Creating Virtual Acoustic SpacesSkip other details (including permanent urls, DOI, citation information)
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. Please contact email@example.com to use this work in a way not covered by the license. :
For more information, read Michigan Publishing's access and usage policy.
Page 477 ï~~Sound Utilities 2.0: A Software System for Creating Virtual Acoustic Spaces Dale Stammen McGill University 555 Sherbrooke Street West Montreal, Quebec, Canada H3A 1E3 firstname.lastname@example.org Gilbert Soulodre McGill University 555 Sherbrooke Street West Montreal, Quebec, Canada H3A 1E3 soulodre@ music.mcgill.ca Abstract We have enhanced the software for our computer-based binaural recording and playback system for the reproduction of virtual concert halls [Soulodre and Stammen, 1993] by adding several new tools that allow the user to modify existing impulse responses and create unique virtual acoustic spaces. Existing impulse responses may be edited or unique impulse responses may be created with the Impulse Response Editor. This editor allows the user to combine specific portions of various impulse responses. The amplitude of any portion of the impulse response may be scaled with the Amplitude Editor. Virtual reflections may be added to an impulse response using the Reflection Editor. The angle, intensity, and location of each reflection is easily adjusted using a graphic display. The software can analyze any impulse response to determine the standard acoustic parameters of RT60, clarity, early decay time, lateral energy, interaural cross-correlation, etc. All parameters are measured on an octave band basis. The newly designed impulse response may then be convolved with anechoically recorded music, electronic music or noise and played through our binaurally accurate reproduction system. 1 Introduction We have developed a software package called Sound Utilities [Soulodre and Stammen, 1993] which enables a user to edit existing impulse responses and create unique virtual acoustic spaces. The software is currently running on a Macintosh Quadra 840AV computer and uses the DSP56001 on a Digidesign Audiomedia II card for audio processing and playback. The software accepts impulse responses in RAMSoft [Halliwell and Bradley, 1993] and Maximum Length Sequence System Analyzer (MLSSA) formats and outputs binaural sound files in Sound Designer II (SDII) format. The current version of the software may be divided into several modules, each of which is discussed below. 2 The Impulse Response Editor After an impulse response (IR) has been measured by the RAMSoft or MLSSA systems, it may be displayed and edited using the Impulse Response Editor (IRE). The IRE offers several display modes including a waveform display, an Amplitude Editor, a Reflection Editor, and a FF1 power spectrum display. The IRE uses two screen cursors that allow the user to select regions for analysis or editing. The display may also be zoomed to view the IR at various resolutions. The IRE permits regions of an IR to be cut and pasted to create new hybrid IRs. Crossfades may be applied between these regions to smooth out any discontinuities between the regions. The IRE can be used to truncate an IR so that the user may listen to only the early portion of a sound field in isolation. Regions of the IR may also be viewed in the frequency domain using the FF power spectrum display. 2.1 The Amplitude Editor The Amplitude Editor (AE) presents the user with a rectified waveform display of the amplitude of an impulse response as shown in Figure 1. The AE enables the user to select a region of the IR with the screen cursors and apply a scaling function to that region. The AE may also be used to apply amplitude scaling to a specific frequency bands thereby allowing control over the reverberation time at various frequencies. dBA LSJ I *ms Figure 1: Amplitude Editor 2.2 The Reflection Editor Virtual reflections may be added to an impulse response using the Reflection Editor (RE). The RE presents the user with a time line to which reflections may be added. The angle, intensity, and delay time of each reflection is easily adjusted using a graphic display similar to the one shown in Figure 2. In the RE display, each vertical line represents a reflection, with the height indicating its intensity in dB. The direction of each reflection may be selected from a library of impulse responses obtained from binaural measurements in an anechoic chamber. We are currently adding to our reflection library a complete set of HRTF measurements of a KEMAR dummy head [Gardner and Martin, 19941. These ICMC Proceedings 1994 477 Acoustics
Page 478 ï~~measurements will allow the user to select reflections at any elevation from -40 and +90 degrees. The Reflection Editor can be used to create simplified sound fields to conduct studies into the properties of early reflections in a manner similar to that done in anechoic chambers. A powerful feature of the RE is that a given set of reflections may be superimposed onto an actual binaural impulse response from a concert hall. This allows the user to add individual reflections to an existing binaural impulse response so that the effects of these reflections may be systematically investigated. dB Ii +30Â~ -120Â~ -45" 0 " o ms Figure 2: Reflection Editor 2.3 Acoustic Measures The Sound Utilities program calculates the standard acoustic measures on IR's obtained from RAMSoft or MLSSA. The program can determine C80 or Cn, EDT, RT60, LF, G and TS on any monaural IR, IACC from any binaural IR, and LF from any IR obtained using a combination of a figure 8 and omni-directional microphones. These acoustic measures are calculated and displayed on an octave band basis. 2.4 Signal Processing After an IR has been prepared by the IRE, it can be auditioned by convolving it with a sound source. The source material (anechoic music or noise) is preprocessed by convolving the source with the inverse of the diffuse field response of the dummy head used to obtain the binaural IR's. The source material is then convolved with the left and right channels of the binaural impulse response. Finally, cross-talk cancellation and speaker frequency response corrections are added to the output. Since this signal processing stage is very time consuming, a batch processing mode is provided. 3 Binaural Playback System The resulting binaural sound files may be auditioned with our binaural playback system (See Figure 3). We have designed a loudspeaker enclosure system to reduce the acoustic cross-talk inherent in loudspeaker playback [Soulodre and Stammen, 1994]. In this system, crosstalk cancellation is done in Lwo stages. The speaker enclosures provide a mechanical barrier which limits the amount of energy from the loudspeakers that diffracts around the head. The second stage uses the approach used by [Schroeder and Atal, 1963] and is accomplished during the signal processing stage described above. Recent measurements of the system have shown that the mechanical barriers provide approximately 20 dB of broadband cross-talk cancellation while the signal processing provides an additional 10 dB. The extremely close proximity of the loudspeakers to the listener eliminates the need for an anechoic environment. Figure 3: Binaural Playback System 4 Summary A desktop computer-based binaural system has been described. The software, Sound Utilities, permits extensive editing and analysis of impulse responses. The unique loudspeaker enclosures provide cross-talk cancellation and obviate the need for an anechoic listening environment, thus making the system extremely portable. 5 Acknowledgments We are grateful to the Concert Hall Research Group for their support of our research. References [Gardner and Martin, 1994] Bill Gardner and Keith Martin. HRTF Measurements of a KEMAR DummyHead Microphone. MIT Media Lab Perceptual Computing Technical Report #280, May, 1994. [Halliwell and Bradley, 1993] R. E. Halliwell and J. S. Bradley. RAMSoft II: A Computer-based Room Acoustics Measurement System. Journal of the Acoustical Society of America, 89(4), p. 1897. [Schroeder and Atal, 1963] M. R. Schroeder and B. S. Atal. Computer simulation of sound transmission in rooms. IEEE Int. Cony. Rec. 7, pp.150-155, 1963. [Soulodre and Stammen, 1993] Gilbert A. Soulodre and Dale R. Stammen. A Binaural Recording and Playback System for the Reproduction of Virtual Concert Halls. Proceedings of the 1993 International Computer Music Conference, 1993. [Soulodre and Stammen, 1994] Gilbert A. Soulodre and Dale R. Stammen. A Binaural Simulator for Conducting Subjective Studies of Concert Hall Acoustics. Presented at the 126th meeting of the Acoustical Society of America, Boston, 1994. Acoustics 478 ICMC Proceedings 1994