A short and focused description : (PLEASE NOTE : all rights reserved)
This project (first introduced at the University of Edinburgh in 2007) explores the psycho-acoustics of human’s perception over the audio spectrum. The purpose is to propose an alternative sound reproduction system that allows a more flexible and innovative sound design. The final outcome is an improved and expandable experience of the original film soundtrack.
// Some relevant theory
Sound, free from reflected areas, mostly refers to natural sound, thus monophonic sources. In such environments the differences in time that the same waveform needs to reach the listener’s ears gives to the brain all the information needed to calculate the location of the source. This means that two separate detectors perceive a single source. “This ability of human spatial hearing is determined by the term ‘binaural delay’.” “Binaural delay is the time difference that same sound source needs to reach each of the listener’s ear.” (Rumsey F. 2001)
In addition, the frequency response of the ear is different from various angles. For example, low frequencies have been proved to correspond to the frontal perception, while higher frequencies seem to be perceived mostly from the rear of the head’s frontal position. (Rumsey F. 2001). Moving sources, like a car on the street, change their spectral characteristics in relation to the listeners, which are received and calculated by the auditory system. The function reminds of the Doppler effect, where the variable distance of the moving source changes the sound’s wavelength and frequency. Therefore, “monaural” localization seems to be affected by the filtered outcome occurred by external factors. (Blauert J. 1996)
Alton Everest describes the Pinna, the outer part of human ear, as a crucial element, which provides information about the directionality of a sound source. That is also responsibly for the differences in perception angles over the sound spectrum. However, because of the differences of the individual, it is difficult to calculate and conclude on standard results. (Everest A. 2001)
Precedence Effect (Haas Effect)
When experiencing stereophonic sound reproduction the auditory mechanism works quite differently. Stereo reproduction creates individual binaural delay to each of the listener’s ear. The difference in time and sound level pressure leads the spatial experience to depend on the first waveform received. (Rumsey. F. 2001)
Alton Everest adds to the above phenomenon that “human perception is able to merger all the reflections of the direct source, arriving within 50 msec and combines them with direct signal.” (Everest A. 2001) That gives the illusion that all these sound elements are being projected from the same position. The above spatial attribute it is also called “The law of the first waveform” a term coined by Cremer (1948). (Everest A. 2001)
Haas Helmut (1949) has concluded that reflections between 5-35 msec has to be 10dB higher than the direct waveform in order to be perceived as an echo. “Otherwise the result within this bandwidth of time, the sound seems louder and pleasant. Sound reflections beyond 50 msec will occur in discrete echo effects.“ (Haas H. 1949)
The reflection can be calculated with the following equation:
RF=RP-DP / (C/sec)
Were: RF= Reflection Delay, RP= Reflection path in feet or meters, DP= Direct path in feet or meters, C = the speed of sound (344/sec on meters and 1,130/sec in feet)
While the level of the reflected amplitude, assuming that is 100% reflected, can be found as followed:
Were: RL = Reflection level at listening position, DD = direct distance and RD = reflection distance
The latter equation uses the “Law of Inverse Square”, which is briefly a reduce in sound pressure level by 6dBSPL every time the distance from the source is doubled. This is true for free sound field but not for indoors sound. Reflections apart from providing information for space, also add loudness to the direct signal. However, using the above calculations, an acceptable arrangement can be implement, in order to avoid the phenomenon of a reverberant sound as much as possible. (Everest A. 2001)
// The System
A specific system of coordinates is used to describe the spatial perception over the different angles of the head.
// FIGURE – Head-Related coordinates (Spatial Hearing, Jens Blauert pg.14)
The horizontal plane (φ) provides information about the azimuth. The median plane is the virtual line that is symmetrical in relation to the head. The elevation (δ) and azimuth are used together to provide specific coordinates of sound events. The distance (r) can be measured both in feet and meters. (Blauert. J 1996)
Therefore, the most relative to this project question would be: How is the auditory system response over the median plane?
Blauert (1996) uses the term “Localization Blur”, which measures the smallest possible changes on a sound event that can affect the spatial perception. Human perception seems to have a maximum value of localization blur at the forward position (azimuth 00). Blauert also states that the minimum angle is at 10 for a change to be noticed in the horizontal plane.
Localization at the median plane is different; at the minimum angle a signal is being set at 40 (δ=40, φ=00) for white noise, while for an unfamiliar voice sample the localization blur can reach an amount of elevation at 170. However, based on an experiment where five loudspeakers were placed on the upper half of the median plane, the spatial perception can be improved by the familiarity and the passages of time of the signal. (Plenge and Brunschem 1971)
Pratt has experimented with the spectrum quality of the sound and has concluded that events with rich high-end register are being located more accurately than the “low tones” over the elevation angle. (Praat C. 1930)
On a later example he confirms that the auditory event in the median plane is related to the spectrum content more than any other factor. Roffler and Bulter concluded to similar results stating that localization blur can be most successful from the differences in the sound spectrum content. (Roffler and Bulter 1968, cited in Blauert)
// FIGURE – Results of Roffler and Bulter experiment 1968
(Spatial Hearing, Jens Blauert pg.106)
As a conclusion to this section it seems that the auditory perception in the horizontal plane depends mostly on the binaural delay and amplitude. Moreover, we may conclude that our perception in the median plane appears to be more effective over the spectrum filtering and the resonant of the auditory system.
// More than one sense ?
Audio and Visual
Audio within image projection seems to be quite different, yet confusing. Klemm (1909) has formulated the “Law of Spatial Complication”. Assuming that sound and image are parallel in time progression, according to his statement, more that one sense, such as auditory and visual events can occur to spatial perception.
Michael Chion seems to share Klemm’s law “When kinetic sensations organized into art are transmitted through a single sensory channel, through this single channel they can convey all the other senses at once.” (Chion M. 1994)
Ewest (1930) has presented an exception, by distorting the visual content, he found that the sound localization could be taught and learned so to response to the auditory event. Blauert (1970) has proved with the method of masking and multiple sound sources that the spatial perception of sound does not depend on the image. Based on Gottesberge (1940), the sound event has to be combined with an eye movement in order to achieve the most successful perception.
However, it seems that this attribute of the brain to react on the sound localization over the image has given the opportunity to the filmmakers to use “off” and “on” screen sound more effectively. “Sight appears to have the priority against hearing, when it comes to space definition and sound/image movement.” (Sonnenschein D. 2001)
Proposed Alternative sound configuration (Quad-Wall)
Set up the “Quad-Wall”
The configuration adds to the existing standard of surround sound configuration (5.1) two extra loudspeakers. The main left right and centre units should be at the same horizontal line. The Dolby Surround Mixing Manual as the best solution for critical listening suggests that the tweeters of the loudspeakers should be placed in the same high, as possible, to the listeners’ ear. It has the compromise that it is not suitable for large audiences, as the physical body will mask the high frequencies. However this makes the comparison of the proposed system more direct against the standard stereo and 5.1 configurations.
The extra two loudspeakers should be placed at the top of the screen as shown in the figure. If the installation permits, the top loudspeakers should be angled, so that they point at the listener’s ear.
A sound pressure level meter is needed to make sure that all the speakers produce the same energy at the listening position. (Dolby Lab “Surround Mixing Manual”) This configuration has been studied and tested at a typical projected screen. However, it can be possible to deliver this concept to large movie theatres with big screens, but with a different set up.
“Quad-Wall” in Movie Theatres
Normally in big cinema rooms small “fill” loudspeakers are used to cover more space (fig). This happens because of the high frequency directionality behaviour and of the individual loudspeaker specifications. (Gary D. & Ralph J. 1989)
// FIGURE – A typical large room with fill units. (modded img, Gary D. & Ralph J. 1989)
More speakers are needed to cover the depth of the room but less for the width of the room. This means that sometimes and depending on the physical space the top units can be mounted horizontally, in order to cover better the area over the horizontal line. In order to transfer the standard surround soundtrack from the critical listening environments to the large theatres, Dolby Lab’s Surround Mixing Manual suggests that more units can be added in series to the left and right surround loudspeakers.
In the same way more units can be added to the proposed extra channels. The loudspeakers can be mounted at the ceiling, with the option of the vertical or the horizontal mounting depending on the individual needs of the movie theatre. As shown in the figure 10 the loudspeakers can be smaller than the previous example (fig) that uses main clusters and fills units.
// FIGURE – (Dolby Lab Surround Mixing Manual 2005)
However, if the depth of the theatre is large enough, time-delay (also known as delay-line) it might be really useful to retain the synchronization of the sound and image. This is automatically equalized for the added units of the figure 9 by the hardware Dolby surround devices. (Dolby Lab’s Surround Mixing Manual 2005)
// FIGURE – More units added to expand the existing two extra channels.
The audience that refer to the source (b) wont hear the sound coming for the source (a). Thus, phase problems would be eliminated, because the human ear has the attribute to concentrate on the louder signals and musk the lower ones. (Katz B. 2003) This can be calculated effectively by the individual loudspeaker’s polar response specification and the inverse square equation. (Gary D. & Ralph J. 1989)
Sound Design Benefits
The task was to create a way to expand the creative use of the films’ soundtracks and the spectator’s experience. Therefore, one question that needs to be answered is which are the mains benefits of this system against the others. The sound can be used, within the Quad-Wall, without being affected or dependent on the recording method. Against popular formats like Ambisonics, which can offer a similar experience, but in order to make complete use out of this arrangement, a specified microphone (see SoundField) technique has to be used. (Elen G. R. 1998) On the contrary, the sound can be manipulated and diffused, being only limited by the needs of each scene and the aesthetic approach of the designer.
// Some reading
Birker C. (2004), “Practical Recording-Surround Sound”, London, SMT productions
Blauert J. (1996), “Spatial Hearing – The Psychoacoustics of Human Sound Localization”, Translated by John S. Allen, London, MIT Press
Chion, M. (1994) “Audio-Vision – Sound on screen” (Trans. G. Gorbman), New York, Columbia University Press.
Dolby Lab (1999) ”Past, Present, and Future. A history of multichannel audio from mag stripe to Dolby Digital”. (http://www.dolby.com/assets/pdf/tech_library/2_Surround_Past.Present.pdf)
Dolby Lab, (No Date), “Evolution Of Sound”. (http://www.dolby.com/assets/pdf/tech_library/53_EvolutionOfSound.pdf)
Dolby Lab, (Press Release 2002), “Sonic Whole Overhead Sound”. (http://www.dolby.com/assets/pdf/press_releases/673_mp_pr_0209_Soundelux.pdf)
Dolby Lab, (2005), “Surround Mixing Manual”. (www.dolby.com/assets/pdf/tech_library/44_SuroundMixing.pdf)
Elen G. Richard (1998), “Ambisonic Surround Sound in the Age of DVD” (http://www.ambisonic.net/ambidvd.html)
Everest A. (2001), “The Master Handbook of Acoustics” (4th edition). New York, McGraw-Hill Education.
Everest A. and Streicher R. (1992), “The new Stereo Soundbook” USA. TAB Books – McGraw Hill.
Gardner G.W. (1998), “3-D Audio using loudspeakers”. London, Kluwer Academic Publiser
Gary D. & Ralph J. (1989) “ Sound Reinforcement Handbook”. USA. Hal Leonard Corporation.
Glasgal R. (2005), “Improving 5.1 and Stereophonic Mastering/Monitoring by Using Ambiophonics Techniques”. (http://ambiophonics.org/papers/RG-Improving_TonSym0510.pdf)
Haas H. (1972) “The Influence’ of a Single Echo on the Audibility of Speech” Audio Engineering Society, (English translation by Dr. Ingr. K.P.R. Ehrenberg of Haas’ original paper in Acustica (1951)
Holman T. (2004), “5,1 Surround Sound and Running”, USA, Focal Press
Katz B. (2003), “Mastering Audio – the art and the science”, USA. Focal Press
Kenny. T. (2000).”Sound for Picture – The art of sound design in film and television”. Canada, Mix Pro Audio Series.
Ostwald, P.F. (1963). “Soundmaking, The Acoustic Communication of Emotion”. New York: Charles C Thomas.
Schoenherr. E. S. (2001), “Stereophonic Sound”, Recording Technology History, (http://history.acusd.edu/gen/recording/stereo.html)
Sonnenschein. D. (2001) “Sound Design – The power of music, voice and sound effects in Cinema”, USA, Michael Wiese Productions.
Sergi G. (2004), “The Dolby Era: Film Sound in Contemporary Hollywood”. Manchester University Press Melland Schill Studies
Sider L. (2003), “Soundscape: School of Sound Lectures 1998-2001”. Wallflower Press
Rumsey F. (2001), “Spatial Audio”. USA. Focal Press
Weis E. & Belton, J. (1985) “Film Sound: Theory and Practice” New York, Columbia University Press.