Electrical audio signal processing systems and devices – Binaural and stereophonic – Pseudo stereophonic
Reexamination Certificate
2000-05-25
2004-09-21
Harvey, Minsun Oh (Department: 2644)
Electrical audio signal processing systems and devices
Binaural and stereophonic
Pseudo stereophonic
C381S001000, C381S061000, C381S307000, C381S310000
Reexamination Certificate
active
06795556
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to the reproduction of 3D-sound from plural-speaker audio systems or headphones, and, in particular, to the creation of Head-Related Transfer Functions (HRTFs) which are used for the synthesis of 3D audio signals. The synthesis of 3D audio signals has been described in a number of previous patent applications including U.S. Pat. No. 5,666,425, WO98/52382, and co-pending application GB 9805534.6. The latter contains a comprehensive description of how HRTFs are used in the synthesis of 3D sound, and is incorporated herein by reference.
2. Background of Related Art
One of the most important applications for 3D audio at present is 3D positional audio processing for computer games, and it is becoming increasingly important for other applications such as consumer electronics products (virtualisation of surround-sound via only two loudspeakers or headphones) and in the production of recorded music. In order to synthesise audio material bearing 3D-sound cues for the listener to perceive, the signals must be convolved with one or more appropriate HRTFs, and then delivered to the listener's ears in such a way that his or her own hearing processes do not interfere with the in-built 3D cues. This is achieved either by listening through headphones, or via loudspeakers in conjunction with a suitable transaural crosstalk-cancellation scheme.
If the listener is to perceive synthesised 3D-sound cues correctly, it is important that the synthesised cues are similar to their own natural ones, and so the HRTFs used for synthesis must be similar to the listener's own HRTFs. This is accomplished by creating an artificial head and ears having the dimensions of an average human adult, and making HRTF measurements on it. Our own method of achieving this has been fully described in WO 98/52382.
This method has been implemented and is very satisfactory for most users. However, there is a small proportion of listeners who have ear and head dimensions which are significantly different to the average, and for them the perceived 3D-sound effects can be spatially less accurate, or tonally incorrect. Accordingly, it would be advantageous to accommodate these individuals by creating HRTF sets which are based on head and ear dimensions closer to their own, such that they can choose to use this particular option, if it were provided, during a set-up option prior to using a 3D-audio system. The present invention relates to the creation of scaleable HRTF data, suitable for accommodating a wide range of physiological variation amongst listeners, and based on a single, average HRTF data set.
It is well-recognised that there exists significant physiological variation in the dimensions of the ears, head and neck, and that these influence any related acoustic measurements, including HRTF data. The use of artificial head systems has been employed in research and development for the optimisation of hearing-aid technology, in which microphones have been incorporated into rubber-type replicas of the human outer-ear and then built into an artificial head assembly. Some of these artificial head systems have also featured auditory canal simulators, together with neck and torso assemblies. In one well-known study by Burkhard and Sachs (“Anthropometric manikin for acoustic research”, M D Burkhardt and R M Sachs, J. Acoust. Soc. Am., July 1975, 58, (1), pp 214), ear dimension measurements were made on twelve male and twelve female volunteers, such that average dimensions of the various physiological features could be calculated. Next, the individual having ear dimensions which were closest to this average was identified, and then his ears were used as replication masters, from which copies were moulded in a flesh-like rubber compound (a mixture of two silicone rubbers to provide similar mechanical properties to flesh). This work produced a manikin and various ear types which are available from the Knowles Electronics Company (Knowles DB series product information (S-554-109), Knowles Electronics Inc., 1151 Maplewood Drive, Itasca, Ill. 60143, USA, under the trade name KEMAR (Knowles Electronics Manikin for Acoustic Research). According to the literature, there are four different ear types available. The original ears (DB-060/DB-061) are small and typical of American and European females, as well as Japanese males and females. Large ears (DB-065/DB-066) are more typical of American and European male pinna sizes. The DB-060/DB-061 and DB-065/DB-066 ears have been used extensively for tests of BTE (behind-the-ear) hearing instruments and for sound recording. Their ear canal openings are relatively small, which make them less suitable to use with ITE (in-the-ear) and ITC (in-the-canal( hearing instruments. The DB-090/DB-091 are large rubber ears with larger ear canal openings for use during the development of ITE and ITC hearing instruments. They permit the use of a common earmold so the hearing instrument can be quickly installed or removed. The DB-095/DB-096 are another variation of the large ears designed to be used with ITE and ITC hearing instruments which have user earmolds.
Details about the construction of the DB-065/DB-066 larger ear replica have been published (“Larger ear replica for KEMAR manikin R J Maxwell and M D Burkhardt, J. Acoust. Soc. Am., April 1979, 65, (4), pp. 1055-1058). The ear was based on one of the original sample of twenty four volunteers, and chosen to be two standard deviations larger than the average ear. For example, the original, average dimensions of various features of the standard ear (with the +2 standard deviation value in square brackets) were as follows. Concha length: 2.4 cm [2.9 cm]; concha volume: 4 cm
3
[5.7 cm
3
]; concha breadth: 1.7 cm [2.0 cm].
In principle, therefore, it would be possible to employ an artificial head system, such as the KEMAR, which is capable of mounting various ear replicas of differing sizes, for the measurement of a complete library set of HRTFs for each ear type. The listener could then choose which of these particular HRTF libraries to use, in order to obtain HRTFs which are the best match to their own. However, it must be appreciated that a typical HRTF library might contain more than 1,000 individual HRTFs (each containing both a left- and right-ear function, and an inter-aural time delay), and so it might take several weeks and much effort to carry out the acoustic measurements on a single ear type.
Also, it will be appreciated that if a number of ear-types were being measured sequentially, then experimental variations would inevitably occur between the measurements, leading to some imperfections in the matching of the inter-type data.
More importantly, there is physiological variation not only in ear size, but in head size, too. If this factor were also incorporated into the artificial head, it would render the measurements much more cumbersome and time consuming. For example, there would be nine measurement sessions needed for combinations of small, medium and large heads with small, medium and large ears. Although this is not impossible, it is the nature of technology that small incremental improvements occur frequently, and so if an improved average ear structure were to be developed, clearly it would be advantageous to make only one single set of measurements, rather than nine sets.
The use of differing HRTF libraries based on different shaped ears is known: in the early 1990s a commercially available binaural sound processor made by Crystal River Inc. offered several HRTF filter set options based on different ear types. There is also a headphone virtualiser for Dolby Pro-Logic audio material which offers the user a selection from 15 different HRTF types, as disclosed in WO 97/25834. This is apparently based on making a database from numerous measurements on a large number of volunteers, and then grouping the HRTF characteristics into 15 different categories, from which one typical HRTF type is selected by the listener.
As f
Clemow Richard David
Percy Michael
Sibbald Alastair
Creative Technologies LTD
Creative Technology Ltd.
Grier Laura A.
Harvey Minsun Oh
Swerdon Russell N.
LandOfFree
Method of modifying one or more original head related... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of modifying one or more original head related..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of modifying one or more original head related... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3216682