mskuhn@faui09.informatik.uni-erlangen.de (Markus Kuhn) (05/03/91)
Some days ago, I reported in this group about a sound compression method developped at the University of Erlangen, Germany. I got a lot of requests for literature, so I went to one of the local scientists (Rolf Kapust). He gave me a recent publication (unfortunately only in German): Karlheinz Brandenburg, Bernhard Grill, Horst Jonuscheit, Rolf Kapust, Dieter Seitzer, Tomas Sporer: Uebertragung von hochwertigen Tonsignalen mit Datenraten im Bereich 64 bis 144 kbits/s, Rundfunktechnische Mitteilungen, Jahrgang 33, Heft 5, 1989. But the article has an English summary: Transmission of high quality audio signals with bitrates in the range 64 to 144 kbits/s Methods enabling bitrate reduction before transmission and/or the recording of high quality audio signals up to factor 7 without loss of sound quality and up to factor 11 with a slight loss of quality are presented. More economic transmission or recording methods than those currently used are employed for this effect. The electrotechnical faculty of the University of Erlangen-Nueremberg has implemented some modern methods of bitrate reduction based on psychoacoustic considerations, which exploit the properties of the human ear. The error signal that results obtained with these methods are described. The possibilities of realizing these methods by means of digital signal processors are particularly emphasized. The faculty has -- as far as I know -- a patent on this method. Especially interesting in the above article is a table comparing several compression techniques: Name bits/sample sampling data comments transmitted rate rate [kHz] [kbits/s] CD 16 44.1 705.6 just a reference NICAM 10.1 32 324 simple method LC-ATC 3 48 144 1 chip solution OCF 2.5 48 120 quality better than CD! 2 44.1 88.2 no difference to CD! 1.45 44.1 64 slight differences to CD The article contains 15 references. Some of them are: Zelinski, R.; Noll, P.: Adaptive transform coding of speech signals, IEEE Trans. on Acoustics, Speech and Signal Processing, Vol. ASSP-25 (1977), p. 299--309. Brandenburg, K.: OCF - A new coding algorithm for high quality sound signals. Proc. 1987 Int. Conf on Acoustics, Speech and Signal Processing (ICASSP), p. 141--144. Brandenburg, K.; Kapust, R. et al.: Fast signal processor encodes 48kHz/16bit audio into 3bit in real time. Proc. from ICASSP, New York 1988. I am not an expert in this field. If you are seriously interested in this topic, you should contact these people. I don't know whether they have an email address. There is no PD source code available. Have fun ... Markus -- Markus Kuhn, Computer Science student -- University of Erlangen, Germany E-mail: G=Markus;S=Kuhn;OU1=rrze;OU2=cnve;P=uni-erlangen;A=dbp;C=de
mskuhn@faui09.informatik.uni-erlangen.de (Markus Kuhn) (05/03/91)
mskuhn@faui09.informatik.uni-erlangen.de (Markus Kuhn) writes: [...] > methods of bitrate reduction based on psychoacoustic considerations, > which exploit the properties of the human ear. The error signal > that results obtained with these methods are described. The > possibilities of realizing these methods by means of digital > signal processors are particularly emphasized. Sorry, I mixed up the lines, as I copied this from the article. It has to be [...] methods of bitrate reduction based on psychoacoustic considerations, which exploit the properties of the human ear. The error signal that results from these techniques has a spectral colouration that remains under the threshold of perceptibility. The results obtained with these methods are described. The possibilities of realizing these methods by means of digital signal processors are particularly emphasized. Markus -- Markus Kuhn, Computer Science student -- University of Erlangen, Germany E-mail: G=Markus;S=Kuhn;OU1=rrze;OU2=cnve;P=uni-erlangen;A=dbp;C=de
richard@ear.mit.edu (Richard Kim) (05/04/91)
In article <mskuhn.673269191@faui09>mskuhn@faui09.informatik.uni-erlangen.de (Markus Kuhn) writes: > Some days ago, I reported in this group about a sound compression > method developped at the University of Erlangen, Germany. > I got a lot of requests for literature, so I went to one > of the local scientists (Rolf Kapust). He gave me a recent > publication (unfortunately only in German): Since there seems to be much interest in audio coding, I decided to gather some of my references and post to this news group. The following are some of the coders that I'm aware of. This is not an exhaustive list by any means. The references are listed at the end. o OCF (Optimum Coding in the Frequency domain) is a transform coder. This performs a modified discrete cosine transform (MDCT) and an entropy coding of the coefficients. The designers are from Erlanger in Germany as was mentioned earlier. o Dolby's coder is a transform coder. They actually have encoder/decoder systems as products. o AT&T people published several papers on their version of the transform coder. o Scientific Atlanta also demonstrated their coder a year ago at National Association of Broadcaster's convention in Atlanta. I don't know of any published article on this coder. o MUSICAM (Masking-pattern Universal Sub-band Integrated Coding And Multiplexing) is a sub-band coder with 32 sub-bands each with 750 Hz bandwidth. This system is a joint effort from three institutions: 1. CCETT (Centre Commun d'Etrudes de Telediffusion et Telecommunications) 2. IRT (Instut fur Rundfunktechnik) 3. Philips, The Netherlands This system was part of a proposal on Digital Audio Broadcasting by the European Broadcasting Union. o Audio Processing Techonology, (a subsidiary of Solid State Logic) Oxford, England This company offers a sub-band coder. They have a demo CD disc of original and coded/decoded music. I don't know of any publications on this one either. o OCF references: This is the latest report that I know of. %A E. Eberlein %A H. Gerhauser %A S. Drageloh %T Audio codec for 64 kbit/sec (ISDN channel) -- requirements and results %B ICASSP (IEEE Int. Conf. on Acoustics Speech and Signal Processing) %D 1990 %P 1105 This is the earliest paper that describes OCF. %A K.H. Brandenburg %T OCF - A new coding algorithm for high quality sound signals %B ICASSP %D 1987 %P 141-144 Other articles on OCF are: %A K.H. Brandenburg %T Evaluation of quality for audio encoding at low bit rates %B AES (Audio Engineers Society) Convention 82nd %D 1987 Mar. %P Preprint 2433 %A K.H. Brandenburg %T Low bit rate codecs for audio signals implementation in real time %J AES Convention 85th %D 1988 Nov. %A K.H. Brandenburg %T High quality sound coding at 2.5 bits/sample %B AES Convention 84th %D 1988 %P Prep. 2582 o Dolby references %A G. Davidson %A L.D. Fielder %A M. Antill %T High-Quality Audio Transform Coding at 128 kbits/s %B ICASSP %D 1990 o AT&T references %A J.D. Johnston %T Transform coding of audio signals using perceptual noise criteria %D 1988 Feb. %V 6 %P 314-323 %J IEEE J. Sel. Areas Comm. %A J.D. Johnston %T Perceptual transfrom coding of wideband stereo signals %B ICASSP %D 1989 %P 1993 o MUSICAM references The latest report on MUSICAM that I'm aware of is: %A G. Stoll %A Yves-Francois Dehery %T MUSICAM: High quality audio bit-rate reduction system family for different applications %D 1989 Oct. %O Presented during the CCIR Final Study Group Meetings, Geneva Earliar ones are: %A G. Theile %A M. Link %A G. Stoll %T Low-bit rate coding of high quality audio signals %B AES Convention 82nd %D 1987 March %P Prep. 2432 %A G. Theile %A M. Link %T Low bit-rate coding of high-quality audio signals -- An introduction to the MASCAM system %B Advanced digital techniques for UHF satellite sound broadcasting %I EBU (European Broadcasting Union) %D 1988 Aug. %P 71-94 -- ------------------------------------------ Richard Y. Kim ryk@athena.mit.edu (617) 253-8142 (W), (617) 449-7347 (H)