Disc Search

	Artist Search: Advanced Search

Published Articles - (Issue 16 of Widescreen Review)

	Back To Articles Main Page DTS' Universal Audio Platform Technology Coherent Acoustics® Disclosed Members of the DTS Coherent Acoustics® engineering and product development team presented its newest technology for consumer and professional applications at an event held on Wednesday, September 13, 1995 at the Alfred Hitchcock Theatre on the Universal Pictures studio lot. The formal presentation was limited to an audience from an invited guest list of 1200 representatives of the creative community, consumer electronics manufacturers and press. The presentation of the design and implementation of Coherent Acoustics was enlightening and compelling as the specifications were indicative of a remarketably flexible and capable audio platform technology which allows any playback system equipped with the Coherent Acoustics IC chip or algorithm to play encoded audio at fixed or variable bit rates ranging from 32 kilobits per second up to 4096 kilobits per second. The system is positioned to operate in a wide range of applications including DVD, LaserDisc, multitrack CD, digital VCR, direct broadcast satellite (DBS), digital radio, HDTV, video gaming and computer multimedia. Coherent Acoustics provides multiple modes of operation, including lossless coding (zero error) for the most demanding applications, reaching beyond CD-quality and even current professional PCM linear technology for a new standard of audio excelence possibilities. Before reporting on the actual specifications in outline form and our analysis the following is the transcript of the opening re-marks to the evening presentation by Terry Beard, President of DTS Technology. Thank you for coming this evening. The clip you have been watching is from Ron Howard's movie Apollo 13 with DTS digital sound. We now have DTS digital audio sound systems installed in 5028 theatres around the world and released 105 films in DTS digital sound as of today. DTS is the undisputed leader in theatrical digital sound. But we are here this evening to introduce our DTS consumer digital audio coding technology. DTS coding is unique in many technical ways, which will become evident as Stephen Smyth makes the technical presentation. However, I would like to highlight first the single most important and unique characteristic of DTS coding and that is that it can provide as high a quality encoding as the user chooses. It is capable of operating up to lossless coding; that is, it can produce a bit for bit reconstruction of the original digital audio coding. No quality compromise of any kind. Yet it can operate at rates as low as 32 kilobits per second per channel in applications where compromise is acceptable and it operates at these low rates with coding efficiency which surpasses other proposed coding methods at equivalent coding rates. As the coding rate increases the DTS coder evolves from a perceptual coder to a replica coder. A single DTS decoder chip is capable of automatically operating in any mode indicated by the incoming data stream. I believe this feature of DTS coding is its most unique and important capability. DTS, be-cause of its uncompromised quality capability, can satisfy the most critical demands of the creative community. It is truly an artist's medium. It does not compromise the art. It will deliver exactly what the artist creates. Quality compromise is not an inevitable feature of coding. In fact, this new technology can actually encourage even higher levels of quality and fidelity and innovation, and innovation in that form could open a whole new creative era in audio. The coding technology selected now, rather than dampening the enthusiasm of the creative community, can ignite it. That is the first reason for DTS. The second is to provide a technology that through its flexibility can avoid the tower of babble syndrome sure to develop in the absence of a single audio coding technology capable of operating across the full spectra of current and future media. It is our intention to make DTS technology universally available in chip and algorithm form to any and all users who request it. Basically DTS technology will be very inexpensive and hence applicable to all media. We have handed out an audio quality evaluation CD that can be played on any CD player to allow you to do your listening tests and technical evaluation of DTS operating over a range of encoding rates. We have tried to make this evaluation disc as fair, thorough, and complete as possible. Of course, since DTS coding can handle up to 24-bit digital audio the upper quality limit cannot be demonstrated on the CD because its quality simply exceeds that of the current CD standard. A number of examples encoded using Dolby AC-3 are also included on this CD for comparison purposes. All tracks were generated by passing original 16-bit linear PCM stereo signals through a real time encoder and de-coder operating at the various fixed bit rates. All signals remained in the digital domain throughout the recording production and mastering process. The prime objective of this evening is to insure that the competing coding technologies will be put before the professional and creative audio communities before we have a de facto standard set. Not withstanding the confident declarations to the contrary, the standards have not been set. Let's put the technologies out there for widespread unbiased evaluation before they are set. Frankly, I believe that will be the unavoidable consequence of the presentation this evening. At and subsequently thereafter our initial A-B listening evaluations conducted at our high end home theatre facilities I have experienced DTS Coherent Acoustics at the 128 kilobit and 240 kilobit per second per channel data rate to be a truly superior format for use in LaserDisc, DVD, and CD players, and I am convinced of its suitability for other futuristic digital audio applications now in development, such as digital video tape and multimedia virtual environments. The specifications described at this presentation certainly convinced me of DTS' commitment to establishing a new high fidelity audio standard. The demonstration at Universal Pictures was the result of a request on the part of members of the SD-DVD alliance who had participated in a demonstration of DTS Coherent Acoustics in Japan in late June (see Issue 15 July August 1995 "CD Digital Video Disc Format Wars&emdash;Part Three: The Nightmare On DVD Drive) to again demonstrate Coherent Acoustics before a panel of impartial listeners for the purposes of scientific evaluation in Los Angeles. While the intent was for an A-B evaluation of DTS Coherent Acoustics operating at 384 kilobits per second and Dolby AC-3 compared to original master recordings, this presentation fell short because no such comparison was conducted. That is not to say that DTS was not demonstrated to be superior in performance, flexibility, and capability to Dolby AC-3 even at comparable data rates, just that comparisons were not made at the 384 kilobit per second data rate be-cause no Dolby AC-3 encoder/decoder was provided. That event is still to come and is being supported by both this magazine, at least one other magazine, and a petition sponsored by the professional community (please see the preceding article "Selection and Testing of the Audio Coding Format for DVD"). DTS Technology has forthrightly agreed to participate and we will have to see whether Dolby Laboratories will or will not participate when the formal invitation is extended. The encoder/decoder used for the presentation and to produce the "Sound Quality Evaluation Disc" was a two-channel Dolby AC-3 digital audio encoder (Model DP523, Serial Number 500010, Software Revision 1.01/9441) and digital audio decoder (Model DP524, Serial Number 500013). These units were purchased by DTS in late June of 1995 with the data rate code configuration upgraded from 905 to 901, the latest version of Dolby's algorithm operating at bit rates of 64 kilobits per second mono and 96 kilobits per second per channel (192 kilobits per second stereo). Both the Coherent Acoustics encoder and the AC-3 encoder were operated as independent channel coders, each channel totally independent from the other and at a fixed rate of 96 kilobits per second. Although multichannel audio coding systems involve a number of important processes such as joint stereo coding, inter-channel bit pooling and inter-channel psychoacoustics, the key issue is still the raw numerical efficiency of the algorithm which codes and compresses the digital audio, i.e. its ability to reduce the bit rate without affecting the perceived quality in any given channel when converted back to analog. Hence, as stated in the CD pamphlet, the purpose of the evaluation CD is to demonstrate the coding efficiency of the Coherent Acoustics core technology and to compare where possible with the Dolby AC-3 core technology. Its purpose is not to facilitate final judgement on the performance of multichannel systems using either core technology since the efficiency of bit pooling, frequency joining etc. must be assessed on multichannel playback systems. As demonstrated on the CD and at the presentation the performance of DTS Coherent Acoustics exceeds that of Dolby AC-3 at lower data rates as well as higher data rates. The determination of the performance evaluation was derived from objective test signals and music material coded onto the test CD. That included a number of examples encoded using the Dolby AC-3 system for comparison purposes. All tracks were generated by passing the reference 16-bit linear PCM stereo signals through a real-time encoder and decoder operating at various fixed bit rates (64, 76, 96, 128, 192, 240 kilobits per second). All signals remained in the digital domain throughout the recording, production, and mastering process. The individual sounds are castanet, harpsichord, xylophone, glockenspiel, grand piano and female voice. There are three music selections as well, performed by Mary Black, Aaron Neville, and the Grateful Dead. The Mary Black selection was chosen to test the capability to reproduce a clear and pristine female voice, guitar and vocal harmony, the accurate capturing of the attack and decay of the overtone structure. "Everybody Plays The Fool" was chosen to test a codec's ability to faithfully reproduce the distinctive timbre of Aaron Neville's voice and capture the subtleties and drama of the male singing voice such as vibrato. In addition, this cut along with the Grateful Dead's "Touch Of Gray" tests an algorithm's ability to handle complex polyrhythms. "Touch Of Gray" also stresses a codec's ability to accurately reproduce large tone masses and distinctive chordal structures such as generated by rhythm guitars and pianos. Other material for evaluation of codec performance includes the difference signals between the original 16-bit PCM glockenspiel test signal and the Coherent Acoustics encoded version at bit rates from 64 to 240 kilobits per second per channel. At the 240 kilobits per second data rate the difference signal was virtually undetectable, thus resulting in a virtually transparent reproduction of the signal. The disc also contains multi-frequency test signals designed for FFT analysis of codec resolution characteristics and spectral response, and coherence test signals designed for x-y oscilloscope display of the coherence behavior of the codec. As stated previously the DTS Coherent Acoustics delivered as promised, and outperformed Dolby AC-3 at the comparable 96 kilobit per second per channel data rate which is the standard date rate for stereo on those platforms using Dolby AC-3, with 384 kilobits per second allocated for full 5.1 channel discrete output. What could be heard on the musical instrument and group selections with Dolby AC-3 was that the extended high frequency spectrum was absent, less air surrounded the instruments reducing the sense of spatial impact, midrange presence was characteristically hard sounding, transient attack was slightly smeared, bass extension was not as deep or defined, and significantly, imaging focus and specificity was veiled when compared to the original and to DTS Coherent Acoustics (both exhibiting virtually zero multi-tone phase error&emdash;see charts) which, while not absolutely transparent to the original either, did not share these attributes, at least less significantly, and was respectively nip and tuck to the original. Decidedly at 240 kilobits per second data rate there was no perceived difference between the original and DTS Coherent Acoustics. But critics will and should insist on an objective A-B between 384 kilobits per second Dolby AC-3 and DTS Coherent Acoustics on a high end matched discrete surround sound system. Then too, there is a mountain of comparative evidence that points to DTS' superiority at comparable Dolby AC-3 data rates and data rates significantly surpassing Dolby's coding capability. But then the evidence does not always make the case and politics and industry fraternalism often have the last say in the matter. To those critics of DTS that came away from the presentation feeling less than satisfied, their criticism should not be directed at DTS but Dolby, who has steadfastly refused to participate in or support every effort to date to cause an out-in-the-open A-B evaluation of the two competing codecs operating in real time. In the end our editorial position is to support "the best that it can be" while standard setting is in process. In our view media nullification of this issue, which has been the case with other publications, is not an act of journalistic responsibility while the opportunity exists to make a difference that could meaningfully advance high-end audio. As our readers know I and our editorial staff have experienced much frustration with DTS' rather weak effort to market its powerful Coherent Acoustics technology. The official DTS policy has been not to speak on behalf of any company it is presently doing, or is considering doing business with, nor to announce manufacturing relationships without the full participation of each new alliance partner. DTS also has refrained from promoting new technologies and products as "ready and available" to members of the stereo and home theatre community when this is not actually the case. Nor has there been vocal support from DTS' partners, which are MCA/Universal Studios and Stephen Spielberg. That was to be expected because until this moment in time, there was no DTS Coherent Acoustics "product." But now that Motorola has formally announced the availability of their DTS programmed DSP56009 Symphony™ 24-bit IC chip, DTS has now begun to enter into formal discussions with key hi-fi and home theatre component manufacturers in anticipation of ramping up to manufacture easily integrated Coherent Acoustics ASIC chips this Fall, leading to controller and processor decoder product introductions at the January 1996 Consumer Electronics Show (CES). Those companies expected to introduce processors or controllers incorporating a DTS mode include Enlightened Audio Designs (EAD), Perreaux, Audio Design Associates (ADA), Harman Kardon Citation (Model 7.5), Lexicon, Angstrom, Krell, Proceed, Kinergetics Research, Chiro, McIntosh, Meridian, McCormack Audio, Counterpoint, B&K Components, Audio Alchemy and Bose. No word whether the Japanese directed companies will be at CES with DTS featured product with the exception that Denon has previously expressed an interest in the DTS technology. The following is the published DTS Coherent Acoustics specifications released on September 13, 1995. DTS Coherent Acoustics Specifications DTS Technology Objectives: Facilitate major advancements in home audio reproduction in terms of fidelity and stage imagery Universal audio platform for many applications&emdash;present and future Provide future-proof decoding platform Provide cost effective decoding technology Utilize phase coherent coding methodologies Provide real-time encoding Coherent Acoustics Specifications: 1 to 8 channels of multiplexed audio Sample rates from 8kHz to 192kHz 16- to 24-bit word length (138 dB dynamic range) Compression ratios from 1:1 to 40:1 Total data rate operating range 32 kbps to 4096 kbps Lossless coding mode (zero error) Fixed or variable rate coding Down mixing from n channels to n-1, n-2 channels etc. Down mixing from 5.1 discrete to Left Total (Lt) / Right Total (Rt) Embedded dynamic range control Channel re-equalization Accurate synchronization of audio to video Embedded time stamp and user data Motorola Single Chip Decoder Solution: Operates from 32 kbps to 4096 kbps Decodes up to 6 channels Low frequency effects in any mode Lossless coding mode 20-bit PCM resolution at outputs Requires single 32K x 8 SRAM and 1M EPROM Controlled via Motorola host serial port Coherent Acoustics Applications: Discrete multichannel LaserDisc using PCM track Discrete multichannel CD using PCM track Discrete multichannel DVD at 192, 384 or 768 kbps Discrete multichannel DVD audio at 2-4 mbps at 96kHz sampling Discrete multichannel in-car entertainment Discrete surround for multi-media PC and Mac audio systems Audio networking (Internet) Electronic distribution ISDN, ATM, Fibre, Satellite For additional detailed technical information, contact DTS Technology at 818 706 3525 or on the Internet at: www.dtsonline.com. Back To Articles Main Page