Post Decoder Algorithms
As the number of homes with 5.1 channel supportive home theater sets increases, it was realized that users would also like to experience their legacy audio sources (eg : tapes, CD, Video Tapes etc) in a surround environment. To this end, Dolby Laboratories and DTS have created matrix algorithms - stereo source to multi-channel output - to expand a two channel source to up to 6.1 channels. These algorithms (Dolby Laboratories' Dolby Prologic II and DTS's Neo6) also have a number of modes to vary the listening experience depending on the source stream. These include Music Modes, Movie Modes and various speaker configurations.
Research has shown that DVD sales have outgrown video tape players in one third of the time. Many original audio systems were based on a stereo configuration, and for these users, many home theater manufacturers and IP holders have created "virtualization" algorithms, which take a multi-channel source, reduce it to a two channel output, and process the phase information of the audio signal to give a surround experience to the listener from just two speakers.
To increase the "surround experience", the configuration of 5.1 speakers has been moving toward a configuration of 6.1 and 7.1 speakers. Dolby Laboratories, in partnership with LucasFilms created SurroundEX, and DTS has created DTS-ES. These matrix algorithms process the Surround Left and Surround Right channels to re-construct the Surround Back channel.
In the final output stage, THX, a patented technology from LucusFilms, is used to change the audio characteristics of the decoded multi-channel PCM to a home environment. As movie soundtracks are engineered for cinemas with unique environmental qualities, the same experience may not be reproduced in a smaller home. THX processing contains several processing methods to give a cinema quality sound to the home, but the main routines are listed as follows:
* Bass Management Electronic Crossover. This module takes the low frequency components of the main channels and re-directs them to the sub-woofer to allow use of smaller speakers for increased dynamic range
* Re-equalization. This module smoothes out the high frequency component of the audio to give a smoother sound.
* Timbre matching. A method to balance the audio levels between the front and rear speakers to deliver smoother sound movement
* Adaptive Decorrelation. Adds surround spaciousness to mono surround signals for a greater stereo feel.
* Bass Peak Level Manager. A module to prevent overload of the sub-woofer during large bass bursts.
The bass manager is one of the final stages in the audio signal processing. This module re-routes the main channels through various LPF, BPF, HPF to create each home theater vendors' "sound", and to control the frequencies that are sent to the speakers. Small speakers, that do not have a large dynamic range, will pass most of the low frequency components to the Sub woofer channel, whereas a large speaker setting will use a lower frequency cut - off HPF on the main channels as the speakers can handle a larger frequency range.
The delay library is used to match the position of the speakers to the position of the listener such that the sound from each speaker reaches the listener at the appropriate time. To create the true movie theater experience, the Bass and Delay settings are of extreme importance, and for high end systems, specialist are sent from the vendor to "calibrate" the system parameters to match the viewing environment. Up to recently, this calibration has been done by hand, but now Automatic Room Equalization techniques have been devised to bring the perfect cinema experience to the complete beginner.
Automatic Room Equalization.
Due to the differences in speaker frequency characteristics and room characteristics, the audio signals from multiple speaker sources will have different frequency components, and the multiple audio speaker outputs will not be time aligned at the listening point. To overcome this and maintain an optimal listening environment, the signal processor in the home theater application will carry out an automatic calibration routine called Room Equalization. The main steps involved in this equalization are: Frequency characteristics correction, Level correction, Delay correction.
Figure 8 : Effect of Automatic Room Equalization.
Figure 9: Automatic Room EQ process flow.
The signal processor will generate the calibration sound and pass the sound through each of the main speakers, one at a time. A special calibration microphone is placed in the listener's area and records the output from each speaker. The microphone output is sent back into the AVR and is re-converted to a digital signal via an analog to digital converter. This re-converted signal becomes an input to the signal processor. The signal processor then analyzes this input and varies the bass filter coefficients and delay parameters to achieve a uniform frequency response and time alignment.
Audio Signal Processor.
Audio processing involves intensive use of FIR and IIR filters. In recursive calculations, the quantization error due to the digital representation of the signals can cause a degradation of the audio quality. High-end audio processors, such as Analog Devices' SHARC processors, use a floating point representation for audio signals to reduce this error.
In the home theater experience the quality of the sound is generally measured by how accurately the low amplitude, or very quiet sounds, can be reproduced. As the amplitude of an audio signal gets smaller, the ability of a fixed point processor to accurately reproduce this signal is limited, but for floating point processors the accuracy with which the audio level can be maintained is contained within a fixed boundary, with a minimum SNR of 186dB.
Figure 10: SNR values for fixed and floating point processors.
Another important feature for a home theatre audio processor is the dynamic range. Dynamic range is defined as the ration of the minimum to the maximum signal amplitude that the audio processor can reproduce without underflow or overflow. Once again, floating point processors far exceed the limits of the fixed point processor.
Figure 11: Dynamic Range comparison for floating point and fixed point processors.
With the increasing complexity of pre-decoder algorithms and post decoder algorithms, the number of MIPS, or execution cycles needed to complete the many combinations needed for the home theater experience is forever on the increase.
To combat these issues, the obvious answer is to increase the clock frequency of the signal processor. Due to silicon process constraints, there are many obstacles to this method, which has lead signal processor vendors to approach the problem with architectural improvements. Some signal processor vendors have used a MIMD architecture approach, which involves executing multiple instructions in a single cycle while performing multiple data moves. The architecture requires more memory, which directly influences the chip cost. The SHARC processor architects took the novel approach of a SIMD architecture, in which the same instruction can be used to implicitly exercise a second parallel arithmetic unit. The result is that the code size is dense so there is a reduction in the MIPS requirement to perform the algorithms. Due to this SIMD architecture, the audio signal processor can operate on stereo signals in parallel without any extra processing overhead.
Figure 12: The SHARC SIGNAL PROCESSOR architecture, optimized for multi-channel audio processing.
As home theater systems must input encoded and non-encoded streams, output multi-channel audio, and carry out complex audio processing in real time, it is imperative that delays, which may inhibit the calculation process, are reduced to a minimum. In the SHARC processor, a dual-ported memory architecture has been implemented to overcome this hurdle. The I/O processor can read and write multi-channel audio samples from outside the processor to the internal memory without inhibiting internal memory read and write accesses by the core during audio processing. This architecture ensures that processing power of the signal processor is fully exercised, preventing detectable drops in the audio output.
Home Theater - the future
To further enhance the movie theater feeling of home theater, high-end AVR amps have already moved toward increasing the number of speakers supported. High-end systems now support a 10.2 speaker configuration, which includes left and right LFEs, multiple surround, and multiple surround back speakers. There is also a movement toward supporting multi-channel 192kHz high quality audio signals, and processing these signals with various post processing effects. Microsoft is in the process of releasing WMA-Pro, a multi-channel format to rival DTS and AC3, that is planned to support up to 7.1 discrete channels at a 96kHz sampling rate. Today, the DVD-Video standards only support 48k sampling streams, but with the introduction of blue ray technology and DVD-AR, which include encoding capabilities, we will see the emergence of even greater sample rates for increased quality, and a need for greater processing power in the audio signal processor.
As the home theater systems become more popular, the aesthetic beauty of the system has also come under the spot light. Large, 60kg boxes are not practical or pleasing for many homes. This has lead the engineering community to push technological frontiers in developing digital amps. Digital amps, in comparison to their analog counterparts, are smaller in size and require less power, resulting in a slim-line form factor for the home theater set. The disadvantage of the digital amp is that it has yet to reach the power amplification needed for larger high-end systems, and has had low acceptance by the audiophile community who has not embraced digital amps due to the lower quality response. It is envisioned that as technology moves ahead, most home theater systems will become fully digital systems, with the same quality as their analog counter-parts.
1 IEC 61937 : Digital Audio - Interface for non-linear PCM encoded audio bitstreams applying IEC 60958
2 White Paper :An Overview of the Coherent Acoustics Coding System, Mike Smyth, June 1999
3 ATSC standard : Digital Audio Compression (AC3), Revision A
4 ADSP-21161 SHARC Hardware reference manual, Analog Devices
Thanks to Jasmin Infotech in Bangalore India for all of their insight into audio processing and help with gathering of information.