YAMAHA AND HOME THEATER

Furthermore, Yamaha products are elaborated in very different conditions than other AV products. Yamaha is the world's largest maker of musical instruments using digital techniques, they also design concert halls all over the world. As a result Yamaha's broad background in all these areas has played an important role in the design of DSP technologies applied to Home Entertainment. They do really know how to measure and reproduce with high quality sound the acoustic soundfields of actual places such as jazz clubs or concert halls.

Yamaha rendering sound systems have nothing to see with "DSP processors" of any other customers. The soundfield created is much more complex and the results are by far much more realistic. Even THX with its enhancements may not reach the level of space and liveness of Cinema DSP concerning the simulation of movie theater sounfields.

HOW DOES CINEMA DSP WORK ?

You need to understand what is a soundfield to realize the Yamaha Cinema DSP process. Basically a soundfield is a mix of sounds generated by a places having typical reflections or sounding capacities. Usually the parameters describing the origins of a soundfield depends upon frequency. The resulting soundfield in a room depends on characteristics such as geometry, wall texture,... If a signal is emitted in the room you'll hear a soundfield depending on all these parameters but initialy all the soundfield was created by the single signal you generated. This is known as signal correlation. If you change the frequency of the original signal, the soundfield will be different since resonnance effects may occur at this frequency. If you want to recreate the soundfield of a place, you need to make very precise measures of the reverberations, reflections or resonnances for all the audio frequency range (20Hz-20kHz) with corresponding amplitudes and time delays or time required for a signal to decay below a stated level. For example a jazz club raises direct sounds with high early first reflections and poor reverberations rate. On the other hand, a concert hall will have very little direct sounds but plenty of long time decaying reverberations and long first reflections. All this work is then converted into software instructions that will be used by the DSP to process the soundfield using input signal. Consequently the process of an AUDIO DSP now appears to be quite obvious to understand :

The two input signals are converted by sampling and quantization into digital data. Converters are usually 16 to 20 bits resolution (the higher the better) at 44.1kHz or twice the audio bandwidth according to the Nyquist Cryterion.
Samples are stored into RAM. They are not processed firstly since next samples are required. They will be used after a few clock cycles by the DSP.
The DSP compares previous samples to older ones and current ones. The results are analysed using software stores in ROM. This stage is very important since comparison between samples allows correlation index to be found and gives all the frequencies comprised in the little part of signal being processed.
Using the results of the signal analysis the DSP processes the output data to be sent to the different output signals. A new correlation and a frequency based process are applied as well as amplitude control following the specifications of the soundfield to simulate (software). A set of errors is implemented to avoid distortion in the signal or processing errors.
The DSP may store samples in another RAM for a few hundreeds of clock cycles in order to generate the desired delay times. Reverberation is achieved by sending the same data many times with a decaying intensity on the channel where it occurs and by sending a corresponding amount of the same signal to other channels to simulate the spaciousness of the reverberation.
All output signals are finally converted into analog form and are rearranged by analog processing to provide fully musical signals.

Many different soundfields may be processed with the Yamaha Cinema DSP chip (usually about 10 to 20 of them are available on amplifiers) but they are all based on the parameters to simplify the process and make it running faster (delay time, amount of reverberations, amplitude of the different channels and spaciousness are controlable). The speed of the processor is of a few megahertz. The figure below describes the main digital features surrounding the DSP.

CINEMA DSP FACILITIES

Very useful On Screen Display with DSP parameters control, test signals, center modes,...
Center speaker 5 frequency bands for adjustments so that tonal character of center matches that of the main stereo pair.
DSP Parameters control and storage (channel level, delay time, liveness, room size)
Full functions on remote control (programmable for the DSPA-970 and DSPA-2070)
Pre outputs for main, center, effects and sub on many models
Front mix switch for seven or five speakers configurations
Input level trim to give all your sources the same input level (very useful)