ANC Digital vs Analogue
In active noise cancellation there is often debate over whether digital or analogue systems provide better performance. Lets examine the differences and similarities and see how they might affect the noise cancellation performance
The principle of noise cancellation is based on the observation that overlapping sound waves add together. So if we have a noise sound wave we want to reduce we can create a negative of that sound wave so that when the two add together the result is zero … no noise. This affect doesn’t care if the negative sound wave comes from digital or analogue electronics, only that the canceling sound wave be a pretty accurate negative of the noise sound wave. So at this level there is no difference between digital and analogue.
In order to create a negative sound the normal method is to have a sensor microphone pick up the noise, invert that signal and pass it out through a speaker. That sounds simple enough but in practice it is not so simple. The speaker used to generate the negative sound will be far from perfect, the noise signal will change significantly from the point at which the microphone picks it up and the point at which it reaches the ear and the microphone itself will convert the noise signal to an electrical signal in an imperfect way. The net result of this is that a noise cancellation system requires a compensation filter to ‘undo’ the all these effects and get a accurate negative sound wave to occur at the point its needed. It is this compensation filter that can be implemented by a digital or analogue system
The analogue system uses operational amplifiers and passive components, resistors and capacitors, to create the compensation filter, whereas the digital system uses computation to create the compensation filter.
An analogue filter has the main advantages that it takes essentially zero time for the signal to propagate through the filter and come out of the speaker ( actually there is a time delay but its well beyond anything that would be relevant to the speed of sound) and the filter can be designed to have a very low power consumption. The disadvantages are that the filter accuracy is limited by the tolerance and physical characteristics of the components used, the PCB requires space to place the components and the designer must decide in advance what filter structure to put in place
The digital filter has the main advantages that the filter itself is not a physical structure and can be changed at will simply be changing some numbers, many devices incorporate digital processing and so no additional parts are needed to implement an ANC compensation filter, the digital filter is not subject to tolerance or other component variations and can implement complex filters with high accuracy. Digital filters have the disadvantages that processing takes a relatively long time and can consume significant power, signals must be converted from electrical signals into data and then back to electrical signals after processing and accuracy is limited by sampling rates, ADC and DAC precision.
On subject of time delay the analogue system is the clear winner. With an analogue system the time delay roughly approximates to zero for audio signals. One can imagine that once a sound wave is picked up by a microphone there is a limited time until it reaches the ear. For low frequencies the time constraint is not a great issue because the wave changes gradually through space but at higher frequencies time delay becomes a critical limiting issue. Having said that digital ANC systems are already available that produce excellent results so it is evident that digital systems are already fast enough.
The analogue system has a beauty to it that the entire chain of microphone amplifier, compensation filter and output amplifier is one cohesive system of coercing electrical energy to get a desired outcome, whereas by contrast the digital system seems more of a brute force approach of number crunching and bouncing in and out of digital space. For a suitably designed acoustic system the analogue method could deliver great ANC performance in an elegant fashion.
A digital ANC system by contrast requires perhaps a little less finesse as it does not depend on an experienced engineer to build a filter structure. Digital can also tolerate inopportune changes in the acoustic system in that the filter can be modified without needing to change any physical component. This may be a draw card for digital systems in that the adaptability makes practical use easier and more widespread.
For the time being analogue ANC systems are more common with digital systems tending to be more expensive. Both however are delivering excellent performance with the right design. The industry may gravitate one way or the other over time but we can expect that both digital and analogue systems will find favour.