cMP² | CMP / 01WhyPCs

Chapter 1 - Why are PCs good for audio?

It is increasingly recognised even in the sometimes conservative high-end audio community that there are sound technical reasons why a PC can supply a better-quality digital audio signal than even the most sophisticated of CD transports.

A conventional CD player reads a disk, whatever its imperfections, in real time and must therefore deal with the many read errors that are encountered with any CD as well as its design allows. It also suffers from the phenomenon known as ‘jitter’, defined by the CCITT as ‘short-term variations of the significant instants of a digital signal from their ideal positions in time’. In digital audio, jitter is the ‘mother of all evils’, its effects accumulating as signals pass down the processing chain even when a music signal is captured accurately in the first place. Controlling jitter in a real-time read operation is neither easy nor cheap. As one audio manufacturer explains:

A simple CD player has multiple motors or actuators and associated control loops in order to perform disc reading. There are [e.g.] the spindle motor that turns the CD, the sledge motor that performs axial tracking and actuators for focus and radial tracking.
Each of these motors/actuators will add a portion of noise to the power supply of the player and this noise will affect accurate switch timing. So each of the motors/actuators adds jitter to the digital audio signal and each adds a different kind of jitter (different in frequency, amplitude, waveform) and each will affect audio reproduction in different ways.

Read errors and jitter-related distortion introduced at the reading stage can be overcome by using a PC and competent software to read CDs, however flawed, prior to and separately from playing them back. It can take as long as it needs to capture perfect or near-perfect data and transfer them to disk (ripping, in the jargon) virtually without error. This is a significant advantage.

PCs with adequate processing power offer superior ‘upsampling’. A technique based on sound mathematical principles known as Bandlimited Interpolation (research paper) is used:

"In digital audio, what matters is the audibility of interpolation error between samples. Since Shannon’s sampling theorem says it is possible to restore an audio signal exactly from its samples, it makes sense that the best digital audio interpolators would be based on that theory. Such ‘ideal’ interpolation is called bandlimited interpolation.

. . .

the problem is to correctly compute signal values at arbitrary continuous times from a set of discrete-time samples of the signal amplitude. In other words, we must be able to interpolate the signal between samples. Since the original signal is always assumed to be bandlimited to half the sampling rate (otherwise aliasing distortion would occur upon sampling), Shannon’s sampling theorem tells us the signal can be exactly and uniquely reconstructed for all time from its samples by bandlimited interpolation."

In short, bandlimited interpolation allows an exact re-creation of the analogue audio signal within defined error margins.

In playing back from PCs, bit-perfect data is read from hard disk to memory and upsampled with a precision better than most, if not all, conventional transports without the need for real-time clocking mechanisms – jitter is not encountered during the data preparation stages.

However, the final stage involves transfering data to an output device connected to the system by an internal, USB or Ethernet bus. This device (here called the soundcard) adds a clock and generates an SPDIF or equivalent signal for a DAC. Inevitably, it suffers from jitter and needs clean power and a high quality clock (unless the DAC performs input buffering or reclocking). Traditional transports face the same challenge but PCs are notoriously noisy electrically (and thus jitter-prone), a phenomenon that has hitherto left high-end CD players with a performance advantage.

This paper describes constructing and configuring a PC designed and configured to enable significant reductions in the noise floor and a concomitant reduction in jitter. The result is a system dedicated to its role as a CD transport and offering a level of performance that equals or exceeds that of the highest quality conventional products at a fraction of the price.

Although called a Computer Transport, it remains in essence a standard PC. The chosen components are good quality examples of industry-standard parts and the changes described are readily reversible. Most of the software is free.

Much of what follows will be familiar to anyone who has configured computers for any high-end use but there is also much that is not mainstream and two of the programs are unique. They are discussed in Chapters 11 and 12.

Nothing in what follows is beyond the skills of anyone who has moderate experience of working with Windows XP Professional SP2 or is willing to learn a few basic skills. There are excellent tutorial sites on the Internet. The reader can confidently use the same or similar components as the recommendations made here or can ‘cherry pick’ the text for ideas.

It’s just noughts and ones – why the fuss?

In a perfect world, a computer's internal processing proceeds without error but, in the real world, either proceeds with occasional errors which are detected and corrected by CRCs, parity bits, etc or corrupts data too severely to be (fault) tolerated, leading to a “crash”.

External activity (output) is a different matter: it is a real-time process with no scope for any “send-that-again” feedback.

In designing high-end, computer-controlled machinery for manufacturing, engineers deal on a daily basis with problems over which audiophiles anguish at length. Whatever the industry, such machines issue billions of instructions per minute to a plethora of devices under conditions in which errors of a ten-thousandth of an inch decide between a product and scrap metal.

To maintain accuracy, the control electronics, motors, transducers, power supplies etc. must be excellent and vibration suppressed to ensure accurate timing. True, the system is controlled by digital signals but the not-uncommon notion that “it’s just noughts and ones – why the fuss?” is ill-informed.

On a more modest scale, the same holds true for high-resolution digital audio. There is, however, an additional, and often overlooked, problem with PC audio: the “real-time” processing of the output is performed inside a more or less stock computer. Although PC audio bypasses real-time data capture (reading CDs “on-the-fly”), RF noise, supply fluctuations and a myriad of other pollutants which are relatively unimportant in data processing (on a PC whose operating system is designed to handle almost everything except precision audio) can severely degrade sound quality. That needs to be addressed. To configure a computer so that it preserves the integrity of the “real-time” output is to enhance sound quality. CMP² is a pioneering design in that it does this more thoroughly than has been the practice hitherto.