In the Pono Player, Ayre began their work by designing the circuitry after the main processor retrieves the audio data from the memory and presents it in "3-wire" form - audio data, bit clock, and word clock. They chose the just-released version of ESS's top-of-the-line ES9018. It has two channels, comes in a very small package (5 mm square), and is extremely customizable, able to tackle the rigors of sensitive, low-level signal path design.
The filter generally favored by Ayre is a minimum-phase digital filter (to eliminate pre-ringing), with a "slow" roll-off, to minimize the overall amount of ringing (ringing can be thought of as an oscillation in the digital signal, causing all sorts of errors if misconstrued as actual signal to be converted to analog, which is engineer-speak for music). For the Pono Player’s D/A (digital-to-analog) converter Charlie went a step further and used a moving average filter for both the double and quad sampling rates because it has no pre-ringing, no post-ringing, no overshoot, and no undershoot (these create inaccuracies in the rendering of digital signal and sacrifice fidelity). In other words, it has none of the digital artifacts (digital artifacts also add to distortion and occlude signal) at all.
The DAC chip’s output comes in the form of current, so Ayre designed a proprietary, fully discrete, fully- balanced, zero-feedback current-to-voltage stage. This then goes to a fully discrete, zero-feedback buffer stage to drive both the headphone output and the line stage output. The output impedance is roughly 5 ohms, allowing the Pono Player to drive any headphone on the market with minimal frequency response errors.
