Many people who purchase an A Cappella III do so for I2S. As most of you know, it's the preferred output here at Musica Pristina.
What you may or may not know is that we follow what's been lovingly referred to as the "PS Audio Spec."
TL;DR Version: We really like I2S here. Since I2S carries a word clock, DACs can "auto-adjust" to the bit rate, meaning we can pad the data samples of a 24 bit track with 8 empty bits and transmit a 32 bit signal. The older versions of PS Audio gear required that we convert all 32 bit signals to 24 bit when interfacing with their DACs. The "Sunlight" update has improved upon this. We now send 32bit signals for everything 24 and 32 bit content.
What follows is a brief history of the transmission of digital audio signals.
See, back in the stone ages of audio, there was a device that spun plastic discs containing laser engraved audio as digital data. Most of these ancient devices (we can call them CD Players) had a DAC chip and analog outputs. The digital audio signal was read off the disc then passed to the DAC chip as an I2S signal. (Technically, it's I2S, but it's often just written as I2S... and spoken as "Eye Two Ess")
There was another excellent use case for the CD player flavor of these devices, too. One that could only output that digital signal to be decoded (and amplified) by an external DAC (or home theater receiver, which performs the D-to-A conversion). The engineering question was "how best do we send this signal from the CD Player (or Transport, if the player had no DAC chip of its own) to the external DAC?" The spec for I2S (Inter-IC Sound) was for a board level signal, not something to be passed around outside a chassis.
Let's understand something for a minute. People had just invented a way to take an audio recording and burn it onto a disc using a laser, then read it back using another laser. That's quite an accomplishment. Surely sending this digital signal containing clocks and data is not nearly as complex, is it?
Nope.
Sending I2S outside of a chassis was not too difficult for the engineers at Sony and Phillips.
What was an insurmountable engineering problem was the pressure from the team in marketing.
See, engineers are tasked with inventing and creating the cool new things people never knew existed but now need to have. Marketers are responsible to figure out how to make people want it and buy it.
The easiest way to ensure mass adoption of this new technology was determined to be the path of least resistance. (no, not an Ohm's Law joke) In this case, least resistance meant not creating a new physical connection to go from the CD Player to the DAC. A new cable capable of sending I2S would require a minimum of 3 signal lines (SCK, WS & SD) and ground. That cable didn't exist in mass production.
An I2S cable didn't exist, but the RCA cable did.
A one cable solution is ideal... for marketing. It eliminates the need (cost) of multiple cable connectors in both the CD Player and the DAC.
An existing cable solution is ideal... for marketing. It eliminates the need for machinery to be built to create a new type of cable along with all the supply chain headaches to support it. Not something you want to tackle during the rollout of a brand new product idea.