in my honest opinion, i2s is a superior protocol
hmm.. i think this is a question for the company. However, I believe the "clock" becomes more stable, therefore the signal; much like having a separate master clock on a CD player - that’s BNC, but CD player outputs also use i2s. Hardware almost always outperforms software. What you’ve added here is almost like me selecting CD input on jriver, jitter correction and 4x. This results in improved PRAT -pace rhythm and timing. --> is this the improvement?
In theory, they would be seperate units. In practice, they would be just one unit, optimized for better performance.The Femto TCXO and FPGA would be doing the heavy lifting...