I could not characterize the jitter levels as inconsequential and below the hearing threshold, since my own experience, in blind tests, with several participants, led to undeniable conclusion that the choice of digital input, transport, and consequently the jitter, had clearly audible impact, with unanimous preference of Ethernet over USB or AES, heard on superbly engineered BR-20-like product costing 16K USD. Re-clocking does not eliminate jitter, all it can do is change its audible character. The clearly audible superiority of and unanimous preference for Ethernet input can be explained only by the fact that rendering happens inside the unit, without any need for re-clocking, with digital audio signal sent directly to DAC chip input over I2S. The same was observed on another less costly BR-20-like product.