I think you are missing the point with your reply to my post. While the numbers are indeed difficult to ascertain using instruments ($150K for an Agilent 5052 for exampleÂ…we prefer to let our key suppliers amortize this expense) the proof is in the listening. While design decisions are informed by the numbers and engineering decisions driven by those numbers, the final decisions are driven by listening tests! We easily hear the improvement between the excellent clocks in the mLink versus the uLink, close in phase noise (below 100 hertz) is nearly identical but the phase noise from 100 Hz to beyond 1MHz is some 10-20dB improved on the Ultra-Low Phase-Noise clocks in the uLink and REFLink. This phase noise represents some 250fS (0.25pS) in the mLink and some 70fS (0.07pS) in the uLink and REFLink. The REFLink improves on the noise floor of the clock by offering further isolation from the computer (the USB Input section and processor are isolated from the clock section yet they are still self-powered with a low noise isolated internal supply-unlike the BADA which powers this function from the computer) and overall lower noise power supplies in the REFLink. The results of these 2 improvements are easy to hear between the uLink and REFLink (although difficult and expensive to measure as you state). My conclusions are that simple measurment of jitter is no longer adequate to understand the impacts of clock noise on audio reproduction quality. You reply also understates the critical importance to audio quality of advancements represented by our new Links and products like the BADA.
- ...
- 48 posts total
- 48 posts total