Yes, you are absolutely correct. The addition of both the Phoenix USB and NET would close the gap considerably, to the point that both (S or ZIII) would deliver more or less the same startlingly pure, massively spacious, highly dynamic and rhythmical musical performances, with the edge still going to the Statement, as you’d expect. And of course there’s Statement + PN MMMMmmmmmm!
One other thing to bear in mind is that the better the system becomes, the wider the gap between Roon and Innuos 2.x.x sound quality, to the point that using Roon becomes a major SQ sacrifice.
The reason that Innuos weren’t promoting SQ with 2.0 is because the difference between it and say iPeng 1.4.9 varies according to how optimized the network data stream. 1.4.9 is a ‘pull’ application, so its constantly polling to see if there’s any outstanding demands to service. With Roon, network traffic back and forth is constant. The whole Innuos design is about reducing EMI, which is why they chose a less powerful CPU, so network traffic and CPU interrupts should be minimised where possible, with corresponding positive impact on SQ. Innuos 2.0 is a ‘push’ application, so remains completely ‘silent’ until there is a command to execute. If you extend this same philosophy with the rest of your network, you can achieve some truly stunning results.
My network is based on wireless isolation (I could use optical here and maybe will when I get an FTTP service), using the following:
Tri-band router (TPLink Archer AC5400) with one 5GHz band dedicated solely to audio into a highly modified* TPLInk RE650 set up as a wi-fi to ethernet bridge, with polling and 2.4GHz switched off. I get about 250Mbps, transmission rate, so the GUI runs at speeds similar to a local interface and downloads of albums and metadata take fractions of a second. With music tracks loaded into RAM, there is virtually no network traffic during music replay, and the CPU is barely interrupted from its main task of music processing. With so little activity, heat generation is minimised and power supply loads are benign. I have applied the same design philosophy to rest of my network, which uses Sean Jacob’s DC4 regulated supplies throughout, thereby minimising harmonics where the noise from different power supplies interact. I use a Synergistic Research Atmosphere X Ref Ethernet loom with all shields star wired to a low impedance ground. All DC cables and internal PS cabling is Mundorf silver/gold with JSSG360 shielding on all external cables, providing very low impedance DC supplies to all network modules. All modules and power supplies are mounted on anti-vibration bases or racks.
- The RE650 network bridge is wall mounted on an anti-vibration platform that maintains antennae orientation. The integral wall plug and 5V SMPS are removed and replaced with a DC4 regulated LPS. Cabling as mentioned above.
Every step taken to improve the network resulted in enhanced sonics that I thought well worth the investment. Some, like the Mundorf silver/gold cabling gave improvements that honestly sounded like I’d substantially upgraded my speakers. Adding SJ DC4 regulation gave the music huge amounts of drive and vitality, with SOTA PRAT, that practically bounced me out of my listening chair, something you’d usually only achieve by major upgrading of an amp or adding additional power supplies. Above all, the improvements were all fundamental rather than cosmetic, which made the music far more believable and generated subconscious emotional responses, which ensures increased long term enjoyment and a compelling interest in listening to music