Ydee wrote on Today, 08:15:

MattRocks wrote on 2026-10-30, 15:07:

The clue for me is the two oscillating crystals by the main processor because that's the arrangement seen on ENVY24 processors to switch between native CD and native DVD frequencies without introducing resampling distortions - the current PCIe Juli@ does not have that.

Interesting, on ESI Maya44 (same DSP ICEnvy24 HT-S) is crystals configuration different.

That is peculiar and interesting to engineers, but it's not a capability difference for users.

Precisely how any PCB derives the CD (44.1kHz) and DVD (48kHz) frequencies from its oscillating crystals is opaque to me - sometimes I can see the mathematical relationship and sometimes I can't. For my simple mind, what matters is that the PCB has two separate sets of frequencies for one audio processor - if it had only one oscillating crystal then I'd be concerned that it might be forcing some inputs to be resampled because it has no physical clock to sync them to.

Unfortunately the ENVY24 has to pick one input frequency in any given moment - it can't process both at the same time, and having the whole thing flip to the undesired native frequency because of a contaminated upstream concerns me (e.g. Windows alert sound).

To get mix streams without loss of detail you need something like a Digigram PCX used by broadcasters to fade between recordings, commentary, announcements, adverts, etc.

I just found it curious that they decided to use the standard 22.5 crystal for 44.1 kHz on one card and derive 48 kHz from 49.1 MHz, while on the other they used the standard set of 22.5 and 24.5 crystals. But @Julia is a higher-end model than @Maya, so who knows...