Michael had pointed out last year that a global master clock would be ideal having an Esoteric audio and tarnsporter front end feeding a tact audio tcs and boz amp chain.

Imagine my disapointment when Boz walked with me over to the Esoteric room where some of the best sounding music was playing and it was confirmed that in order for the master clock to work properly it has to be manually reset for 44, 96, or 192 in redbook cd or 44 in dvd. The esoteric master clocks require a manual setting of the operating rate. Bummer.

It does seem like if you have the esoteric Po3u and dO3 DtoA's that then if you set the G0s to 100khz it will auto coordinate the frequencies for all sources including DSD but only for these propietary Esoteric devices and certainly not the slimdevices Transporter or Boz amps which include word clock input capability.

There goes my ambitious plan.

Could a pc be built with automatic word clock resynching to dvd cd at the different cd frequecies?

Well, that really blows. :( Surely there ought to be a way to autoselect the proper word clock rate. How do the proprietary two-box systems do it?

The short answer
most likely over RS232

Which aint there in the esoteric clocks.

In my book wouldn't switching clock frequencies be akin to swapping media ?

Insert CD DVD, etc , close the tray, select the freq for that type of media.

cd's alone can have 3 types .

992 Master Clock

The dCS 992 Master Clock is a high stability reference generator designed for studio and live recording applications. The unit provides reference clocks to support systems sampling from 11.025kS/s to 192kS/s or DSD, via AES3 or SDIF-2 Word Clock formats. It offers a variety of "Superclock" options up to 256Fs and can generate all the commonly used sample rates from 11.025kS/s to 96kS/s, and output up to 10 different sample rates simultaneously phase locked together. In addition, the relative phase of each output may be adjusted by plus or minus 1 sample, in steps of 1/128th of a sample. With the dCS 992, synchronizing all the components of a multiple sample rate digital audio system becomes a simple and painless task. The unit can operate in Master mode from its own precision TCXOs, or can slave to an external source, such as GPS. It can "learn"

I am going to check it out, thanks Bio.

If these word clock inputs could accept 28.224MHz as a reference clock, then that would solve everything. That's 640fs for Redbook, 588fs for DVD, 294fs for 96kHz, and 147fs for 192kHz. It would not be difficult for a DAC to autodetect the sample rate under that and adjust its filter rate accordingly. What's more, that exact frequency is in common use in other applications, including modems and software-defined radios. Heck, some computers use that frequency for a real-time clock interrupt.

EDIT: Oh, it's also 10 times the standard DSD/SACD sample rate.

The Boz amp probably can you just dial it in, but how do you clock back to the source say a transporter or Esoteric ux-1/po-3?

Well, the idea would be something like this. The transport, say the Esoteric, would simply divide the 28.224MHz reference clock as needed to correspond to its desired output rate. The DAC would have to do the difficult work: to figure out the sample rate and the phase offset used by the transport. But actually that's only marginally more difficult than the sample rate detection employed when using a standard S/PDIF connection.

I'm not saying that the Boz or some other DAC can do it, but I am saying it shouldn't be that difficult.

BTW, 28.224MHz is also 80*352.8kHz, which is the standard sample rate for DXD. And it covers 88.2 and 176.4kHz, too. So this really is a "magic" sample rate for all the standard audio formats.

Actually, now that I think of it, there's no reason that a fixed 48kHz or 44.1kHz word clock can't be used to synchronize any of these sample rates.

For example, say the output sample rate is 192kHz and you have a 44.1kHz word clock. The transport then has to output 640 samples every 147 clocks. Sure, it looks hairy, but really, it would not be difficult to implement, just some standard clock multiplication or a PLL with a 28.224MHz VCO.

Jitter, which can be a concern with clock multiplication, wouldn't be an issue here because the DAC is controlling the clock.

So I have to confess I'm stumped as to why these guys just can't get 'er done. I'm free for consulting work these days if they want to call me :)