A
Anonymous
Guest
Craig M.:maybe, although with my dacmagic i think the optical from mac is (very) slightly better than usb (although the difference is so small i might be imagining it). i'd love to know the reason, because the theory behind using a hdd based source makes sense to me. hopefully the naim dac will make an appearance before i can afford the cyrus combo, and i can then demo a dac which doesn't rely on the timing of the incoming signal. having said that, i wouldn't be suprised if the cyrus dacxp reclocks the signal at input, using a PLL or something. or i might stop overloading my brain with it all and just get what sounds best.
It is feasible that there is the difference between usb/optical as the 'work' is done by different components in the chain, with differences in quality, and in this particular case this may work out to the benefit of optical. Reclocking by the receiver is not so straightforward as it seems - even slight timing differences between the two clocks will accumulate, and a dac slave can not make the incoming optical signal speed up or slow down. So I doubt that the cyrus does it, it will rely on a jitter free incoming signal. Also, while USB has to reclock the data (since there is no clock signal in the data transfer, just packets of bits arriving every msec), methods to adjust for these timing differences due to clock differences between source and receiver vary. The TI / BurrBrown method appears to work quite well. Making the DAC the master of the data transfer to begin with would have been easier. It seems that DNLA will be the answer here. Until then, stick to your last sentence & enjoy..

It is feasible that there is the difference between usb/optical as the 'work' is done by different components in the chain, with differences in quality, and in this particular case this may work out to the benefit of optical. Reclocking by the receiver is not so straightforward as it seems - even slight timing differences between the two clocks will accumulate, and a dac slave can not make the incoming optical signal speed up or slow down. So I doubt that the cyrus does it, it will rely on a jitter free incoming signal. Also, while USB has to reclock the data (since there is no clock signal in the data transfer, just packets of bits arriving every msec), methods to adjust for these timing differences due to clock differences between source and receiver vary. The TI / BurrBrown method appears to work quite well. Making the DAC the master of the data transfer to begin with would have been easier. It seems that DNLA will be the answer here. Until then, stick to your last sentence & enjoy..