conspicuous wrote:gents, thanks a lot for sharing your inputs.
indeed luckily enough the sound can fixed with just a switch of a button. i've kept my cayin pretty much on the tube (6922) output. in a few days i'll switch over to the ss output. it's been a while but i remember the ss being less warm (duh, no surprise there) but still enjoyable in a different kind of way. like a good wine vs. a good scotch
i also read somewhere that 96/192khz are not ideal sampling rates because they are not multiples of red book 44.1khz rates. the very expensive mark levinson transport use multiples of 44.1khz. but i don't feel a need to upgrade my cd player and i'm enjoying it
i've now bumped up to 25% digital/75% analogue from just 10% digital just a month ago.
Multiples = generally oversampling (integer upsampling / synchronous upsampling)
Non-multiples = generally upsampling (non-integer upsampling / asynchronous upsampling)
"Oversampling" players and DACs may have DACs capable of running 192kHz or 96kHz (often advertised), but being limited by the built-in oversampler of most DAC chips, they normally just upsample to 176kHz or 88kHz when playing redbook cds. This is a cost-effective implementation, and can possibly have lesser losses. However, since the DAC is running synchronously with the source, it becomes very susceptible to transport and interconnect jitter (external DACs suffer the most, as they don't get the advantage of running I2S). With an oversampling DAC, a cheap DVDP vs. a decent CDP as transport may turn out to be a night and day difference. The main advantage of this design over NOS is that it shifts the aliasing noise to somewhere beyond the audible range (making it easier to filter out in the analogue domain). The simpler filters tend to introduce less phase distortion, and the fact that they can be set higher up the frequency range means that the effect on the audible range can be far less detectable.
"Upsampling" players and DACs generally have DACs running asynchronously from the transport (through an ASRC), which allow for resulting sampling rates that are not multiples of the original. This feeds the DAC chip with the final sampling rate (192kHz for example), (usually) with no further need for on-chip oversampling. This is clearly a costlier implementation, but the reclocking of the DAC's input eliminates some of the negative effects of jitter. These designs can tolerate far higher jitter levels without too much output degradation.