Dear Mr Nishio,
For years I’ve been in awe of
your previous work in audio technologies. Today, I’ve plucked up the courage to
do whatever it takes to get in contact with you to solve some of our (very
little understood) matters.
I wish to ask for some sort
of detailed explanation of SBM-Direct (How might it differ from a AudioGate
conversion for example). I know it uses separate processors to do its 30000+
taps but what actually happens after that? Does it go to a SBM processor and if
so, how is this sbm different from the one for pcm processing? All we currently
know is an experimental real-time down conversion processor was built using a
special 1-bit convolution circuit and fixed point digital signal processors
(DSP). A 32767-tap convolution is processed by 8 paralleled 4096-tap
convolution processor cards per channel. The processor card outputs are
added and then processed through extended precision Super Bit Mapping (SBM) on
a DSP card. Via a digital I/O card, word sync and the DSD signal are
input and the 44.1kHz/16bit signal is output by SDIF-2 format.
I also wish to know whether
the popular VAIO bundled DSD-Direct is truly a reversed version of the
algorithm for SBM-D. Does that process involve the reversal of the SBM process
as well?
Some people have in the past
used ADA-7000R and other rare equipment to do DA-AD-DA. My question is why ‘アナログ吸い上げ’ is done at SMC and why can’t we just use DSD-D and
then SBM-D if it’s simply the DSF file characteristics that we want (lack of
harshness or something). In other words, how does ‘all digital’ compare with
‘part analog’ when doing format conversions? Why would the quality be
different?
Finally, who actually decided
to add the infamous CXD9872 chip to VAIO and do you have a datasheet for it?
Why doesn’t Sony release a USB DAC that uses such a chip to allow PCs to record
DSD – a rather nonsensical question in fact! I’ve had people say that any of
these consumer devices (VAIO and PCM-D100 and the HX record player) are NOT
true DSD audio. Do you agree with this?
