In article , Arny
Krueger
wrote:
I seem to recall that there is a freebie HDCD decoder kicking around.
What does it do?

Good question. :-)

The only one I know of is the 'hdcd.exe' file that 'cjk32' released for
Windows some years ago. (Hence my looking for C J Key.) No source code was
released. Nor any specific explanation of the algorithm used, or what parts
of what the HDCD patent *claims* to do it decodes.

I *think* he also let someone working on BSD have a copy of the source. But
as yet I've not tracked that down if true.

I have a copy of hdcd.exe as it is still on the web. But it isn't much use
to me without a clear specification of what it does, and how. Without that
I can't tell if it works 'correctly' in terms of whatever any 'official'
HDCD dac/decoder is supposed to do.

AIUI MicroSoft took his code and say they put it into their software
players. But they bought the HDCD Patents. I have my doubts they actually
know or care about the details beyond it being another 'feature' they can
offer.

If you know of something else, please let me know.

If you read the HDCD Patent, etc, it makes a whole series of general
claims. And says the control codes are 'dithered' into the LSB stream every
now and then. It gives the example of a maximal length sequence as seems to
say that is how the control is made to mimic 'noise'. But no real details
of what the control codes *actually* are, nor the *actual* 'spreading'
process. Nor the details of any of the claimed applied effects.

An AES paper by Johnson shows the curve said to be representing the 'soft
limiting' HDCD can optionally apply to squeeze up the waveforms by 6dB. But
again, no real detail of how this is actually flagged, etc, etc.

I can easily apply a reverse gain correction for the soft clipping. If I do
that *assuming* it is present, then some recordings sound much better (less
compressed). But I still have no idea if the result is genuinely
'corrected' or just less nasty as the waveforms aren't so badly mangled. I
may just be making 'less bad' some of the compression applied by 'clever'
remastering gurus. Not really undoing that part of HDCD.

For all I know, that is all 'hdcd.exe' does... or it may do something
totally different. The Patents and blurbs claim HDCD lets you mimic wider
frequency bandwidth as well by furtling about with the digital filtering
and giving output upsampled to 88.2k. But zero details...

If I had an HDCD player I'd try seeing what that does to the discs. But at
present I don't have one. And it seems a bit mad to buy one just for a few
tests on a tiny number of discs. Particularly if it confirms my growing
feeling that HDCD is a combination of some (reversable) soft peak limiting
and much smoke and mirrors...

All an interesting example of how a Patent - which is supposed to tell
'skilled practitioners in the field' how to duplicate something - fails to
provide the real details required for the actual system in use.

I don't know if CJK reverse engineered the spreading and managed to assign
the codes for different purposes, etc. At present I doubt it. e.g. How
would you work out from the discs what the different replay FIR taps of the
reconstruction filter were meant to be? But I am keeping my fingers crossed
that he will eventually reply and give me some useful info.

Slainte,

Jim

--
Please use the address on the audiomisc page if you wish to email me.
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Armstrong Audio http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc http://www.audiomisc.co.uk/index.html