MC uses the same HDCD software decoder as foobar, which is pretty well documented. It is based on a decompilation of the Microsoft HDCD software. Microsoft bought HDCD from Pacific Microsonics so their software code was probably pretty close in function if not exactly what was in the original chips, with one caveat. HDCD uses two main techniques - Peak Extension and Low Level Gain Adjustment. . Peak Extension allows the dynamic range to be increased by up t 6 dB. Low Level Gain Adjusment allows low level signals to be temporarily increased so as to not be lost in the noise floor. Decoding undoes the boost HDCD also had an option to use one of two transient filters, although there is evidence that only one was ever used. Both of those are implemented in the software decoders. The people who developed the software decoder based on the decompilation decided that the filters were not all that important and did not implement them. So, I should presume that the MC implementation does Peak Extension and Low Level Gain Adjustment and not the filters.
HA’s statement that HDCD is a scam typical of HA. They think that many techniques that the audiophile community users are scams.
Many CDs have HDCD markers but do not implement Peak Extension and Low Level Gain Adjustments. That is because many engineers used Pacific Microsonic A-D/D-A converters (PM Model One and PM Model Two) because they were extremely good units, well ahead of the competition at the time. They are still coveted today. The converters did put a marker in the files even if none of the HDCD techniques were used. That was not a scam, since those CDs were not marketed as HDCDs.
According to one of Matt's post, in detecting HDCD he added a check to be sure that actual changes were made, so that a simple marker would not be detected as a HDCD.
Reference Recordings was the last company that I know of doing HDCDs. They did them up to a few years ago, but I am not sure I they are still making them. One of the principles at RR was part of the Pacific Microsonics team.
The HA list of HDCDs is quite incomplete. A more comprehensive list is on Goodwin’s website, although it does not list if HDCD code is actually used in the recordings. That website also has a more neutral writeup on the intent and the implementation of HDCD.
http://www.goodwinshighend.com/hdcd.htmIf you want more detail of the foobar development, I can dig up those references. EDIT - Added below.
EDIT : Here is a link to the original AS paper on HDCD if you want to get into the gory detail.
http://web.archive.org/web/20020124220637/www.hdcd.com/partners/proaudio/AES_Paper.pdfEDIT: If you want to read the early discussions of the software decoders, take at look at these forum threads.
https://hydrogenaud.io/index.php/topic,30999.0.htmlhttp://forum.doom9.org/showthread.php?t=129136cjk32 on the doom9 thread is Christopher Key, the guy who got this started and did most of the early work on this.