Ghostrider - I read this before but have hard time to understand importance of it and also why wouldn't this be applied to regular redbook CD in downmixing. This is more of preprocessing technique and has nothing to do with HDCD. If I learn that, for instance, TELARC is using similar technique to downmix their redbook CDs I wouldn't called TELARC CD a different standard. TELARC currently uses DSD recording format/technique - does it change CD standard?.
I'm also not certain how much it affects the sound. There is probably very little energy above 20kHz and it is already filtered out in downmix processing. Whatever left above 20.5kHz is folded into passband starting from 0Hz but people who own NOS players don't complain. They even claim better sound.
So - now they claim that what differentiate them from just simple downmixing process is that they hide somehow information about status of adaptive filter at given moment in the music itself. Hiding it in the lowest bits while it's used only 2% of the time maximum (as they claim) is strange. Maybe I'm slow to understand it, but if they used adaptive filter in downmixing why do they need to match it in playback? They stated before that when high frequency info was present (like cymbals) they applied sharp antialias filter and when it wasn't present they bypassed filter. All this was done to avoid using sharp antialias filter in the player. Fine, but now there is no high frequency above 20.5kHz in the mix and even phase is the same for all signals - I don't understand what they are matching (and how)? Oversampling and Bessel filtering in my DAC sounds simpler to me.
Error correction code (Reed-Solomon) is pretty weak and the player bypasses the data with the wrong checksum. That would mean that adaptive filter info hidden in the lowest bits (as a pattern or a sequence) can be lost (fingerprints or scratches). It cannot be that important.
Again - I don't question that it sounds great. I'm just trying to understand. Perhaps the fact that it is not very popular has nothing to do with technical merits (SACD was killed by greed in my opinion).
I'm also not certain how much it affects the sound. There is probably very little energy above 20kHz and it is already filtered out in downmix processing. Whatever left above 20.5kHz is folded into passband starting from 0Hz but people who own NOS players don't complain. They even claim better sound.
So - now they claim that what differentiate them from just simple downmixing process is that they hide somehow information about status of adaptive filter at given moment in the music itself. Hiding it in the lowest bits while it's used only 2% of the time maximum (as they claim) is strange. Maybe I'm slow to understand it, but if they used adaptive filter in downmixing why do they need to match it in playback? They stated before that when high frequency info was present (like cymbals) they applied sharp antialias filter and when it wasn't present they bypassed filter. All this was done to avoid using sharp antialias filter in the player. Fine, but now there is no high frequency above 20.5kHz in the mix and even phase is the same for all signals - I don't understand what they are matching (and how)? Oversampling and Bessel filtering in my DAC sounds simpler to me.
Error correction code (Reed-Solomon) is pretty weak and the player bypasses the data with the wrong checksum. That would mean that adaptive filter info hidden in the lowest bits (as a pattern or a sequence) can be lost (fingerprints or scratches). It cannot be that important.
Again - I don't question that it sounds great. I'm just trying to understand. Perhaps the fact that it is not very popular has nothing to do with technical merits (SACD was killed by greed in my opinion).