HDCD - is it worth it?

[johndoe]

I recently bought some King Crimson remastered CDs, when I played them on my PC it said that they are HDCD. I had never heard of that format. They sound nice, but is there a real (audible) difference from the regular CDs?.
Is purchasing a HDCD capable player worth it?
Will a universal CD player (CD, DVD, DVD-A, SACD, etc) decode the HDCD content?
Will MS Media Player decode the 20 bits or just tell you that the disc is HDCD?
Is this one more of those DOA technologies?

 

[MDS]

I have 5 or 6 HDCD - the Van Halen Remasters. I personally can't tell any difference between them and the regular Rebook CDs.

A CD player (or receiver if you use a digital connection) has to contain an HDCD decoder to play the full 20 bits. I would not discount a player or receiver just because it doesn't have an HDCD decoder (none of mine do). HDCD will play in any cd player.

Microsoft bought Pacific Microsonics, the company that created High Definition Compatable Digital,and I believe WMP does actually decode them.

 

[Buckle-meister]

I recently bought some King Crimson remastered CDs, when I played them on my PC it said that they are HDCD. I had never heard of that format. They sound nice, but is there a real (audible) difference from the regular CDs?.

I doubt there's any way to tell. The only way that I can think of would be to have a 16-bit recorded CD and compare it to a 20-bit HDCD of the same album.

Regards

EDIT: I didn't realise one could actually buy two versions of the same CD.

 

[bigbassdave]

This is the first time I have heard of HDCD. Is it better than SACD or DVD audio? Thanks

 

[johndoe]

It is not better or worse, it is different: It's stereo, and it's 20 bit technology as opposed to the regular 16, supposed to have more audio information. I can't really say more without risking getting a foot in my mouth

 

[Buckeyefan 1]

This site may answer some of your questions.

http://www.microsoft.com/windows/windowsmedia/forpros/hdcd/hdcdfaq.aspx

 

[mtrycrafts]

It is not better or worse, it is different: It's stereo, and it's 20 bit technology as opposed to the regular 16, supposed to have more audio information. I can't really say more without risking getting a foot in my mouth

Red book is 16 bits. human perception is not much above this, maybe 17 or 18 tops. If you place actual 20 bits on a CD, that is not red book and the player will not read it. Today's CDs are mastered in 20+ bit encoders already.
SACD may also have a level difference issue to make it sound different.
And, one needs to be absolutely sure that the mastering was identical to both formats before comparing.

 

[MDS]

HDCD are 16 bit on the disc so that they conform to the Redbook standard and will play in any cd player. It is encoded somehow so that an HDCD player can extract the extra resolution (probably analogous to how dolby surround is matrixed into a 2 channel signal).

 

[mtrycrafts]

HDCD are 16 bit on the disc so that they conform to the Redbook standard and will play in any cd player. It is encoded somehow so that an HDCD player can extract the extra resolution (probably analogous to how dolby surround is matrixed into a 2 channel signal).

I think it would be easier to understand matrixing than what HDCD is doing, or not doing to extract 4 more bits from 16 bit data.
Pro logic uses left right signals that are in phase and the same for the center ch. and phase differences for the surround. But how exactly they encode 20 bits into 16 and extract, I need to think on that some more

If this was such a great discovery, it would be standard for all CDs? Just think of the data bits saved using 16 bits and achieve 20 bit resolution, if that was audible in the first place. A whole bunch of storage space saved or much more music on a disc?

 

[MDS]

I wonder too. A few years ago I did try to find info on exactly how it is done, but didn't find any. I can speculate on how it might be done though.

A cd player reads 2,352 bits at a time. After all the redundancy and circ bits are removed, you are left with the actual pcm samples. The extra 4 bits would have to be allocated among those samples because if you play or rip the cd, the result sounds like a regular cd. So maybe it uses every 4 samples to form one, using the least significant bit of every sample as an additional bit for a 20 bit sample. That might work because in the case where you don't have an hdcd decoder, the least significant bit being 0 or 1 won't make a difference in what you hear.

Using 4 bits for brevity:
sample1: 1111 (15)
sample2: 1010 (10)
sample3: 1100 (12)
sample4: 1110 (14)

sample5: 1011 (11) - accumulate the lsb from the prior 4 samples and use them to form an 8 bit sample: 1000 1011 (139). If you didn't have an hdcd decoder and the lsb of each sample didn't really 'belong' to that sample, it wouldn't make a difference in the sound because changing the value of sample1 from 15 to 14 (change the lsb to zero instead of 1) won't be audible. [of course in my example, sample values that low are inaudible anyway.]

It may instead use the subchannel bits (q,r,s,t etc) to store the extra bits because they aren't generally used for audio cds.

Who knows, I'm just speculating. Being patented, we may be able to get some info by using http://www.uspto.gov and looking for hdcd patents.

 

[mtrycrafts]

I wonder too. A few years ago I did try to find info on exactly how it is done, but didn't find any. I can speculate on how it might be done though.

A cd player reads 2,352 bits at a time. After all the redundancy and circ bits are removed, you are left with the actual pcm samples. The extra 4 bits would have to be allocated among those samples because if you play or rip the cd, the result sounds like a regular cd. So maybe it uses every 4 samples to form one, using the least significant bit of every sample as an additional bit for a 20 bit sample. That might work because in the case where you don't have an hdcd decoder, the least significant bit being 0 or 1 won't make a difference in what you hear.

Using 4 bits for brevity:
sample1: 1111 (15)
sample2: 1010 (10)
sample3: 1100 (12)
sample4: 1110 (14)

sample5: 1011 (11) - accumulate the lsb from the prior 4 samples and use them to form an 8 bit sample: 1000 1011 (139). If you didn't have an hdcd decoder and the lsb of each sample didn't really 'belong' to that sample, it wouldn't make a difference in the sound because changing the value of sample1 from 15 to 14 (change the lsb to zero instead of 1) won't be audible. [of course in my example, sample values that low are inaudible anyway.]

It may instead use the subchannel bits (q,r,s,t etc) to store the extra bits because they aren't generally used for audio cds.

Who knows, I'm just speculating. Being patented, we may be able to get some info by using http://www.uspto.gov and looking for hdcd patents.

Would this account for the way the Cd is encoded with the Reed Solomon Interleaving? Do you have Pohlmann's Principles of Digital audio? I am trying to understand it

 

[MDS]

Would this account for the way the Cd is encoded with the Reed Solomon Interleaving? Do you have Pohlmann's Principles of Digital audio? I am trying to understand it

Yes, I have the book. The circ is one reason why the number of bits on the disc is far greater than just sampling rate*bit depth*number of channels*number of tracks and why it doesn't just read the samples directly like would be the case for a wave file.