Re: GCR Encoding

[Eric Smith <eric-no-spam-for-me@brouhaha.com>]

"Bryan Parkoff" <bryan.nospam.parkoff@nospam.com> writes:
>     Yeah, I already understood your point since we have already discussed in
> the past.  I do understand that data bit and channel bit must be combined
> into sequence 16 bit or a pair of byte per raw data byte.

No, that is not what happens.  THere isn't any "pair of bytes".

When writing (in 16-sector format), the data is read 6 bits at a time,
and used as an index in a "nibblizing" table to get the 8 channel bits
("nibble") to write to the medium.

Because only 6 data bits can be read or written per nibble, a 256-byte
data sector is written in segments (my term, I don't know of any
official name).  The first segment consists of 256 nibbles each containing
six bits from a distinct  data byte.  The second segment consists of
nibbles formed by grouping the "leftover" two bits from three data bytes into
one six bit nibble.  This results in a total of 342 nibbles per sector
(plus additional nibbles for the address field, data field prolog, epilog,
and checksum).  Note that 342 nibbles actually encodes 2052 data bits,
but since there are only 2048 data bits actually used in a sector,
2/3 of the last nibble are effectively wasted.

When reading, the 8-bit channel codes (nibbles) are read then used
to index a denibblizing table to get 6 data bits.

>     I wish that 7+1 can be added, but it will require 3 or 4 (zero) bits
> that GCR encoding can't accept it.  The discussion is enough so it is not
> necessary to repeat over and over.

OK, if you don't want me to anser your questions, I certainly don't
have to.

>     My last question is about copy-protected disks.
> 
> 1)  Can Disk II read A VALID PAIR OF BYTE for each raw data byte inside
> copy-protected disk?

It is *completely* unclear what you mean by "valid pair of byte".  I
never used pairs of bytes ANYWHERE in my description of how this stuff
works.

> 2) Do copy-protected disk contain more than 3 (zero) bits that appear to be
> invalid?

Sometimes.  If they do, they can't be read completely reliably, for
reasons we've previously discussed.

> 3) If you answer yes to question #1, why nibble analyzer fail to read EACH
> PAIR OF BYTE? 

Question #1 didn't make any sens, and neither does this one.  Once again,
there are no "pairs of bytes" anywhere in the process.

> 4) I would appreciate if you provide more information how PAIR OF BYTE is
> stored as WEAK BIT

The aren't.  A weak bit has nothing to do with a pair of bytes.

> since I already have all the copy-protected scheme
> information from Copy II Plus manual.

I haven't read that manual.

> 5) Is it true that the manufacturer convert raw data 140K (probably 90K -- 
> 4+4 only, but no 6+2) into PAIR OF BYTE before one or more PAIR OF ZERO BITS
> may be added between each TWO PAIR OF BYTE that they really use Disk II
> drive, but not other type of PC drive like FM?  Please explain...

No, it's not true.

Are you somehow interpeting FM encoding using clock bits as turning
a data byte into a "pair of bytes"?  If that's what you're talking
about, the Disk II doesn't do that, except in using 4+4 encoding to
store the track, sector, and voume number in the address field.  It
has NOTHING to do with how the data field is encoded.