*** G64 (raw GCR binary representation of a 1541 diskette)
*** Document revision: 1.6
*** Last updated: March 25, 2004
*** Contributors/sources: Markus Brenner,
Immers/Neufeld "Inside Commodore DOS",
Wolfgang Moser
This format was defined in 1998 as a cooperative effort between several
emulator people, mainly Per Hakan Sundell, author of the CCS64 C64
emulator, Andreas Boose of the VICE CBM emulator team and Joe Forster/STA,
the author of Star Commander. It was the first real cooperative attempt to
create a format for the emulator community which removed almost all of the
drawbacks of the other existing image formats, primarily D64.
The intention behind G64 is not to replace the widely used D64 format, as
D64 works fine with the vast majority of disks in existence. It is intended
for those small percentage of programs which demand to work with the 1541
drive in a non-standard way, such as reading or writing data in a custom
format. The best example is with speeder software such as Action Cartridge
in "warp save" mode or Vorpal which write track/sector data in another
format other than standard GCR. The other obvious example is copy-protected
software which looks for some specific data on a track, like the disk ID,
which is not stored in a standard D64 image.
G64 has a deceptively simply layout for what it is capable of doing. We
have a signature, version byte, some predefined size values, and a series
of offsets to the track data and speed zones. It is what's contained in the
track data areas and speed zones which is really at the heart of this
format.
Each track data area is simply the raw stream of GCR data, just what the
read head would see when a diskette is rotating past it. How the data gets
interpreted is up to the program trying to access the disk. Because the
data is stored in such a low-level manner, just about anything can be done.
Most of the time I would suspect the data in the track would be standard
sectors, with SYNC, GAP, header, data and checksums. The arrangement of the
data when it is in a standard GCR sector layout is covered at the end of
this document.
Since it is a flexible format in both track count and track byte size,
there is no "standard" file size. However, given a few constants like 42
tracks with no halftracks, a track size of 7928 bytes and no speed offset
entries, the typical file size will be approximately 333744 bytes.
Below is a dump of the header, broken down into its various parts. After
that will be an explanation of the track offset and speed zone offset
areas, as they demand much more explanation.
Addr 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F ASCII
---- ----------------------------------------------- ----------------
0000: 47 43 52 2D 31 35 34 31 00 54 F8 1E .. .. .. .. GCR-1541?T°?....
Bytes: $0000-0007: File signature "GCR-1541"
0008: G64 version (presently only $00 defined)
0009: Number of tracks in image (usually $54, decimal 84)
000A-000B: Size of each stored track in bytes (usually 7928, or
$1EF8 in LO/HI format.
An obvious question here is "why are there 84 tracks defined when a
normal D64 disk only has 35 tracks?" By definition, this image includes
*all* the tracks that a real 1541 can access, which is at most 42 tracks
and 42 half tracks. Even though using more than 35 tracks is not typical,
it was important to define this format from the start with what the 1541 is
capable of doing, and not just what it typically does. Some 1541 drives may
have problems reading past track 40, and pushing the head past track 42
might be somewhat hazardous to the health of the drive as it could get
stuck.
At first, the defined track size value of 7928 bytes may seem to be an
arbitrary value, but it is not. It is determined by the fastest write speed
possible (speed zone 0), coupled with the average rotation speed of the
disk (300 rpm). After some math, the answer that actually comes up is 7692
bytes. Why the discrepency between the actual size of 7692 and the defined
size of 7928? Simply put, not all drives rotate at 300 rpm. Some can be
faster or slower, so a upper safety margin of +3% was built added, in case
some disks rotate slower and can write more data. After applying this
safety factor, and some rounding-up, 7928 bytes per track was arrived at.
Also note that this upper limit of 7928 bytes per track really only
applies to 1541 (and compatible) disks. If this format were applied to
another disk type with more sectors per track (like the SFD1001 or the
8050), this value would be higher.
In my investigation using MNIB (a utility by Markus Brenner that allows
you to nibble a 1541 diskette to the PC in G64 format) on a cleanly
formatted 1541 disk (using the built-in 1541 format command), I saw the
following numbers, compared with the defaults that MNIB uses:
Track Range Avg Size Tail Gap MNIB
(bytes) (even/odd sectors) Size
----------- -------- ------------------ ----
1-17 ~7720 9/9 7692
18-24 ~7165 9/19 7142
25-30 ~6690 9/13 6666
31- ~6270 9/10 6250
Note that the first size number (7720) is larger than previously
mentioned track size of 7692.. why? Likely the drive that I used to create
and nibble the clean disk was rotating a little bit slower than 300 RPM
(~299 RPM), so more data than normal was stored on each track. I calculated
the percentage difference between my numbers and the established benchmark
of 7692, multiplied all my numbers by this factor, and arrived at the
following chart:
Track Range Size Tail Gap MNIB
(bytes) (even/odd sectors) Size
----------- ------- ------------------ ----
1-17 7692 9/9 7692
18-24 7139 9/19 7142
25-30 6666 9/13 6666
31- 6247 9/10 6250
See how close the real numbers come to what MNIB uses? I can attribute
the differences of a few bytes to my own rounding errors. Therefore I
conclude that the numbers MNIB uses can be taken as the standard that all
1541-compatible G64 tracks should be created with.
All of the above calculations are shown here to establish a safe
benchmark to create G64 images in the event that someday we can copy them
back to a real 1541 disk. If the G64 track size was too large, it might
happen that the track cannot be written back out. By using the above MNIB
track size numbers, this problem should be alleviated.
Below is a dump of the first section of a G64 file, showing the offsets
to the data portion for each track and half-track entry.
00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F ASCII
----------------------------------------------- ----------------
0000: .. .. .. .. .. .. .. .. .. .. .. .. AC 02 00 00 ............????
0010: 00 00 00 00 A6 21 00 00 00 00 00 00 A0 40 00 00 ?????!???????@??
0020: 00 00 00 00 9A 5F 00 00 00 00 00 00 94 7E 00 00 ?????_???????~??
0030: 00 00 00 00 8E 9D 00 00 00 00 00 00 88 BC 00 00 ????????????????
0040: 00 00 00 00 82 DB 00 00 00 00 00 00 7C FA 00 00 ????????????|???
0050: 00 00 00 00 76 19 01 00 00 00 00 00 70 38 01 00 ????v???????p8??
0060: 00 00 00 00 6A 57 01 00 00 00 00 00 64 76 01 00 ????jW??????dv??
0070: 00 00 00 00 5E 95 01 00 00 00 00 00 58 B4 01 00 ????^???????X+??
0080: 00 00 00 00 52 D3 01 00 00 00 00 00 4C F2 01 00 ????R???????L???
0090: 00 00 00 00 46 11 02 00 00 00 00 00 40 30 02 00 ????F???????@0??
00A0: 00 00 00 00 3A 4F 02 00 00 00 00 00 34 6E 02 00 ????:O??????4n??
00B0: 00 00 00 00 2E 8D 02 00 00 00 00 00 28 AC 02 00 ????.???????(???
00C0: 00 00 00 00 22 CB 02 00 00 00 00 00 1C EA 02 00 ????"???????????
00D0: 00 00 00 00 16 09 03 00 00 00 00 00 10 28 03 00 ?????????????(??
00E0: 00 00 00 00 0A 47 03 00 00 00 00 00 04 66 03 00 ?????G???????f??
00F0: 00 00 00 00 FE 84 03 00 00 00 00 00 F8 A3 03 00 ????????????°???
0100: 00 00 00 00 F2 C2 03 00 00 00 00 00 EC E1 03 00 ?????+??????????
0110: 00 00 00 00 E6 00 04 00 00 00 00 00 E0 1F 04 00 ????????????????
0120: 00 00 00 00 DA 3E 04 00 00 00 00 00 D4 5D 04 00 ????+>???????]??
0130: 00 00 00 00 CE 7C 04 00 00 00 00 00 C8 9B 04 00 ?????|??????????
0140: 00 00 00 00 C2 BA 04 00 00 00 00 00 BC D9 04 00 ????+|???????+??
0150: 00 00 00 00 B6 F8 04 00 00 00 00 00 .. .. .. .. ?????°?????.....
Bytes: $000C-000F: Offset to stored track 1.0 ($000002AC, in LO/HI
format, see below for more)
0010-0013: Offset to stored track 1.5 ($00000000)
0014-0017: Offset to stored track 2.0 ($000021A6)
...
0154-0157: Offset to stored track 42.0 ($0004F8B6)
0158-015B: Offset to stored track 42.5 ($00000000)
The track offsets rquire some explanation. When one is set to all 0's, no
track data exists for this entry. If there is a value, it is an absolute
reference into the file (starting from the beginning of the file).
If an image stored here only contains 35 tracks (e.g. a standard 1541
disk), then all the offset values for track 35.5 and higher will be set to
0. This can be used to detect the maximum track count when converting to a
D64 image. Since D64's cannot hold over 40 tracks, and typically only have
35, some information will be lost when converting a G64.
From the track 1.0 entry we see it is set for $000002AC. Going to that
file offset, here is what we see...
00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F ASCII
----------------------------------------------- ----------------
02A0: .. .. .. .. .. .. .. .. .. .. .. .. 0C 1E FF FF ............????
02B0: FF FF FF 52 54 B5 29 4B 7A 5E 95 55 55 55 55 55 ???RT+)Kz^?UUUUU
02C0: 55 55 55 55 55 55 FF FF FF FF FF 55 D4 A5 29 4A UUUUUU?????U??)J
02D0: 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 R??)JR??)JR??)JR
Bytes: $02AC-02AD: Actual size of stored track (7692 or $1E0C, in LO/HI
format)
02AE-02AE+$1E0C: Track data
Following the track data is filler bytes. In this case, there are 368
bytes of unused space. This space can contain anything, but for the sake of
those wishing to compress these images for storage, they should all be set
to the same value. In the sample I used, these are all set to $FF.
Below is a dump of the end of the track 1.0 data area. Note the actual
track data ends at address $20B9, with the rest of the block being unused,
and set to $FF.
00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F ASCII
----------------------------------------------- ----------------
1FE0: 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 R??)JR??)JR??)JR
1FF0: 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 ??)JR??)JR??)JR?
2000: A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 ?)JR??)JR??)JR??
2010: 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 )JR??)JR??)JR??)
2020: 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A JR??)JR??)JR??)J
2030: 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 UUUUUUUUUUUUUUUU
2040: 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 UUUUUUUUUUUUUUUU
2050: 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 UUUUUUUUUUUUUUUU
2060: 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 UUUUUUUUUUUUUUUU
2070: 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 UUUUUUUUUUUUUUUU
2080: 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 UUUUUUUUUUUUUUUU
2090: 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 UUUUUUUUUUUUUUUU
20A0: 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 55 UUUUUUUUUUUUUUUU
20B0: 55 55 55 55 55 55 55 55 55 55 FF FF FF FF FF FF UUUUUUUUUU??????
20C0: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
20D0: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
20E0: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
20F0: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2100: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2110: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2120: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2130: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2140: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2150: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2160: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2170: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2180: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
2190: FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF FF ????????????????
21A0: FF FF FF FF FF FF .. .. .. .. .. .. .. .. .. .. ??????..........
Now we can look at the speed zone area. Below is a dump of the speed zone
offsets.
00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F ASCII
----------------------------------------------- ----------------
0150: .. .. .. .. .. .. .. .. .. .. .. .. 03 00 00 00 ............????
0160: 00 00 00 00 03 00 00 00 00 00 00 00 03 00 00 00 ????????????????
0170: 00 00 00 00 03 00 00 00 00 00 00 00 03 00 00 00 ????????????????
0180: 00 00 00 00 03 00 00 00 00 00 00 00 03 00 00 00 ????????????????
0190: 00 00 00 00 03 00 00 00 00 00 00 00 03 00 00 00 ????????????????
01A0: 00 00 00 00 03 00 00 00 00 00 00 00 03 00 00 00 ????????????????
01B0: 00 00 00 00 03 00 00 00 00 00 00 00 03 00 00 00 ????????????????
01C0: 00 00 00 00 03 00 00 00 00 00 00 00 03 00 00 00 ????????????????
01D0: 00 00 00 00 03 00 00 00 00 00 00 00 03 00 00 00 ????????????????
01E0: 00 00 00 00 02 00 00 00 00 00 00 00 02 00 00 00 ????????????????
01F0: 00 00 00 00 02 00 00 00 00 00 00 00 02 00 00 00 ????????????????
0200: 00 00 00 00 02 00 00 00 00 00 00 00 02 00 00 00 ????????????????
0210: 00 00 00 00 02 00 00 00 00 00 00 00 01 00 00 00 ????????????????
0220: 00 00 00 00 01 00 00 00 00 00 00 00 01 00 00 00 ????????????????
0230: 00 00 00 00 01 00 00 00 00 00 00 00 01 00 00 00 ????????????????
0240: 00 00 00 00 01 00 00 00 00 00 00 00 00 00 00 00 ????????????????
0250: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ????????????????
0260: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ????????????????
0270: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ????????????????
0280: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ????????????????
0290: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ????????????????
02A0: 00 00 00 00 00 00 00 00 00 00 00 00 .. .. .. .. ????????????....
Bytes: $015C-015F: Speed zone entry for track 1 ($03, in LO/HI format,
see below for more)
0160-0163: Speed zone entry for track 1.5 ($03)
...
02A4-02A7: Speed zone entry for track 42 ($00)
02A8-02AB: Speed zone entry for track 42.5 ($00)
Starting at $02AC is the first track entry (from above, it is the first
entry for track 1.0)
The speed offset entries can be a little more complex. The 1541 has four
speed zones defined, which means the drive can write data at four distinct
speeds. On a normal 1541 disk, these zones are as follows:
Track Range Storage in Bytes Speed Zone
----------- ---------------- ----------
1-17 7820 3 (slowest writing speed)
18-24 7170 2
25-30 6300 1
31-4x 6020 0 (fastest writing speed)
Note that you can, through custom programming of the 1541, change the
speed zone of any track to something different (change the 3 to a 0) and
write data differently.
From the above speed zone sample, all the zones use 4-byte entries in
lo-hi format. If the value of the entry is less than 4, then there is no
speed offset block for the track and the value is applied to the whole
track. If the value is greater than 4 then we have an actual file offset
referencing a speed zone block for the track.
In the above example shown, there were no offsets defined, so no speed
zone block dump can be shown. However, I can define what should be there.
You will have a block of data, 1982 bytes long. Each byte is encoded to
represent the speed of 4 bytes in the track offset area, and is broken down
as follows:
Speed entry $FF: in binary %11111111
|/|/|/|/
| | | |
| | | +- 4'th byte speed (binary 11, 3 dec)
| | +--- 3'rd byte speed (binary 11, 3 dec)
| +----- 2'nd byte speed (binary 11, 3 dec)
+------- 1'st byte speed (binary 11, 3 dec)
It was very smart thinking to allow for two speed zone settings, one in
the offset block and another defining the speed on a per-byte basis. If you
are working with a normal disk, where each track is one constant speed,
then you don't need the extra blocks of information hanging around the
image, wasting space.
What may not be obvious is the flexibility of this format to add tracks
and speed offset zones at will. If a program decides to write a track out
with varying speeds, and no speed offset exist, a new block will be created
by appending it to the end of the image, and the offset pointer for that
track set to point to the new block. If a track has no offset yet, meaning
it doesn't exist (like a half-track), and one needs to be added, the same
procedure applies. The location of the actual track or speed zone data is
not important, meaning they do not have to be in any particular order since
they are all referenced by the offsets at the beginning of the image.
Analysing the GCR data stream
-----------------------------
Since the information stored in the track data area is in GCR format, it
is not as simple to analyse as a normal 256-byte sector would be. Here is a
dump of a portion of the GCR data, and what to look for...
00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F ASCII
----------------------------------------------- ----------------
0000: 0C 1E FF FF FF FF FF 52 54 B5 29 4B 7A 5E 95 55 ???????RT+)Kz^?U
0010: 55 55 55 55 55 55 55 55 55 55 FF FF FF FF FF 55 UUUUUUUUUU?????U
0020: D4 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 ??)JR??)JR??)JR?
0030: A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 ?)JR??)JR??)JR??
0040: 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 )JR??)JR??)JR??)
0050: 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A JR??)JR??)JR??)J
0060: 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 R??)JR??)JR??)JR
0070: 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 ??)JR??)JR??)JR?
0080: A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 ?)JR??)JR??)JR??
0090: 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 )JR??)JR??)JR??)
00A0: 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A JR??)JR??)JR??)J
00B0: 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 R??)JR??)JR??)JR
00C0: 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 ??)JR??)JR??)JR?
00D0: A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 ?)JR??)JR??)JR??
00E0: 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 )JR??)JR??)JR??)
00F0: 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A JR??)JR??)JR??)J
0100: 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 R??)JR??)JR??)JR
0110: 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 ??)JR??)JR??)JR?
0120: A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 ?)JR??)JR??)JR??
0130: 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 )JR??)JR??)JR??)
0140: 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A JR??)JR??)JR??)J
0150: 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 R??)JR??)JR??)JR
0160: 94 A5 29 4A 55 55 55 55 55 55 FF FF FF FF FF 52 ??)JUUUUUU?????R
0170: 54 A5 2D 4B 7A 5E 95 55 55 55 55 55 55 55 55 55 T?-Kz^?UUUUUUUUU
0180: 55 55 FF FF FF FF FF 55 D4 A5 29 4A 52 94 A5 29 UU?????U??)JR??)
0190: 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A JR??)JR??)JR??)J
01A0: 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 R??)JR??)JR??)JR
01B0: 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 ??)JR??)JR??)JR?
01C0: A5 29 4A 52 94 A5 29 4A 52 94 A5 29 4A 52 94 A5 ?)JR??)JR??)JR??
We need to establish a "marker" by which one can start to interpret the
data. Always look for a group of at least 10 1-bits (two 'F's in a row and
a bit more), as they establish the SYNC mark. The 1541 actually writes out
a SYNC mark of 40 'on' bits (10 'F's in a row). Note that there are *2*
groups of SYNC marks quite close together, one for the sector header and
one for the sector data. In the above example, there is 2 groups of "FF FF
FF FF FF". The first one is the header SYNC and the second one is the data
SYNC.
An important point here: some documentation refers to the minimum SYNC
mark as being at least 12 bits wide, and claims that one of that size is
still not entirely reliable. Thus Commodore chose to use 40 bits for the
SYNC mark, making it impossible for the drive read electronics to miss.
If the GCR data is not in the standard sector layout, then anything goes
for interpreting the data. If no standard SYNC mark can be found, then
there is no simple way to extract any useful data.
Here's the layout of a standard low-level pattern on a 1541 disk. Use the
above sample to follow along.
1. Header sync FF FF FF FF FF (40 'on' bits, not GCR)
2. Header info 52 54 B5 29 4B 7A 5E 95 55 55 (10 GCR bytes)
3. Header gap 55 55 55 55 55 55 55 55 55 (9 bytes, never read)
4. Data sync FF FF FF FF FF (40 'on' bits, not GCR)
5. Data block 55...4A (325 GCR bytes)
6. Inter-sector gap 55 55 55 55...55 55 (4 to 12 bytes, never read)
1. Header sync (SYNC for the next sector)
The 10 byte header info (#2) is GCR encoded and must be decoded to it's
normal 8 bytes to be understood. Once decoded, its breakdown is as follows:
Byte $00 - header block ID ($08)
01 - header block checksum (EOR of $02-$05)
02 - Sector
03 - Track
04 - Format ID byte #2
05 - Format ID byte #1
06-07 - $0F ("off" bytes)
The "header gap" (#3) is 8 bytes on an early model 1540/1541, but 9 bytes
on a later model 1541 and 4040. The 1541 doesn't read the header gap, but
simply waits it out to write out the sector data. When sector data is
written, the SYNC mark is re-written as well.
There is some controversy over the header gap (#3). Most people assume it
to be 9 bytes of '0x55' characters, but the early 1540/1541 drives used
only 8. This caused an write incompatability with the existing 4040 disks
of the day. In 1541 ROM revision 901225-3 this error was fixed, and now all
drives use a value of 9 '0x55' characters for the gap. The book "Inside
Commodore DOS" by Immers/Neufeld documents the write incompatibilty and
what corruption happens at a low level when writing to a disk with a header
gap of 8 bytes on a disk that normally expects a gap of 9 bytes.
The 325 byte data block (#5) is GCR encoded and must be decoded to its
normal 260 bytes to be understood. For comparison, ZipCode Sixpack uses a
326 byte GCR sector (why?), but the last byte (when properly rearranged) is
not used. The data block is made up of the following:
Byte $00 - data block ID ($07)
01-100 - 256 bytes data
101 - data block checksum (EOR of $01-100)
102-103 - $00 ("off" bytes, to make the sector size a multiple of 5)
The inter-sector gap (#6) can vary from 4 to 19 bytes in length,
depending on the track being read and the manufacturer of the drive. In
tests that the author conducted on a real 1541 disk, gap sizes of 8 to 19
bytes were seen.
The most reliable way to read G64 track data is to read it as bits, not
bytes as there is no way to be sure that all the data is byte-aligned. This
simulates the way a 1541 drive reads data as well as the head only reads
bits as well. The starting location of the track data is know, as well as
the track size so the boundaries of the track limits (start and end) are
obtainable.
What follows is a very simply point-form list of how to read data,
finding sync marks, header blocks and sector blocks.
1. Search for SYNC (at least 10 or more 1 bits)
2. Check for header id after SYNC (GCR 0x52)
3. If header, read the remaining 9 header bytes
4. Decode header and get sector value
5. Search for SYNC again
6. Check for data id after SYNC (GCR 0x55).
7. If data, read and store with previous header.
8. Have we finished reading the track... stop
9. Start over
|