The Name of the disk (Was: Disk imaging with IMD - question

Fred Cisin cisin at
Wed Aug 9 14:08:15 CDT 2017

The structure used on most diskette tracks is based on an IBM design 
(3740, etc.), and was referred to as "IBM format".  BUT, the general 
public think that "IBM" means the machines that followed the 5150, so "IBM 
format" is oft misconstrued to mean PC.  To avoid ambiguity, I sometimes 
call it "IBM/Western Digital style format".

It includes substantial structures, including sector headers, gaps, 
write splice areas, etc., which require a substantial overhead  (about a 
quarter? of the disk capacity), so decisions about how many sectors of 
what size can have an effect on the final usable capacity.  In spite of 
that, in the interest of "uniformity", it is commonplcae to refer to 
different types of diskettes by one of their commonly used formatted 
capacities  ("360K", "1.2M", "720K", "1.4M")  THAT certainly doesn't 
remove all confusion - for example Apple "400K" and "800K" are the same 
diskettes as IBM "720K"

8" disks used a different jacket for double sided disks than single sided, 
with the index hole aperture in a different location.  That let the 
computer recognize which type of diskette was inserted, and therefore to 
refuse to cooperate if the user asked for a different format than that 
diskette had been intended for.

5.25" dropped that, and also reversed the write-protect notch to a write 
enable notch.
Single sided and double sided 5.25" diskettes are interchangeable, and 
single sided ones can be flipped over to use the other side (single sided, 
of course) by punching additional holes in the jackets (cf. Berkeley 
Microcomputer "Flip Jig")
5.25" diskettes were once known as "mini-floppy".

Besides GCR (Commodore, Apple][, Macintosh 400K and 800K, etc.), there are 
a variety of obscure alternate possibilities, such as Amiga MFM but 
without the IBM/WD structures, hard sector, NRZ, FSK?, etc.)

>  the drive and disks are double-sided double density. Are you 
> saying that's quad density?

Would it work to kill off people that say "quad density"?

Originally, 5.25" disks were single sided 35 track, soon changed to 40 
track.   Diskette was 300 Oersted.
Capacity depended on formatting choices, typically between 80K and 100K.
These disks were an FM ("Frequency Modulation") recording. 
There was NO mention of "density", although some engineers might call it 
"half density", since there is one bit of data for every two pulses/flux 

Apple chose to use GCR for their 35 track single sided disks, resulting in 
about 140K.

Then there was MFM ("Modified Frequency Modulation").  The premise was 
that clock pulses/flux transitions weren't reaally needed between adjacent 
data pulses/flux transitions.  That put more space between the 
pulses/transitions, which meant that the data transfer rate could be 
increased (they doubled it), getting about twice as much data per track.
About 1.5? pulses/transitions per data bit.
Depending on format choices, typically between 150K to 200K capacity.
Instead of just calling it "MFM", the marketing people called it "DOUBLE 

AFTER that, they renamed "FM" to "SINGLE DENSITY".  That means that if you 
look back, historically, you'll find earlier mentions of the phrase 
"double density" then the earliest mentions of "single density".
(The same historical principle applies to the phrases "World War TWO" V 
"World War One" (which had previously just been "the great war"))

But, there was also single sided and then double sided.
Depending on format choices, between 240K and 400K for DSDD.

BUT, the marketing people at Intertec (Superbrain) chose to call DSDD: 
"QUAD density".
They were the only ones who did that.

Soon thereafter, 5.25" disks came out with 96tpi, instead of 48tpi, 
resulting in 80 tracks instead of 40 tracks.  The "density" on each track 
was not affected.
Depending on format choices, between 640K and 800K.

The marketing people of many/most? computer companies called THAT "QUAD 
DENSITY".  (DSDD, with 80 tracks, instead of 40 tracks)
I think that that was a very stupid naming choice.

So, is "quad density": 40 track DSDD (Superbrain)?
or 80 track DSDD (MANY CP/M computers)?
or 1.2M (DSDD with twice the linear density/data transfer rate, where they 
really did get 4 times as much data per track)?

BUT, it gets WORSE!!
Intertec (Superbrain) started making 80 track DSDD (800K?) available. 
But, they had already used "QUAD DENSITY" to refer to 40 track DSDD!  So, 
they called the 80 track DSDD, "SUPER DENSITY"!   If that wasn't bad 
enough, they abbreviated "Super Density" as "SD".
OK, is "SD" FM/single density,
Thankfully, NOBODY else was THAT stupid.

I have a special fond spot in my heart for Intertec.  At NCC (National 
Computer Conference) in 1983, I stopped by their booth to ask some minor 
questions about their formats for XenoCopy.  (They had multiple formats, 
with their own unique names, inverted the data bits, short-changed 
some of the IBM/WD style gaps, etc.)  The honchos in the Intertec 
booth could not comprehend why ANYBODY could possibly want to copy files 
from any Superbrain disk to any other format of disk other than to steal 
their proprietary software!  So, they said that if I included Superbrain 
formats in XenoCopy, that they would sue me!   That night, was the first 
time that I added a disk format in a hotel room.   The promised "free ink" 
publicity of a lawsuit never happened.  (At one point a Televideo 
executive made a similar promise; that was my second hotel room format)

Later, when "1.2M" came out (80 tracks MFM, with twice the linear 
density/data transfer rate, further doubling the amount of data per 
track), it was called "HIGH DENSITY".
Those disks are 600 Oersted

Single density 3.5" disks never got off the ground.
40 track 3.5" drives were rare (Epson Geneva PX-8)
The original MFM 80 track ones were [rarely] used single sided,
and commonly used double sided ("720K")
Those are about 600 oersted.

Within a few years, they doubled the linear density/data transfer rate of 
3.5", to come up with an 80 track MFM with twice the linear density, 
which they called "1.44 MB".  (1,474,560 bytes)
Using a slightly different diskette, recognizable by and extra "media 
type" hole.   About 720 - 750 Oersted
Using an honest MebiByte (2^20 = 1,048,576 bytes), they are 1.406M, so 
"1.4M" would be a legitimate name for them.  (That also provides the 
trivial advantage of being the same number of characters as "720K")
The only way that you can get 1.44 "MB" would be to creatively redefine 
"MB" to be 1,024,000 (1000 * 1024).  Not valid as either decimal NOR 
binary, and therefore LACKING IN INTEGRITY.

The difference in Oersteds between "1.4M" and "720K" is small enough that 
one can often get away with using the wrong type of diskette.
A good quality "720K" diskette can be used as a low quality "1.4M";
a good quality "1.4M" diskette can be used as a low quality "720K"
Some drives, including many IBM PS/2, ignore the "media type hole", 
and/or a hole can be punched or covered.

Within a few more years, Barium-Ferrite disks were made that permitted 
doubling the linear density and data transfer rate again (with a 
different drive and controller).  2,949,120 bytes (2.8125 MebiBytes) 
"2.8M" would be an honest name for them, so they called them "2.88 MB".
The media type hole was moved slightly, to be able to differentiate only 
by people paying close attention.

They never caught on.

NeXT used the same Barium-Ferrite "2.8M", but called them by their 
UNFORMATTED capacity, and therefore called them "4 MB".

Unformatted capacity of 40 track 300 Oersted 5.25" (all 5.25" prior 
to 1.2M) is about 250K per side,
unformatted capacity of "1.2M" "HD" is about 1.6M,
unformatted capacity of "720K" 3.5" is about 1M,
unformatted capacity of "1.4M" 3.5" is about 2M.

(tangentially, the earliest 5.25" had been 35 tracks at 48tpi, and later 
increased to 40, and Micropolis created a 77 track 100tpi, Victor/Sirius 
used GCR on 80 tracks, 3", 3.25" disks, etc.)

Naming standards of all of them are nonexistent.
As George Morrow said, "Standards are wonderful;
everyone can have a unique on of their own."


More information about the cctech mailing list