Date   : Thu, 21 Oct 2004 14:41:43 +0000
From   : "W.Scholten" <whs@...>
Subject: tape formats etc.

L.S.

This was in my tmp box for ages (many moths), in unfinished form as I
just don't get around to it. Comments welcome but I may respond with a
long delay.


Atom half bits etc:
=====================================
Greg Cook wrote:

> Acorn would surely have had good reason to record this half bit.  I'm not
> an Atom expert but I believe the cycle length is controlled by a VIA
> timer, so the CPU should be idle and can run the CRC routine during the
> stop bit.  Therefore I assume the half-bit is deliberate; either to
> improve the reliability of recording on tape, or to be compatible with an
> earlier Acorn machine that lacks a VIA.

No, the Atom writes with CPU timing loops. These check a 2400Hz
reference signal that's connected to the 8255's PC4. There's also *NO*
CRC calculation on the Atom, just a checksum.

This is another bad thing about the Atom's tape loading system (which is
truly awful, worse is probably impossible), and I've had a couple of
instances of reasonable quality tapes being loaded in without problem,
but giving a bad program.

I know about the tradeoffs and choices of course, and my comment about
the mess simply meant that it complicates matters, and it's just ugly.

I'm not sure where the extra 1/2 bit comes from. The putbyte routine
seems to be writing just 10 bits so seems to come from another part of
it or maybe a missed start with the CPU loop (I'm not an expert in the
Atom OS routines)

> Historians, though, might be interested in the IBG
> length variations, as they tell apart one copy of a file from another, and
> reveal whether MODE 7 was being displayed when it was recorded...

Hmm, I don't think think there's anyone crazy enough to be interested in
anything as banal as that :)

Something much more interesting, is the background or before/after
noise. For example: I have a tape of snapper on which a person is
actually playing snapper, and someone else in the backround is putting
stuff in boxes or something like that. Could this be an Acornsoft
recording with a microphone left on? It's probably not a home made tape
(original tape, broken off protection bits, no trace of glue over them
to record despite the protection, and the audio/data looks original).

[ -> now we know why Acornsoft games were more expensive, they were
playing games most of the time in the packing dept. so labour costs
there were huge ;-) ]

You are *never* going to be able to put all those things properly into
any tape format, because you're never sure the tape recording might
contain interesting stuff or not. On superficial hearing it might be
noise, on more close hearing, it may be machine noise (tape player or
something). On further hearing, it might be a faint SOS encoded by ticks
on a table somewhere who was a captive in the Acornsoft recording
studio...

No, I'm not kidding, I'm serious here because talk of 'perfect
recordings' or 'perfect emulators' is nonsense. As you say, there are
tradeoffs, and I'm going further than what you're saying; In fact, I've
actually done so long ago when I talked of transistor level emulation;
this was not a joke but serious; emulators are never perfect because not
all the parameters are known. Example: Did I mention temperature related
behaviour changes of machines, like the TV output going B&W after some
time due to overheating? Not even perfection if the abstract working
models of CPUs and other ICs are used are ok, because behaviour differs
for example with busloading (values that stay on the CPUs bus for
differing amounts of time; I've seen this while working on my SWR
detection program)).

What about storing the magnetic tape type used? The composition of the
plastic casing? The recording level? The distortion level of the waves
which can give clues to the reproduction machines or the BBC micro
(maybe one with a flaky serial ULA) used to record the master tape? This
might be of interest to historians, and I'll bet that's more interesting
to them than which mode a program was recorded in.

Another problem is accuracy of the recording and tape degradation. How
sure can I be that single waves are ok? Maybe what I see is a deformed
other wave. CRCs on the BBC won't cover this... (this is esp. for
begin/end of a wave section before/after silence)
=====================================

Regards, 

Wouter
--
BBC/Atom/magazine scans:
     http://8-bit.summerfield-technology.co.uk/