18 Nov
2009
18 Nov
'09
6:36 a.m.
I'm working on a Tomy Tutor tape decoder because, well, no one else is. To
that end, this weekend I managed to crack the encoding and now have a
primitive tape decoder that reads an AIFF audio file and spits out bits for
a higher-level decoder to process. To date I can now see the bit pattern for
the GRAPHIC paintbox, and can even do rudimentary decoding of BASIC programs.
So far so good.
However, playing back that exact uncompressed 44.1kHz 16-bit mono AIFF into
the Tutor doesn't work (before you ask, the Tutor's tape inputs are mono).
The Tutor doesn't see the sync mark, and never loads the "tape." I recorded
this a few times, making sure that all the output got on the audio file,
and no dice. I also played with line levels and varied the output volume
level through all the fine steps the Mac would let me step, and the Tutor
just sits there.
Have people discovered any gotchas in general about using PC audio files
to load and save from tape, besides the obvious ones like don't compress,
etc.? Any suggestions about how to make the Mac's output more acceptable
to the Tutor?
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- Adore, v.: To venerate expectantly. -- Ambrose Bierce ----------------------
18 Nov
18 Nov
8:04 a.m.
On 17 Nov 2009 at 22:36, Cameron Kaiser wrote:
Have people discovered any gotchas in general about
using PC audio
files to load and save from tape, besides the obvious ones like don't
compress, etc.? Any suggestions about how to make the Mac's output
more acceptable to the Tutor?
A low-pass filter? How does the waveform look on a 'scope?
--Chuck
2:04 p.m.
Have people
discovered any gotchas in general about using PC audio
files to load and save from tape, besides the obvious ones like don't
compress, etc.? Any suggestions about how to make the Mac's output
more acceptable to the Tutor?
A low-pass filter? How does the waveform look on a 'scope? 
2:52 p.m.
-----Oorspronkelijk bericht-----
Van: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org] Namens Cameron Kaiser
Verzonden: woensdag 18 november 2009 15:05
Aan: cctalk at classiccmp.org
Onderwerp: Re: tape "load and save" to a PC/Mac
The Tutor doesn't save tape as FSK or KCS; it's actually just
pulses, i.e.,
-||-|--|-||-|--|--
It reminds me more of the Commodore Datasette than the Texas
Instruments machines the Tutor was descended from, actually.
The trick with decoding it wasn't per se figuring out the
frequencies, it was figuring out what pulse pattern meant
which bit, and then shuffling the bits around until I got
some sort of framing that made sense. But that's what makes
the playback problem mystifying because this should be very
simple to generate and play back. I don't see what a filter
would do to help here necessarily unless I'm misunderstanding
the alleged purpose.
--
------------------------------------ personal:
http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com *
ckaiser at floodgap.com
-- LOAD"STANDARD DISCLAIMER",8,1
----------------------------------------------
And putting your output on a tape and trying to load that ?
Because if that's a working soluntion a filter should do something..
If there tape is generating only pulses the rise time could be affecting the
smitt trigger or other signal shaping circuit.
Another thing, is it possible the tape set some kind of signal like write
enable or read etc..
And last but not least, some tape stored programs have a kind of lead-in
with enables the computer to sync on the tape signal.
-Rik
> Have people discovered any gotchas in
general about using
PC audio
files to
load and save from tape, besides the obvious ones like
don't compress, etc.? Any suggestions about how to make the Mac's
output more acceptable to the Tutor?
A low-pass filter? How does the waveform look on a 'scope?
3:02 p.m.
And putting your output on a tape and trying to load
that ?
Okay, I'll try that. Have to dig out the tape recorder though :)
Because if that's a working soluntion a filter
should do something..
If there tape is generating only pulses the rise time could be affecting the
smitt trigger or other signal shaping circuit.
If so, though, I should be able to compensate for that. What kinds of things
could I do to the signal to make it more acceptable?
Another thing, is it possible the tape set some kind
of signal like write
enable or read etc..
No, the Tomy tape recorder is a dumb recorder, I'm pretty sure. Also, there
are only in and out data lines on the cable.
And last but not least, some tape stored programs have
a kind of lead-in
with enables the computer to sync on the tape signal.
Yes, I have the full sync mark accounted for, and made sure I gave plenty
of lead time before hitting SAVE and lead time after, so I'm sure I have
all the data.
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- Must have been the collard greens that made me lose control! -- "In the Mix"
4:23 p.m.
If so, though, I should be able to compensate for
that. What kinds of things
could I do to the signal to make it more acceptable?
Have you compared the outputs with an oscilloscope? It would seem to me
that five minutes with a scope would solve any problems you might be
having, as any distortion in the signal caused by digitizing it would be
obvious.
My two immediate thoughts are: if it's "pulses", maybe the quantization
is causing issues; or maybe the sample rate (or some other randomness)
is introducing noise to the signal.
Also, a scope should tell you if there's clipping or other analog
problems (from overamplification, say).
5:34 p.m.
Exactly how long is the lead-in? I know on the Spectrum it is 4 seconds and I believe it
is 2 seconds for the TRS-80. Not sure how long it is for the C64, but I'm sure you
already know that.
My point is that the lead-in might have to be a precise length and you might have it too
long (or too short).
Regards,
Andrew B
aliensrcooluk at yahoo.co.uk
--- On Wed, 18/11/09, Cameron Kaiser <spectre at floodgap.com> wrote:
From: Cameron Kaiser <spectre at floodgap.com>
Subject: Re: tape "load and save" to a PC/Mac
To: cctalk at classiccmp.org
Date: Wednesday, 18 November, 2009, 15:02
And putting your output on a tape and trying to load
that ?
Okay, I'll try that. Have to dig out the tape recorder though :)
Because if that's a working soluntion a filter
should do something..
If there tape is generating only pulses the rise time could be affecting the
smitt trigger or other signal shaping circuit.
If so, though, I should be able to compensate for that. What kinds of things
could I do to the signal to make it more acceptable?
Another thing, is it possible the tape set some kind
of signal like write
enable or read etc..
No, the Tomy tape recorder is a dumb recorder, I'm pretty sure. Also, there
are only in and out data lines on the cable.
And last but not least, some tape stored programs have
a kind of lead-in
with enables the computer to sync on the tape signal.
Yes, I have the full sync mark accounted for, and made sure I gave plenty
of lead time before hitting SAVE and lead time after, so I'm sure I have
all the data.
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
? Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- Must have been the collard greens that made me lose control! -- "In the Mix"
4:13 p.m.
On Nov 18, 2009, at 9:52 AM, Rik Bos wrote:
And putting your output on a tape and trying to load that ?
Because if that's a working soluntion a filter should do something..
If there tape is generating only pulses the rise time could be affecting the
smitt trigger or other signal shaping circuit.
Another thing, is it possible the tape set some kind of signal like write
enable or read etc..
And last but not least, some tape stored programs have a kind of lead-in
with enables the computer to sync on the tape signal.
This might sound crazy but have you put the output on a tape and then tried bringing the
tape output back into the Mac to watch the pattern?
Rob
Rob Borsuk
email: rborsuk at colourfull.com
Colourfull Creations
Web: http://www.colourfull.com
3:05 p.m.
Hi
It might be the phase of the signal that counts.
The Poly88 used a manchester encoding and It had
a switch to select the right phase ( normal or inverted ).
Dwight
----------------------------------------
From: spectre at floodgap.com
Subject: Re: tape "load and save" to a PC/Mac
To: cctalk at classiccmp.org
Date: Wed, 18 Nov 2009 06:04:41 -0800
The Tutor doesn't save tape as FSK or KCS; it's actually just pulses, i.e.,
-||-|--|-||-|--|--
It reminds me more of the Commodore Datasette than the Texas Instruments
machines the Tutor was descended from, actually.
The trick with decoding it wasn't per se figuring out the frequencies, it
was figuring out what pulse pattern meant which bit, and then shuffling the
bits around until I got some sort of framing that made sense. But that's
what makes the playback problem mystifying because this should be very
simple to generate and play back. I don't see what a filter would do to help
here necessarily unless I'm misunderstanding the alleged purpose.
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- LOAD"STANDARD DISCLAIMER",8,1 ----------------------------------------------
_________________________________________________________________
Bing brings you maps, menus, and reviews organized in one place.
http://www.bing.com/search?q=restaurants&form=MFESRP&publ=WLHMTAG&a…
Have
people discovered any gotchas in general about using PC audio
files to load and save from tape, besides the obvious ones like don't
compress, etc.? Any suggestions about how to make the Mac's output
more acceptable to the Tutor?
A low-pass filter? How does the waveform look on a 'scope?
3:17 p.m.
Hi
It might be the phase of the signal that counts.
The Poly88 used a manchester encoding and It had
a switch to select the right phase ( normal or inverted ).
But wouldn't the sync mark handle getting the proper phase (i.e., the
adjustment to properly read the pulses)? Or do I misunderstand what you're
referring to by phase?
--
------------------------------------ personal: http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com * ckaiser at floodgap.com
-- I would like to achieve [immortality] by not dying. -- Woody Allen ---------
3:28 p.m.
----------------------------------------
From: spectre at floodgap.com
Subject: Re: tape "load and save" to a PC/Mac
To: cctalk at classiccmp.org
Date: Wed, 18 Nov 2009 07:17:21 -0800
Hi
No, it doesn't in the Poly88. Each recorder has a different
number of inversions. The manchester encoding is sensitive
to the phase. The pulses may be a fast rise and slow fall.
This would also make it phase sensitive as well.
Think in terms of a logical inverter only for the audio.
Imagine a pulse detector that is looking for rising edges
and then invert the data. Since it is looking for the
time between pulses, it would be completely wrong.
Dwight
_________________________________________________________________
Windows 7: I wanted simpler, now it's simpler. I'm a rock star.
http://www.microsoft.com/Windows/windows-7/default.aspx?h=myidea?ocid=PID24…
Hi
It might be the phase of the signal that counts.
The Poly88 used a manchester encoding and It had
a switch to select the right phase ( normal or inverted ).
But wouldn't the sync mark handle getting the proper phase (i.e., the
adjustment to properly read the pulses)? Or do I misunderstand what you're
referring to by phase?
3:49 p.m.
-----Oorspronkelijk bericht-----
Van: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org] Namens Cameron Kaiser
Verzonden: woensdag 18 november 2009 16:17
Aan: cctalk at classiccmp.org
Onderwerp: Re: tape "load and save" to a PC/Mac
I think he means, the signal after sync mark should be in fase with the sync
mark or 180 degrees out of fase.
Did you compare the tape signals with the signals you're generating with a
scoop or may be using a spectrum analyser ?
Measuring should give you some clous about the differences between the
original and generated signals.
Is it possible that the sync mark being to long could generate a time-out..
-Rik
Hi
It might be the phase of the signal that counts.
The Poly88 used a manchester encoding and It had a switch to select
the right phase ( normal or inverted ).
But wouldn't the sync mark handle getting the proper phase
(i.e., the adjustment to properly read the pulses)? Or do I
misunderstand what you're referring to by phase?
--
------------------------------------ personal:
http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com *
ckaiser at floodgap.com
-- I would like to achieve [immortality] by not dying. --
Woody Allen ---------
4:40 p.m.
----------------------------------------
From: hp-fix at xs4all.nl
To: cctalk at classiccmp.org
Subject: RE: tape "load and save" to a PC/Mac
Date: Wed, 18 Nov 2009 16:49:47 +0100
---snip---
I think he means, the signal after sync mark should be in fase with the sync
mark or 180 degrees out of fase.
Did you compare the tape signals with the signals you're generating with a
scoop or may be using a spectrum analyser ?
Measuring should give you some clous about the differences between the
original and generated signals.
Is it possible that the sync mark being to long could generate a time-out..
No, I meant just what I said. Since it isn't looking for
a tone, the phase of the pulse makes a big difference.
A "plus to minus" pulse is not the same as a "minus to plus".
Try feeding an inverted signal into your serial port.
It just doesn't work. I doubt there is anything in the
toy to check for correct phase and then invert all the
pulses. It most likely expects rising edges to be rising
and not falling.
Dwight
_________________________________________________________________
Windows 7: I wanted simpler, now it's simpler. I'm a rock star.
http://www.microsoft.com/Windows/windows-7/default.aspx?h=myidea?ocid=PID24…
8:27 p.m.
On Wed, 18 Nov 2009, Cameron Kaiser wrote:
The Tutor doesn't save tape as FSK or KCS;
it's actually just pulses, i.e.,
-||-|--|-||-|--|--
It reminds me more of the Commodore Datasette than the Texas Instruments
machines the Tutor was descended from, actually.
The trick with decoding it wasn't per se figuring out the frequencies, it
was figuring out what pulse pattern meant which bit, and then shuffling the
bits around until I got some sort of framing that made sense. But that's
what makes the playback problem mystifying because this should be very
simple to generate and play back. I don't see what a filter would do to help
here necessarily unless I'm misunderstanding the alleged purpose.
Look at the waveform. In addition to the outgoing waveform that you are
recording accurately, look at the incoming waveform produced by the POS
recorder. Could the Tomy be expecting (and relying) on certain
degradations of the signal? Howzbout a pulse extender circuit or
software?
8:05 p.m.
On Wed, 18 Nov 2009, Chuck Guzis wrote:
A low-pass filter? How does the waveform look on a
'scope?
Doesn't the waveform look a lot cleaner than the waveform coming back from
the OEM recorder?
Years ago, there were some threads here about difficulties trying to use
high fidelity recordings in place of the POS recorder with the TRS80.
--
Grumpy Ol' Fred cisin at xenosoft.com
9:37 a.m.
-----Oorspronkelijk bericht-----
Van: cctalk-bounces at classiccmp.org
[mailto:cctalk-bounces at classiccmp.org] Namens Cameron Kaiser
Verzonden: woensdag 18 november 2009 7:37
Aan: cctalk at classiccmp.org
Onderwerp: tape "load and save" to a PC/Mac
I'm working on a Tomy Tutor tape decoder because, well, no
one else is. To that end, this weekend I managed to crack the
encoding and now have a primitive tape decoder that reads an
AIFF audio file and spits out bits for a higher-level decoder
to process. To date I can now see the bit pattern for the
GRAPHIC paintbox, and can even do rudimentary decoding of
BASIC programs.
So far so good.
However, playing back that exact uncompressed 44.1kHz 16-bit
mono AIFF into the Tutor doesn't work (before you ask, the
Tutor's tape inputs are mono).
The Tutor doesn't see the sync mark, and never loads the
"tape." I recorded this a few times, making sure that all the
output got on the audio file, and no dice. I also played with
line levels and varied the output volume level through all
the fine steps the Mac would let me step, and the Tutor just
sits there.
Have people discovered any gotchas in general about using PC
audio files to load and save from tape, besides the obvious
ones like don't compress, etc.? Any suggestions about how to
make the Mac's output more acceptable to the Tutor?
--
------------------------------------ personal:
http://www.cameronkaiser.com/ --
Cameron Kaiser * Floodgap Systems * www.floodgap.com *
ckaiser at floodgap.com
-- Adore, v.: To venerate expectantly. -- Ambrose Bierce
----------------------
I suppose the Tomy Tutor uses FSK or something like that as codec.
You should find out the base frequentie and use a (analog)low pass filter to
filter out any spikes and smoothing the signal. Old cassette tapes have a
band pass between 200-6000Hz (10k for metal tapes) using a filter would
smooth the signal with more change for succes.
-Rik
8 p.m.
Be warned that I don't even own a PC with audio recording facilities, and
still use real cassettes for this sort of thing. But I can suggest some
things to try..
I'm working on a Tomy Tutor tape decoder because, well, no one else is. To
that end, this weekend I managed to crack the encoding and now have a
primitive tape decoder that reads an AIFF audio file and spits out bits for
a higher-level decoder to process. To date I can now see the bit pattern for
the GRAPHIC paintbox, and can even do rudimentary decoding of BASIC programs.
So far so good.
OK, I think we can assume you've got valid data onto the PC/Mac.
However, playing back that exact uncompressed 44.1kHz 16-bit mono AIFF into
the Tutor doesn't work (before you ask, the Tutor's tape inputs are mono).
The Tutor doesn't see the sync mark, and never loads the "tape." I
recorded
this a few times, making sure that all the output got on the audio file,
and no dice. I also played with line levels and varied the output volume
level through all the fine steps the Mac would let me step, and the Tutor
just sits there.
What I would doo is grab the 'scope. Try loading a program from a real
cassette and see what the signal looks like _inside the Tutor, on the
input socket_. If you can get to the output of the digitising circuitry
(often just a comparator) -- it's not hidding inside an ASIC -- look at
that too.
Then try 'playing' the signal from the PC/Mac into the Tutor. First check
there's actually a signal getting throuigh the input socket (I've lost
cound of the number of times I've had a mis-wired cable, or when a mix-up
of stereo and mono jack plugs ('phone plugs' to you) has shorted out the
signal). See if the signal is about the same amplitude. Make sure there's
no mais-freqeuncy (50Hz, 60Hz, 100Hz,120Hz) ripple on it. If you can
check the ouutput of the comparator stage, does that look much the same?
-tony
5520
days inactive
5520
days old
16 comments
9 participants
participants (9)
-
aliensrcooluk＠yahoo.co.uk -
ard＠p850ug1.demon.co.uk -
cclist＠sydex.com -
cisin＠xenosoft.com -
dkelvey＠hotmail.com -
feedle＠feedle.net -
hp-fix＠xs4all.nl -
rborsuk＠colourfull.com -
spectre＠floodgap.com