Re: Importing .TAP files from internet into PET using datassette

From: Marko Mäkelä <>
Date: Wed, 7 May 2014 11:20:34 +0300
Message-ID: <20140507082034.GA1689@x60s>
On Tue, May 06, 2014 at 11:57:42PM -0700, chitselb wrote:
>I've got the PETdisk from which has 3.5mm audio jacks for
>tape in and out.  Even better, it has a 2K microSD card reader that pretends
>to be a disk drive!  You might be able to use audacity on a laptop to send a
>.wav file to that audio jack, if you are really committed to your first

For what it is worth, around 2002 I implemented WAV PCM output in the 
"c2n" tool that I wrote when developing the C2N232 adapter.

I do not clearly remember how I tested the WAV output. It could be that 
I connected the audio jack to an analog tape deck, recorded the sound 
and then put the tape in a Commodore 1531 tape drive. I probably used a 
PET 8032SK in this experiment.

Around that time, I also played with TAPir. I digitized some tapes with 
it, using an analog tape deck for input. I think I also tried mtap 
(connecting a 1531 to the parallel port).

One tape from 1985 was in bad shape, but I was able to manually 
resurrect the contents by tweaking the *.cas file in Emacs. You know, 
the *.cas format represents the raw bytes in a Commodore tape recording 
just like *.d64 represents the raw bytes in a Commodore GCR disk image.  
In this format, there are 2 copies of each block (2 copies of the 
192-byte file name block followed by 2 copies of the file contents 

Sadly, few emulators support the native high-level tape format; they go 
with the lower-level *.tap (ok, fastloaders will need it) or with the 
completely arbitrary *.t64 format.

Side note: You can clean up *.tap files by quantizing the pulse widths.  
This will allow them to compress better.

The pulse widths when executing SAVE on the C64 is varying a lot.  IIRC 
the PET had the cleanest pulse widths (measured with the 8MHz AVR 
AT90S2313 in the C2N232).

When tuning the tape head, it can be useful to look at the pulse width 
histograms. Back when I made these experiments about 12 years ago, I 
wrote a simple program to display the histograms in real time.


       Message was sent through the cbm-hackers mailing list
Received on 2014-05-07 09:00:03

Archive generated by hypermail 2.2.0.