All news

Aug 27, 2017
Category: All news
Posted by: FUJI
  • version 1.05
  • small bugfixes, mostly regading conversion from binary files to Turbo Blizzard tapes
  • files with unknown extension are now treated as binary files instead of audio
Jul 3, 2016
Category: All news
Posted by: FUJI
  • a new turbo system support: Low Silesia Turbo 2000 (lowsil2000). This is almost identical with cs2000, the only difference is the length of piloting signal
  • a fix to --endtime option, end time is now determined more precisely
  • a change to output files naming scheme was made in the routine which avoids overwriting existing files (filename_1.log, filename_2.log ... instead of new_filename.log, new_new_filename.log ...)

Jun 26, 2016
Category: All news
Posted by: FUJI
  • small bugfixes
  • a new option "--pilot" which allows specyfying the length of piloting signal for conversion from binary files (bin/xex/com/bas...) to standard tapes in cas/wav/hex format
  • each input file can now have it's own set of conversion options. You can specify options for input file specific options by separating them from filename with colon, for example to create cas file from hybrid wav file (starndard + turbo format) you can use the following syntax:

    a8cas-util.pl conv hybrid.wav:endtime=65 hybrid.wav:starttime=65:turbo=turbo2000:amplify=2 hybrid.cas
Jun 3, 2016
Category: All news
Posted by: FUJI

Addedd possibility to convert binary files to standard tapes.

Now You can do this:

          a8cas-util.pl conv loader.bin program.xex program.cas

or this:

          a8cas-util.pl conv loader.bin program.xex program.wav

(Note however, that init calls which require longer gaps between blocks are not supported. Yet.)


Jan 25, 2015
Category: All news
Posted by: FUJI
It should be released very long time ago.
  • added TapeWiz support (very basic, this format is checksumless)
  • added generic PWM support, any PWM encoded recording can be decoded into pwml-only data chunks. This is similar to fsk-only decoding for FSK encoded tapes. In fact this mode can be used to archive any Atari tapes (FSK tapes too) and to archive tapes dedicated for any other computers (for example I have successfully tested it with TRS-80 recordings).
Sep 15, 2013
Category: All news
Posted by: FUJI

Here comes version 1.00. Don't be so excited ;), it is just a next number. Changes introduced in this release follow:

  • Multiple input files can be specified. Name of output file from now on is mandatory. The complete command now looks like this:

    a8cas-util.pl <command> [options] <input file> [input  file ...] <output file>

    You can mix different types of input files (audio, cas, hex).

  • PWM (turbo): edge detection algorithm is now simpler, cleaner and probably somewhat faster
  • support for Auto Turbo decoding was added
  • some bugfixes
  • the design of web page was changed (using CMS now)
Oct 30, 2011
Category: All news
Posted by: FUJI
  • FSK decoder is now more precise, even very short signals are sometimes important
  • very long series of raw FSK signals are now supported, they are correctly stored in CAS/HEX files
  • thanks to above two and several smaller changes tapes in Speeder Tape (Speeder 1400) format can be converted to CAS/HEX
  • PWM decoder was significantly improved, options "--amplify", "--highpass", "--lowpass", are no longer so important (though they are still usefull for recovering poor quality tape dumps). The most important addition is automatic (dynamic) adjustment to the amplitude of a signal, when option "--mindiff" is absent. From now on 16-bit samples are used, remember about it when using "--mindiff", previous values must be multiplied by 256
  • timings for all turbo formats were revised, now they are based on code of loaders
  • support for Hard Turbo system was added (--turbo hardturbo)
  • CTRL-C handling was fixed/improved
  • short (several bytes long) data blocks at the end of standard tapes are not being ignored anymore
  • (Examples need updating...)
Jul 31, 2011
Category: All news
Posted by: FUJI
  • average baud rate in audio files produced from cas/hex files is now much more accurate (i.e. number of bits divided by time)
  • several bugfixes (like writting to closed log file)
Oct 17, 2010
Category: All news
Posted by: FUJI
  • added support for Turbo 6000 decoding
  • fixed, tuned and improved Turbo 2000 decoding
  • added option --fskonly; it is intended for recovery purposes - for decoding small fragments of recordings into "fsk " chunks
  • removed workaround for bug in Ecasound versions earlier than 2.7.2 - use Ecasound 2.7.2 or newer if You want to use --starttime
Aug 16, 2010
Category: All news
Posted by: FUJI
  • some bugfixes and other internal changes (endless loops in FSK decoding routines should be finally eliminated)
  • proper checksum verification for AST and UM
  • proper checksum verification for Turbo ROM (only for recordings used with cartridge, it is more complicated in case of tape loaders)
  • turbo decoding: option "--invert" usually will not be needed anymore - now the script automatictally detects, which edge is first. Autodetection does not work for Turbo ROM, sometimes it can fail also for other formats.
  • turbo decoding: parameters tuning for czechoslovakian formats
  • FSK decoding: baud rate stored in cas files is now computed as an average baud rate of the whole data record; this should allow to achieve better synchronization with left audio channel (when it exists)
  • improved informations displayed by "info --detail"
  • syntax of pwms block has changed in HEX format - hex files in examples section were updated accordingly
  • switched from "pwmi" to "pwml" according to previous news entry
Aug 12, 2010
Category: All news
Posted by: FUJI
  • quick update - the name of "pwmi" block was changed to "pwml" ("levels" intead of "impulses"), to avoid confusion with the meaning of word "impulse" in PWM definition. Almost certainly no tape images were using this type of block until now, so the change should not break anything.
    However this information is important for for those, who already use the CAS extensions proposed here in their software. Please update Your programs or give me a reasons for not making the mentioned change.
  • updated hex files on examples page (format of pwms block has changed in hex files)
Jun 1, 2010
Category: All news
Posted by: FUJI
  • corrected new FSK decoding subroutines; effectiveness is almost as good as it was in version 0.94, but there are still some minor problems - sometimes the script enters endless loop, it is also a bit too sensitive to changes in frequencies of FSK signals
  • changed the appearance of status/progress line
  • fixed the command "info", detailed informations were slightly improved
  • addedd checksum verification for Turbo 2600 tapes
  • addedd additional type of checksum for FSK encoded tapes, custom loaders use it sometimes instead of default OS-based checksum; it is chosen automatically when needed
  • added support for more Czechoslovakian turbo formats: Super Turbo (csst), Turbo Tape (cstt) and B-Tape (csbtape); there was very little testing done, because I have no access to real tapes recorded in these formats
  • dropped commands "split" and "merge" - HEX files are good enough for any manipulations
  • a number of other changes/fixes/tunning in the code
Apr 30, 2010
Category: All news
Posted by: FUJI
  • 0.95 will follow soon, I must finish several things before that
  • subroutines handling tapes in standard format were completely rewritten. It was needed for support of the additional kind of tape protection - data blocks built up from fragments of different baud rates. The ability to recover heavily broken tapes is a bit worse now (unfortunatelly), so please pay a bit more attention to sampling quality. Older version(s), which work better in really hard cases, will remain available for download.
  • option "--autoadjust" is no longer needed for FSK encoded tapes
  • option "--nofsk" (was undocumented) has been transformed into "--allfsk". Normally a8cas-util.pl stores only significant FSK signals in output file (cas, hex or audio). If You want all "junk" to be stored also, for debugging or recovery purposes, You must specify "--allfsk".
  • the new command "squarify" was added. It can be used to convet original FSK ecoded signal into square wave, in which zeros and ones are represented by signal levels instead of different frequencies. It works like the old shell script, which started it all (available for download too). It differs from the effect caused by "--squarify", because the output is unaffected by any additional processing. You can use both files (created by "squarify" command and with "--squarify" switch) to precisely adjust IRG lenghths and baud rates to values matching original tape.
  • renamed the name of turbo format czech2000 to cs2000 (it was developed in former CzechoSlovakia)
  • Removed last addition to specification of "pwms" block - it seeems that it won't be needed at all
  • updated examples to match changes in HEX format
Apr 10, 2010
Category: All news
Posted by: FUJI
  • let's name the new release: version 0.94 ;)
  • a new prerequisite: you need the additional perl module, Audio::Ecasound, it can be installed with the help of CPAN or downloaded and installed manually from the following location: http://www.cpan.org/authors/id/B/BO/BOWMANBS/
  • conversion to audio: addedd ability to produce audio files containing both FSK and PWM encoded parts in one step (for tape images holding loaders in standard format and main programs in turbo format)
    Because of that feature the script from now will run only on unix-like systems (anyway - I doubt anyone would use it on Windows)
  • added possibility to specify start and stop time for processing audio files (options --starttime and --endtime)
  • Turbo Rom (original format) is now fully supported. Remember to sample Your tapes properly.
  • Added support for "Unknown1" turbo format (name: unk).
  • Added support for Czech version of Turbo 2000 format (name: czech2000)
  • Changed the type of highpass filter (associated with --highpass option) to different one which works better
  • improvements in checksum calculations
  • improvements in raw FSK signals handling
  • FUJI blocks between parts of the recordings can be automatically added (option --autofuji); it is probably a bit buggy and based only on gaps between data blocks
  • added option "--squarify" for FSK encoded tapes - it can be used to produce audio files, in which "mark" and "space" tones are represented by low and high levels of signal instead of different frequencies; this is mostly for debugging purposes
  • Turbo 2000 (K.S.O) tape images can be converted to binary files (.xex)
  • binary files (.xex, .com) can be converted to Blizzard Turbo or Turbo 2000 format
  • Some more internal changes, tweaks and tunning.
  • Small change in "pwms" block definition
  • The layout of the web page has changed
  • New examples added
Mar 7, 2010
Category: All news
Posted by: FUJI
  • Fixed error in Turbo Blizzard parameters
  • Added ability to make monophonic tracks when converting to audio
Mar 6, 2010
Category: All news
Posted by: FUJI
  • Turbo Rom format is now supported. The most important part is making a proper dump of tape to audio file. Conversion works fine for good quality dumps. See examples.
  • Highpass filter is now optional. It breaks things too often.
  • Several bugfixes and other small changes.
Feb 28, 2010
Category: All news
Posted by: FUJI
  • Added highpass filter for all turbo formats. It centers signal towards level 0, thus allowing more amplification without loosing information. Conversion from audio to cas/hex can be more reliable now. Cutoff frequency is specific for every format.
  • Resonant lowpass filter can now be used for all turbo formats. Cutoff frequency is specific for every format. This filter amplifies narrowest impulses (usually they are also weakest) and blocks some of high frequency noise.
  • Support for Turbo Rom slowly improves.
Feb 25, 2010
Category: All news
Posted by: FUJI
  • Sample rate calculations for turbo formats during conversion to audio files shoud work correctly from now on. Sample rate specified as command line option (--samplerate) will cause appropriate recalculation of widths of impulses.
  • "--invert" should finally do what it is supposed to do. Note, that when You specify "--invert" to convert audio -> cas, then conversion cas -> audio without "--invert" will produce the wave which looks like the original
  • small modifications of HEX format
Feb 22, 2010
Category: All news
Posted by: FUJI
  • Added support for HEX files (CAS in human readable and editable form), both reading and writing to such files is supported.
  • Full support for conversion between any of formats: audio, A8CAS, HEX (any direction)
  • Preliminary support for conversion from audio/A8CAS/HEX to binary files; currently basic Turbo Blizzard programs can be converted to xex (files in DOS FMS format), tapes in native Atari format can be converted to .xex, .bin (one part boot programs) or .bas (programs written in basic)
  • Added support for ATT (Atari Turbo Tape) and UM (Unerring Master), which are the clones of AST.
  • Many fixes, improvements and tweaks
Feb 16, 2010
Category: All news
Posted by: FUJI
Added primitive wave shapers (option --shape) allowing to choose sine-like or sawtooth-like shape of wawe in output file created during cas -> audio conversion (the default is to create square wave). For turbo formats only.
Feb 14, 2010
Category: All news
Posted by: FUJI
  • Fixed some settings for turbo formats
  • Check for existance  of the given input file
  • Added optional resonant lowpass filter. It selectively amplifies weak parts of the signal without clipping the rest of it. For now works well with AST format.
  • 'pwmc' chunk is now stored according to description
  • FSK - fixed detection of "long zeros" in some situations.
Feb 13, 2010
Category: All news
Posted by: FUJI
  • the license was decided (GPLv2)
  • possibility to specify checksum type in command line was added; some turbo formats may use more than one algorithm of checksum calculation
  • possibility of inverting the signal (--invert) was re-activated
  • forgotten description of --mindiff was added to the manual

Next page: Documentation