• Please review our updated Terms and Rules here

Need help identifying a tape game's platform

LocalH

Member
Joined
Sep 2, 2007
Messages
12
Location
Kingsport, TN
Over at Lemon 64, somebody posted a WAV recording of a cassette tape that had no label, and they need to identify which system the game is for. Here's the thread, I'll quote the first post here since it has a link to the WAV:

Macc said:
Dropbox link;
https://www.dropbox.com/s/cd89dn35taxtcgy/unknown.wav

Its not C64, Speccy or Amstrad, appreciate it if someone could tell me. Files 25mb btw in .wav

The person just now posted a bit better recording with no distortion, here's the link.

They are saying the game is Red Moon by Level 9, which came out for the following systems on tape: Amstrad CPC464, Atari 400/800, BBC 32K, C64, Enterprise 64, Memotech 500/512, MSX 64k, and Spectrum 48k. They originally ruled out C64, Speccy, and CPC, but I'm not sure if the newer recording would change anything. I had found a site that also mentioned a PC release, and just on the off chance that it may have been released on cassette for the 5150, I looked into that. Here is my research, copied from my post on the Lemon64 thread:

LocalH said:
Looking at the IBM PC 5150 cassette format, this isn't it. The header tones for each block are shorter on the PC than the ones present in this file (which actually seem to be variable length, first one is approx 3s, second through seventh are approx 1.6s, eighth one is almost 6.4s, ninth one is back to 1.6s, tenth one is back to 6.4s, last one finishes the file with no actual data after it at a length of just under 3s), and the pulse widths don't match (IBM PC pulse widths are 1ms for a 1 bit and 0.5ms for a 0 bit, this file seems to have long pulses of approx 0.4ms and short pulses of roughly 0.2ms). For completeness' sake, I'm going to document the frequencies of the pulses as well, at least approximately, in case it helps anyone determine which computer it's from. The long pulses rest at approximately 1250Hz, while the short ones appear to be at almost 2500Hz. This comes really close to the modified KCS used on the Beeb and Electron, which runs at 1200 baud raw bitrate, with 0 bits being one 1200Hz cycle and 1 bits being two 2400Hz cycles (meaning that both 0 and 1 bits would take up roughly the same amount of time in the recording, even with the different frequencies and thus different pulse widths).

Doing some reading, I can confirm it's not BBC or Electron, however. Both of those machines store data in 256-byte blocks with sequence numbers, and those blocks have the 2400Hz tone in between. This file does not have that, in fact it has a quite large block next-to-last.

I haven't inspected the newer recording yet. Does anyone here have any idea which system this game might be for? With the first recording, I was unable to get any tools to make any sensible conversion to any of the various tape image formats. The emulators I have tried that allow direct WAV input (like Altirra) gave no results as well.
 
Last edited:

LocalH

Member
Joined
Sep 2, 2007
Messages
12
Location
Kingsport, TN
Yeah, I was just doing a bit of analysis on it and I'll agree with you. A poster on Lemon was able to partially convert the WAV to a CAS file and get it to display a filename in an MSX emulator, before throwing an I/O error. This is just probably due to the massive amount of noise in the WAV and bad arguments for the WAV-to-CAS conversion. I found a known-good CAS of the game and converted it to WAV, then compared the waveforms. Other than the CAS-to-WAV being cleaner and having shorter gaps, I think you are indeed correct and that it is an MSX game.

Here's my screenshots if you're curious:
Second posted WAV file, without clipping
Converted from CAS

Data blocks look identical, same relative sizes and same number of them. The bit of repetitive data in the last third or so of the last data block seems to match up well, too. Between my analysis and yours, I'm willing to call it MSX. Thanks for the help!
 
Top