Identifying and recovering programs from tapes

Question

A couple of days ago, I was given some old tapes with programs written on them. However, I am having some trouble with them.

I do not know what computer they were programed for. After recognising the ZX Spectrum preamble in some of them, I proceeded to obtain their .tzx file. They all work fine but the following two, that I have not been able to recover (continuous Tape Errors):

Tape 1

Tape 2

On the other hand, I also have the following tapes, which were not written for the ZX Spectrum. I have not able to identify the computer they were programed for. On Tape 3, the preamble is easily heard. Tape 4 starts at around 0:40. Both of them should have been written for the same computer (they were both sides of the same tape):

Tape 3

Tape 4

Any help or advices for recovering the first two tapes and for identifying the language of the latter ones will be highly appreciated.

UPDATE:

• Tape 1 seems to have a copy of a Chess game called El Turco, with an unknown program called HEADER related to recording unprotected headers into tapes (?, I do not know what than can mean or be useful for). Link to its ROM:

HEADER.tzx

• Tape 2 is a damaged tape of LEMS TAPE RECORDER 5

• No special updates for either Tape 3 or Tape 4; unknown compuerts. They seem not to be for ZX80, ZX81, Spectrum, Amstrad CPC, Atari, MSX nor Apple II.

UPDATE 2: Thank you very much for the support and the ideas! Now, all of the tapes have been either identified or restored but Tape 3.

These days we are suffering some inconvenience with our FTP server, I am sorry. Here are some pictures of the wavform of that recording:

It seems very strange to me that it does not seem to use any type of FSK modulation. Nevertheless, I hope this can help narrow down the different possibilities.

Thank you again!

Answer 1

A partial answer that I nevertheless hope contributes:

It doesn’t sound like an Acorn machine because the data isn’t in chunks, and probably isn’t an Amstrad because that uses the same physical encoding as the Spectrum so the lead-in tone for samples 3 & 4 would be the same as for 1 & 2.

By ear, my best guess for Tape 4 is a Commodore but I couldn’t tell you which — it sounds like each file is the same tones repeated twice with a very brief bridge, which is something Commodore did for error detection. Statistically the Commodore 64 is most likely based on sales figures.

Tape 3 sounds like it’s for a different machine again, but I can’t quite place it. The ZX80 and ZX81 leave tiny pieces of silence between bytes, which I couldn’t discern, and the MSX sounds much closer to the Spectrum end of the scale with regards to harshness.

Is there any chance it was an Oric saving in slow (/more reliable) mode? That uses collections of 4 or 8 cycles of tone at the same frequency, which would sound more like a soft babbling than the usual one or two cycles of higher baud rate schemes.

Of course, there were almost as many micros as audio tapes, and I’m completely oblivious as to the encoding used on most of them; other options may be more relevant.