Well, here's something I've never seen before.
You can see a wobble, kind of a double peak, towards the lower left. That's a missing bit.
The bit isn't missing in the classic sense, which would be a volume drop-out. It's missing in the timing sense! As if someone has snipped 1-bit worth of time out of this WAV. It's detectable because the so-called 8N1 serial encoding of tape data (the default for normal beeb blocks) has a start bit (0) and a stop bit (1) surrounding every 8-bit byte. In the image above, it starts with a 2-cycle 2400Hz bit which is the stop bit at the end of a byte. Then there's a 1-cycle 1200Hz start bit for the byte that has the error.
Because the serial format allows this error to be located with some precision, I was able to fix it by looking at the peak gradients at the drop, and looking at how many bits are needed to re-align the start and stop bits. In this instance, the wobble should be 2 bits, "01". The fixes everything, including the eventual block CRC.
Unfortunately, this same situation occurs again (at least once; maybe more) later in the tape, and it's less clear what the missing bit should be. I've not attempted any further fixing at this time.
I was wondering what might cause this. Perhaps a micro-crinkle in the tape is eating a bit's worth of time? Anyone seen that?
Or could there be something in the WAV capture pipeline that isn't keeping up and is losing data?
Would you be willing to capture it one more time just in case? And if that gives the same result, I at least we know. Given that this title is MIA (best we know?), I'd be willing to try and recover the additional missing bits with what we've got. I love a bit of MIA recovery