Regenerating tapes .. even when "load error? ready.

davidv_ · Post by **davidv_** » Wed Apr 26, 2006 6:28 pm

Hi.

Did anyone ever did some extreme tape salvaging using PC sound editors and heavy DSP? Theres one particular tape that i need to convert and sadly only the caracter redifinition sub program was salvageable.

The deal would be to digitize and volume correct with envelopes all analog "dips" and more advances things, fix sudden pitch changes. Since im a PC audio guy i could go ahead and try (need to buy a better tape player first)

So am i wasting my time?

nbla000 · Post by **nbla000** » Wed May 03, 2006 1:17 am

But tape works in real Vic-20?
If yes my experience suggest you to use if possible a x1531 cable or similar adapters, it's more simple and useful, take a look in other posts.

Mayhem · Post by **Mayhem** » Wed May 03, 2006 3:52 am

Indeed, that's the first question to find out the answer to: can you get it to load on a real Vic20? Once I know that, then I persevere trying to make an image of it

davidv_ · Post by **davidv_** » Wed May 03, 2006 8:53 am

Mayhem wrote:Indeed, that's the first question to find out the answer to: can you get it to load on a real Vic20? Once I know that, then I persevere trying to make an image of it

Nope not even on real vic.

But i managed to salvage a few things now using Tape64 and a high quality tape player (Yamaha Natural Sound deck).

I had to create volume ramps and a few tricks in Sound Forge to achieve some backups. I now have to find an utility to fix a very odd variable DC offset/volume loss in one part. (I think ill try high pass filter then volume ramp)
Another thing that was obvious is that one section of tape had two consecutive headers so tape64 was all confused. (fixed this)

Tape64 has a nice -e setting which helps you recover the "broken" data too.
I ran Tape64 in Visual Studio 6.0 and stepping through the code, but it lacks comments on the structure of the signal.

If i understand correctly, the data is actually loaded twice for integrity checks, but is the second pass a mirror image of the first? bit mask swapped? I do see (graphically) where the dip is and its in the first section.
Just copying the second section on top of first doesnt work as is.
But the -e option doesnt say which of the two passes it restores...
If you know that you have better chances of success with one or the other.. then .. ok ill hack at the code more when i have the time.

Its just a hobby of mine, i surely spent more time trying to get the data back than coding the game in 1982

)