Flakey disks

[LordCrass]

I have a disk I'm trying to image (Three Stooges) and while my disks work fine on my 1541, they come through a bit on the iffy side when it comes to reading them with my panasonic drive. I've cleaned them as much as I can and the tracks look otherwise great, but there's still 2 or 3 tracks that come back with a 1 or 2 bit error frequently, and it's visible in the analyzer's flux display.

However, if I keep reading the track over and over, it will sometimes come through without the bit error. Is there a way to save this track back into my .scp image once I get a good read of it?

[admin]

There is currently no way to save tracks in the analyzer. I might add this ability at some point.

Most of these issues have to do with the narrow drive head used with 96 TPI drives. Using a 48 TPI drive is recommended for copying disks for C64, Atari 400/800, TRS-80, TI-99/4A, etc.

[LordCrass]

Hmm, I see a pattern in where the error comes up. This is a V-Max 2 formatted disk, so it uses sector headers that follow the pattern

64 XX YY XX YY .... 46

Where XX and YY are EOR'd together to determine the sector number. In every case where I see a bad bit, it's in the $46 end-of-sector marker byte, which contains the illegal GCR combination of three consecutive 0 bits. And in every case where it was bad, it read in as a $5B instead.

Gauntlet Deeper Dungeons (another one I'm having trouble with getting a decent image from) also uses this $46 byte. Bad Street Brawler uses it too. Most of the other games with this format use $4E as the end marker, likely due to issues with the $46 byte.

[admin]

Hey, send me your disk image where the data is showing up as correct. There is a big window for the GCR conversion, but maybe not big enough for that disk. I want to be able to have a skew that you can set to adjust the decoding for cases like yours, but I need some examples. Thanks!