My sympathies. I have hundreds of reels of the crap I have to deal with. You have to supply a LOT of air for it to have any effect. The humidity needs to be low, though that is generally a side-effect of getting a chamber up to 55deg C. Not that I have any personal experience with. The brands pointed to by the spectrumdata page are the WORST of the worst (ie. silver label memorex circa. mid-80's) The only person I know who claims to have 'solved' the recovery problem uses custom transports and industrial-sized tape ovens. He is completely unreliable, though, and has personally ripped me off for several thousand dollars worth of tape transports and heads. The two failure modes I've seen on binder is layer-layer adhesion, which can be reduced through baking, and binder deposit if the tape EVER stops moving during reads. The tape should be baked, cooled, then retensioned on a transport with no read-write head, then read using the best transducer you can find WITHOUT STOPPING. There is a high probablity with Memorex tape that even after all that the head will still drop out from binder clog. Analog data recovery and DSP techniques have been used successfully to recover 7 track tapes that were unreadable using the data separators on the drive. Magnetorestrictive transducers like the IBM 3480/3490 heads are more sensitive than conventional inductive heads, but don't map exactly to the track layout of 7 and 9 track drives.