Re: Resources required to retrieved wiped disk

Paul Rubin wrote:
"Douglas A. Gwyn" <DAGwyn@xxxxxxxx> writes:
If one instead were to search for the closest fit to {the signal
minus the first-order best fit}, perhaps the remanent magnetization,
i.e. the previous data encoding, would become decodable (albeit with
higher noise level).
Do you mean the first order best fit to the random data that's
overwritten the old data? That requires knowing the contents of the
random data, which is probably not on the menu.

The first order fit for the random data is whatever the standard
read of the random data normally reports.

The reason for "wiping" algorithms overwriting tracks several
times with different random data is to bury the remanent or spill
effect behind several layers of stronger (more recently written)
effects, so it gets lost in the noise as you try to wade through
the signal history.
.

Relevant Pages