# Reading old data from a reformatted drive

**URL:** <https://boards.straightdope.com/t/reading-old-data-from-a-reformatted-drive/221773>\
**Category:** Factual Questions\
**Created:** [December 30, 2003, 3:06pm UTC](https://boards.straightdope.com/t/reading-old-data-from-a-reformatted-drive/221773 "2003-12-30T15:06:01Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![El\_Zagna](https://avatars.discourse-cdn.com/v4/letter/e/d2c977/32.png) [@El\_Zagna](https://boards.straightdope.com/u/El_Zagna)\
**Post date:** [December 30, 2003, 3:06pm UTC](https://boards.straightdope.com/t/reading-old-data-from-a-reformatted-drive/221773/1 "2003-12-30T15:06:01Z")

</div>

An [MSN article](http://www.bcentral.com/articles/komando/126.asp) today talks about the need to clean your hard drive before getting rid of your old computer. In it they claim that

> [@](#):
>
> Reformatting will keep most people out of your old files. But specialized shareware exists to reclaim files after reformatting.

They also state that

> [@](#):
>
> There are several programs that write gibberish to the hard drive. They promise that nobody will be able to find your files after the software is utilized.

My question is this - how is it possible to read old data off a hard drive after it has been reformatted or had gibberish written on it?

---

<div class="post-metadata">

**Author:** ![micco](https://avatars.discourse-cdn.com/v4/letter/m/5f8ce5/32.png) [@micco](https://boards.straightdope.com/u/micco)\
**Post date:** [December 30, 2003, 3:44pm UTC](https://boards.straightdope.com/t/reading-old-data-from-a-reformatted-drive/221773/2 "2003-12-30T15:44:43Z")

</div>

There are a number of ways to read data off a drive. In simple cases of erased files, the data is still there but the pointers to the files have been removed from the disk index. This is analogous to “deleting” a library book by removing it from the card catalog - it’s still on the shelf, but no one will find it unless they go look for it. There are plenty of tools which find and recover lost clusters of data on disks and allow you to recreate files.

A more thorough deletion will overwrite the deleted files so these unreferenced areas no longer contain valid data. There are a number of tools for doing this, and many of them contain procedures which meet certain standards for deletion, which specify things like overwriting several times with alternating sequences of all ones, all zeros, pseudo-random data, etc. This will prevent simple file recovery tools from working.

The next level of analysis is to look at the values stored along the sides of each individual bit. When the write head makes a pass along the disk, it may be offset slightly in one direction or another. For example, if the write head was offset lightly left when it wrote your data and was then offset slightly right when it overwrote it, the original data is still recoverable by analyzing a narrow band on the left side of the bit’s allotted space. The tools to do this are expensive and require dedicated hardware to analyze the disk bit by bit, but these methods can be used to recover data from disks even when they have been overwritten and physically broken. That’s why the US government standards for deleting secure data involve not just overwriting but shredding the physical disk and incinerating.

This has been discussed before, so a search on this board will reveal several threads about recovery methods.

---

<div class="post-metadata">

**Author:** ![Mangetout](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mangetout/32/19_2.png) [@Mangetout](https://boards.straightdope.com/u/Mangetout)\
**Post date:** [December 30, 2003, 3:52pm UTC](https://boards.straightdope.com/t/reading-old-data-from-a-reformatted-drive/221773/3 "2003-12-30T15:52:22Z")

</div>

I think that with very specialised techniques, it is even possible to recover data from the sides of a track if the heads are aligned on each pass - as far as I understand it, overwriting data doesn’t completely change the state of the magnetic domain, but that writing the new state can do something akin to squeezing the existing magnetic state to the edge of the domain.
