Could you multiply hd size with just more hd heads?

[trend]

I was thinking.. There are tons of programs out there that can undelete programs (even if they havebeen dumped out the recycle bin). I know how this process works.. deleting just means deleting the entry in the file allocation table.. but some programs can even undelete a program that has been written over.

Also, I know the gov and foreiseinc labs can recover data that has been written over many time. they just use higher power magnets that what you used to delete the file.

Then, with the 2 above statements true (or even just one), wouldn't it then be possible, if you had a hd, with 2 different strength heads (the strong one to read "read-only files.. becuase if you write with the strong one, this will erase the contents that was written with the weak head) you could effectively double your hd size? You would have to have 2 File allocation tables that is..



Just wanted toget that off my chest 🙂

 

[itachi]

Originally posted by: trend
I was thinking.. There are tons of programs out there that can undelete programs (even if they havebeen dumped out the recycle bin). I know how this process works.. deleting just means deleting the entry in the file allocation table.. but some programs can even undelete a program that has been written over.
Also, I know the gov and foreiseinc labs can recover data that has been written over many time. they just use higher power magnets that what you used to delete the file.

higher power magnets would erase the data.. what they're using are extremely sensitive read heads.

Then, with the 2 above statements true (or even just one), wouldn't it then be possible, if you had a hd, with 2 different strength heads (the strong one to read "read-only files.. becuase if you write with the strong one, this will erase the contents that was written with the weak head) you could effectively double your hd size? You would have to have 2 File allocation tables that is..

the cost to develop something like this would be insanely high and would require read/write heads that are 100 times more accurate than the current ones. the 2 read heads would have to be able to read not only their logic 1.. but their logic 1 plus the other's logic 1. the 2 write heads would have to be able to write their logic 1 onto the disk without taking away any of the magnetism from the others logic 1.

ok that's a bit confusing.. try this:
with respect to writing...
if the weak write head writes at 0.5 tesla and the strong head writes at 1.0 tesla.. if the strong head writes a 1.. when the weak head tries to write a 1, the current inside the magnet won't be strong enough and will instead demagnetize the strong head's write.
with respect to reading..
if the strong head already wrote.. when the weak head tries to read, it'll read that the area is magnetized so a 1 will be read.. which would be incorrect. so the heads would have to be configured to read the magnetism of that area in teslas and the controller would have to subtract the strong head's write magnetism from the whole magnetism.

not using any of the heads for writing completely defeats the purpose of your post.. without a write head no information will ever be written for the strong head to read.

 

[trend]

The strong head would have to write it's read-only data first then, the weak head can write whatever then read.

 

[f95toli]

It doesn't work that way. You can not write one thing with a strong field and another with a weak.
You would never be able to recover data that has been written over from an "ideal" HD with only two states (1 or 0).
HDs operates by changing the state of magnetic domains and they should only have two states (up and down). The magnetic field strenght does not matter as long as you "flip" the whole bit, the "weak" field would therefore just erase the data written by the "strong" field.

Real HDs are not ideal and that is why "goverment labs" (and quite a few commercial companies, there is nothing secret about the technology) can recover data, there will always be "traces" of old data in the magnetic layer because the new data has no flipped all the magnetic domains that make up a single bit, if you have a senstive magnetometer you can therefore read out the old bits.

This is a simplified explanation of course, there are also other more sophisticated methods.

 

[itachi]

hmm.. forget what i said hahah.

f95toli.. so logic 0 is a negative phase and logic 1 is a positive phase?

 

[Falloutboy]

ok then instead of only righting 1s and 0s to the drive right it in 0-3, where they would stand for 2digit combonations of 1s and 0s. for instance.

0 would really mean 00, 1 would really mean 01, 2 would really mean 10, 3 would mean 11. this would double the capicty of the drive and in my mind wouldn't be that complex.

 

[itachi]

Originally posted by: Falloutboy
ok then instead of only righting 1s and 0s to the drive right it in 0-3, where they would stand for 2digit combonations of 1s and 0s. for instance.

0 would really mean 00, 1 would really mean 01, 2 would really mean 10, 3 would mean 11. this would double the capicty of the drive and in my mind wouldn't be that complex.

that doesn't change anything. 2 and 3 would still require a magnetic field associated with them. it'd be the same as having 01-01.

 

[kpb]

Originally posted by: trend
I was thinking.. There are tons of programs out there that can undelete programs (even if they havebeen dumped out the recycle bin). I know how this process works.. deleting just means deleting the entry in the file allocation table.. but some programs can even undelete a program that has been written over.

Also, I know the gov and foreiseinc labs can recover data that has been written over many time. they just use higher power magnets that what you used to delete the file.

Then, with the 2 above statements true (or even just one), wouldn't it then be possible, if you had a hd, with 2 different strength heads (the strong one to read "read-only files.. becuase if you write with the strong one, this will erase the contents that was written with the weak head) you could effectively double your hd size? You would have to have 2 File allocation tables that is..



Just wanted toget that off my chest 🙂

Wouldn't work. The reason that the government and others can recover hard drives is they can more accurately read the fields on the drive. When your writing to the drive it's not truely binary. It's not really either a 1 or a 0 it's some value inbetween. It's a .95 or a .05. Typically speaking the drive will read anything above a point as a 1 and anything below another point as a 0 and anything between as an error. So maybe .9+ is a 1, .1- is a 0 and anything between them is an error. Well what happens when they recover the data is they go and look at it and record it as exactly a .974 or a .922 Well the first might be a 1 that was written over by a 1 and the second might a 1 that was written over a 0. You hd would see either as a 1 and wouldn't be able to read the previous version. Additionally the hard drive can be thought of as writing a track like on a record. The exact location in that track isn't going to be exactly the same each. It might be a little further in or out each time it rights. Once again it has particular tolerances about how close it has to be to the exact right spot when it reads and writes to be able to work correctly. The equipement used for recovery can be positioned much more precisely and can say the top half was written to a 1 and then the bottom half was written to a 0 since the top 10% is still a 1 while the rest of it is a 0.

The general idea is just that equipement exsists that is much more sensitive that the heads used in HD's and can be used to read more detailed information. I'm also pretty sure that the stuff doesn't just come off like you'd read it off the hd normally. i'd expect that there is large amount of analysis needed to extract all the different peices of data and figure out the previous written over data.

The problem with using something like this to increase hd capacity is that the equipement the government or labs use to recover this info is much more expensive than could be practical in a hd. HD manufuactures are doing thier best to increase the sensitivity of thier read/write heads with in thier cost and other limitations. The increase in sensitivity is seen as the increasing density on hd's we've seen that increases thier capacity and speed.

 

[Matthias99]

Originally posted by: itachi

Originally posted by: Falloutboy
ok then instead of only righting 1s and 0s to the drive right it in 0-3, where they would stand for 2digit combonations of 1s and 0s. for instance.

0 would really mean 00, 1 would really mean 01, 2 would really mean 10, 3 would mean 11. this would double the capicty of the drive and in my mind wouldn't be that complex.

that doesn't change anything. 2 and 3 would still require a magnetic field associated with them. it'd be the same as having 01-01.

You could, however, use read/write heads that work on more than just two field levels. As the last post above describes, hard drives are really analog devices used to store digital data. It's just that they treat values below a certain threshold as a "0", and above a certain threshold as a "1". It might be something like:

write "0" as field strength 0.0
write "1" as field strength 1.0

field strength 0-0.4 = 0
field strength 0.41-0.59 = unknown
field strength 0.6-1.0 = 1

There's no reason you couldn't store two bits in each field, and do something like this:

write "00" as field strength 0.0
write "01" as field strength 0.33
write "10" as field strength 0.67
write "11" as field strength 1.0

field strength 0-0.2 = 00
field strength 0.25-0.45 = 01
field strength 0.55-0.75 = 10
field strength 0.8-1.0 = 11

But doing this reliably requires read and write heads that are at least 4x more sensitive/precise than just storing a binary value does. At the moment, at least, the added cost of doing this is not worth it -- it's cheaper to build two normal hard drives than one double-density one.