LebGeeks

royal_titanium

Member

Compression

I need a software that compresses (high compression) data for backups before I format my pc.

mir

Member

Re: Compression

What Operating system ?

high compression ? so regular softwares won't do ?
how much data do you have ?

are you willing to pay or just free software ?
you have to test your compressed files before formatting you don't want to get a Corrupted file message

maybe,lezim tist3ir external hard disk w khlosna

Ragnar78

Member

Re: Compression

7Z has the highest rate compression (its the name of a tool and the compression alogrithm)
You can find this type of compression in PowerArchiver and 7Z, Xarchiver (the last 2 are free from sourceforge.net)

arzleb

Member

Re: Compression

Search for Genie Backup Manager or Norton Backup Manager.

They both can back up your computer with high compression and then put it on an external Hard disk or DVD. format your PC and install again the Genie or Norton and recover your backup as it was before format.

rolf

Member

Re: Compression

Windows has a "backup and restore" wizard. Type start->run and "ntbakcup"
I know that's not ur question but maybe this info could help.

Ad for compressionI use winRar instead of winzip, better compression - but 7Z might also be better, never tried it.

mir

Member

Re: Compression

well.. my personal feedback :

I really liked 7zip.. works great and simple
good performance
using it as my Personal compression software
at least till i hear about something new

LebaneseChiphead

Member

Re: Compression

Would U software guys please explain how the compression work excatly? how does the file reduces in size to almost half ? or in other words how does that format of the compressed file differ from the format of  uncompressed file? or simply where is the trick in that? why do'nt  we always use compressed files?? why not make that an industry standard?  what are the disadvantages of a compressed file? is there any?

Thank U
Chips IMR

Ps, Mir. Please notice I am in a good mood Today :)

rolf

Member

Re: Compression

Would U software guys please explain how the compression work excatly? how does the file reduces in size to almost half ? or in other words how does that format of the compressed file differ from the format of  uncompressed file? or simply where is the trick in that? why do'nt  we always use compressed files?? why not make that an industry standard?  what are the disadvantages of a compressed file? is there any?

Thank U
Chips IMR

Ps, Mir. Please notice I am in a good mood Today :)

I'm gonna give you an example.
Consider the following sentence:

"How does compression compress the file and why is the compressed file smaller."

Lets rewrite it using less letters. The first thing we do is revrite is like this:

"How does +ion + the * and why is the +ed * smaller.+=compress, *=file"

See it's already smaller. What we did is find repeated patterns and use a "dictionnary". That is one method of compression.

another example, consider this binary data (for example in a bitmap image file):
000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000

How do we compress that?
simple: "repeat 0 a 100 time"

That is another method, and I'm sure there ary many other methods, and maybe some complex math involved.

I hope that was clear and instructive enough.

The reason why we dont always use compression is that hard disks are cheap and compression uses CPU and memory resources, as you can see, processing is required to convert between compressed and uncompressed formats.

Yet you have the option in windows to use a compression at the file system level, but i dont recommend it, it's slower then the normal filesystem and the gains are ridiculously low.

Another disadvantage of compression is that they are more sensible to data corruption. If a text file gets corrupted, you will only loose part of the text data that is corrupted. If a compressed file gets corrupted, you are likely to loose access to the whole file altogether.

Compression is used on the web too, many servers serve pages in a compressed format which is read and decompressed by your web browser (without you knowing anything of it)

Last edited by rolf (November 29 2006)

mir

Member

Re: Compression

I think rolf gave a short idea
if u are more intrested.. i will do a google search and post u some links
yey! now u are intrested in software stuff ... early sign of wanting to cross to the software side

Ps, Mir. Please notice I am in a good mood Today

well ur today is couple of days ago.. lol so i donno what mood u are in when u read this
but anyway.. i don't wanna ruin ur future-mood by giving you a lecture about how great and mighty are the software ppl
think of it like we can make 10 Litre of water fit into 5 Litre Gallon (example)
it is like magic for u hardware ppl .. isn't ?
so i don't wanna make u wonder how it looks like on the software side with all those mighty ppl around , cuz i don't wanna ruin ur day
Hope u keep the  good mood

Mighty MIR

Last edited by mir (December 1 2006)

mahdoum

Member

Re: Compression

There's also Huffman trees (I'm studying it now). Instead of using the standard 8bit per character what the huffman tree does is give the most frequent letters in a document for example the shortest bits per character and it gets bigger as the letter becomes more ferquent. The technique uses binary trees and creates a unique 01 combination for each letter.

It brings down the size because each character now has (most of the time) less then 8bits, and the most frequent letters have shorter lengths.

LebaneseChiphead

Member

Re: Compression

Mir/Mahdoum,
Thank U both for ur replies, I was just wondering since all the compression does is uses 0101 which is binary why can't we just use "hex" as  the time scale and that will save a lot of 0's and 1's. example : "A" in binary = "1010" but in hex  A=A.

Another point is if we use loops as You suggested "Write 0 a 100 times" I understand every loop or if statment will add a  single time clock" Tick Peroid" to the bus, which will result in a slower performance. Do You both  Agree??

Now :

Mighty Mir,  Thank U ,  I loved  UR example,  it was awesome:
"think of it like we can make 10 Litre of water fit into 5 Litre Gallon''

So, Forgive me  Mighty one ,  We hardware engineers see the invisible and do the impossible :)

Chips IMR

battikh

Member

Re: Compression

Mir/Mahdoum,
Thank U both for ur replies, I was just wondering since all the compression does is uses 0101 which is binary why can't we just use "hex" as  the time scale and that will save a lot of 0's and 1's. example : "A" in binary = "1010" but in hex  A=A.

it's simple:
0101 in binary has a size of 4bits
A in hex has a size of 1 byte=8bits
so 0101 takes less space on disk.

i hope i am not saying stupid and not paying attention to something obvious  ma baddeh edbahdal, it's 1:35am and god knows how my head is functioning right now

LebaneseChiphead

Member

Re: Compression

Battikh,
Not true , A hex number like the number "A" which is really the number "10" in decimal, is still a 4 bits number.

Actually:
A Nibble is 4 bits
A Byte is 8 bits which is 2 Nibbles...... can be 2 hex numbers though  example: "171" in decimal = "AB" in Hex , we just saved a bit... went down from 3 bits "171" to 2 bits "AB".
Dword is 2 bytes
Quadword is 4 bytes
Chunk is 8 Bytes.
Cache line use to be 4 chunks.....32 bytes
And sometimes these size  definition change with the technology available.  :)

Chips IMR

karim

Member

Re: Compression

but i don't see why would you want to use compression with the price of storage nowadays..
you can get a 1TB array for around 400$ or smtg...

nuclearcat

Member

Re: Compression

karim - if u have broadband 256k, u can have with compression ~330K. :-)

About main subject. I recommend bzip2 or winrar.

mir

Member

Re: Compression

sorry for taking this topic as a joke - i apologize for the above reply

well there are 2 types of compression called lossy and losseless compression

i donno if u noticed that when working with compressed files
usually a file with images ,  becomes much smaller than other data
that is due to the lossy compression
that is when the compression allows to loose some of the data
that is totaly affordable and do-able with images
cuz the human eye can miss some pixels without noticing

the loseless compression does not allow data to be lost
cuz every bit is important and meaningful

so beside the stuff mentioned above one way that compression works is loosing some of the data

I think there are lots of algorithm for the compression
i have to do a search myself on them
cuz i am in complete ignorance abut that topic

battikh

Member

Re: Compression

Battikh,
Not true , A hex number like the number "A" which is really the number "10" in decimal, is still a 4 bits number.

i was talking about the letter A, the letter A is coded on 1 byte=8bits in ascii. and even if u code letter A=hex A, letter B=hex B, after F, you will need 2 hex symbols=8bits.
i just re-read my post, and i understand that u did'nt understood what i really meant  it is understandable exactly as u mentionned it, bass that was'nt my point

LebaneseChiphead

Member

Re: Compression

No Problem Battikh :)

Mir.
"that is due to the lossy compression
that is when the compression allows to loose some of the data
that is totaly affordable and do-able with images
cuz the human eye can miss some pixels without noticing"

Okay here is my 2 cents: when data is unstable or "Loose" mean U will lose some bits from UR data load, In Hardware  something called "CRC" is used , which is I believe  is "Cyclical Redundancy check" All it does is validate the authication of the data, In other words: any bit that is dropped will be replace with the proper value of that bit  through that Correction procedure.

Any Way, I just like to wish U all happy Holiday and a have good happy new year :)
Chips IMR