If i have a audio CD, and use EAC secure mode to copy all the audio data into my computer as one big wav file. ( a cue sheet with wav ). And then compress it with Monkey's Audio and write it on to a CD. If i copy back the monkey's audio file to my computer would that ensure the music is the same quality as i ripped with EAC in secure mode??

yes.
if the cd does not currupt over time.

forgot to tell you:
you can make a sfv file (checksum file) so that you can always check that your files are ok.

Just thought I'd let you know, you probably should use MD5 checksums instead of SFV checksums.

SFV stands for "Simple File Verification" and uses only a 32-bit verification sum. Which means, although it is extremely unlikely you would ever run into one, that there are *numerous* possible files with the same SFV.

MD5 uses a 128-bit verification sum. While it's still theoretically possible to generate two different files with the same MD5, your odds of encountering one just got a whole lot smaller. Like, say, by an order of magnitude.

If you need a free MD5 program for the Win32 platform, there are currently two options:

Command line: http://etree.org/md5com.html
GUI: http://www.md5summer.org/

... and if you're using something other than Windows, there are comparable programs - I'm just not sure where their links are, at the moment. Hope this helps!

- M.

As far as im aware, FLAC incorportes MD5 sum error checking so that you can test the FLAC file at any time to see if it has become corrupt, this takes the same time as a standard decode and is performed if u decode anyways i think,

Just a though,

Cheers,
Kristian

thanks you M,
my paranoia makes me use md5 now.

I did a little searching and found Easy MD5 Creator to be better for me than your program.
it can make files recursive and right-click create/verify.

Are CRC generators really so bad ?
With a perfect CRC generator, a 32 bits key will fail one time out of 4,294,967,296.
If you store 4,000,000,000 of files, and 1 out of 1000 are corrupted, you'll have one chance out of 1000 that the 32 bits key won't tell you.

A 128 bits key should be
log-10(2^128/2^32)= 29 orders of magnitude more secure.

79,228,162,514,264,337,593,543,950,336 times more secure exactly.

I've tested "Easy MD5 Creator". By default it made .md5-files that wheren't compatible with md5sum. But you could select to write "MD5Summer"-compatible files.

Isn't the md5-sum-fileformat standardized? Doesn't all official md5sum-utils (available on linux/unix) write the md5sum-file a special way?

Edit: found a program here that could calculate hashes upto SHA2 512bit, but no automatic verifying. So, does anyone know any program that can verify SHA1 or stronger hashes?

More evidence that crc32 is not enough: the file here contains some different test-cases, one is a file with 41 t's (41_T_S.TXT) that will compute to equal crc32 of another file consisting of both t's and q's mixed (41_TQ_S.TXT), and if all q's where replaced by s' it would still compute to the same crc32! I was very surprised by this, md5 is definitely the way to go...