View Single Post
Old 2006-02-11, 02:18 PM
Five's Avatar
Five Five is offline
TTD Staff
Join Date: Oct 2004
Location: Canada
Re: Checksums Tutorial (.st5, .ffp and .md5)

Originally Posted by U2Lynne
A couple notes from the Checksums section (comments in bold):

.st5 aka SHNtool md5. this is a checksum taken from the decompressed audio only. the advantage to this is that it works with any lossless codec at any compression setting. This is the preferred checksum type to be included with shn or ape files.

.ffp aka FLAC fingerprints. These checksums are virtually identical to .st5. some differences are explained below, under "Verifying a checksum file" and "Creating a .ffp file".

For audio, I highly recommend using FLAC and including .st5 and .ffp no wholefile .md5 necessary. Except the 32-digit checksum is going to be exactly the same in the st5 and ffp file. There is really no reason to include both types of checksums for fingerprints, just one type.

I added in "This is the preferred checksum type to be included with FLAC, SHN or APE files." in the st5 section.

The main reason is that when you double-click a .ffp for testing using TLH, it extracts the checksums hidden in each FLAC file's header and compares that to what is in the .ffp file to see if it matches (takes about three seconds). Using the .ffp this way for testing does not test the audio content whatsoever (which could be corrupt after a successful .ffp verification). A second stage of testing is necessary in this case, it is "test" mode, which decompresses the audio to a temp file and generates a fresh hash from this and compares that to the the checksum hidden in the header. So, only after the two-stage test the files are verified. Double-clicking on a .st5 for verification using TLH results in the audio content of the files to be tested being decompressed to a temp file (temporarily) while a fresh hash is generated and compared to the corresponding checksums contained in the .st5 file. So the .st5 test is only one step and verifies both (a) that the files will decompress properly (as with "test" mode) and (b) that the checksums contained in the .st5 file match hashes freshly generated from that decompressed audio. When a .st5 verifies 100% okay everything is tested except the header isn't looked at, thus the .ffp. The .ffp is admittedly a little superfluous, but since it is only 1k and takes about three seconds to generate and verify its not really much to ask.

oww, my brain!
Checksums Demystified | ask for help in Technobabble | ttd recommended free software/freeware webring
shntool tlh eac foobar2000 spek audacity cdwave vlc

Originally posted by oxymoron
Here you are in a place of permanent madness, be careful!
Reply With Quote Reply with Nested Quotes