Finally, I can put together some info about the recovery process I
went through.
1. I lost power while recording on an HD24 before STOP was pressed.
2. I called Alesis and Robert Ray directed to this HD24 group and
emailed a post about using Linux to recover lost audio.
3. I read most everything but did not comprehend much due to my lack
of experience with Linux.
4. I bought a new hard drive exactly like the one with the corrupt
information.
5. I used the HD24 to format and record blank audio onto the new drive
using the same song name, number of tracks, sample rate and enough
length to accommodate my lost recording.
6. Then I borrowed a pc with Susie Linux and it would never see all
three drives no matter what master/slave/cable select jumper settings
I tried.
7. Then I borrowed an old Emachines pc with 333mhz processor and 64mb
ram running Slackware Linux with no GUI, only command line interface.
It saw all three drives fine using cable select jumper setting.
8. I then ran the dd command using hexdump to view contents of bad
drive. The command I used was this:

dd if= /dev/hdc bs=512 count=1000 | hexdump -C | less

dd is Data dump
if is input file
/dev/hdc tells it to look at hard drive C
bs is block size
count is number of blocks
hexdump tells it to display in hexadecimal format in 16 byte lines
-C says to display any printable characters
less is to display a page at time and be able to scroll forward and
backward
9. I could not figure out where the header ended and audio began.
10. Then I did the same command on the new drive with blank audio and
determined where the header ended. Basically, when the header ended
the code became all zeros because of blank audio. I determined the
audio started at the hex address of 1dff0
11. Then I constructed the copy command as the following

dd if= /dev/hdc bs=256 count=96000000 skip=478 of= /dev/hdd seek=478

dd is Data dump
if is input file
/dev/hdc is hard drive with audio and bad header
bs is block size
count is amount of data ( I estimated 24 gigs)
skip means skip down to 478 and start copying from there
of is output file
/dev/hdd is new drive with blank audio and good header
seek tells it where to start copying from

12 Then I hit enter and saw hard drive activity light. I then left for
the night and returned the next morning to the process being finished.
13. I then did the same dd and hexdump commands saw similar info on
both drives.
14. I then put the new drive into the HD24 and the audio was there
completely.

Most of the time was spent making the computer see all drives at one
time. Once that was done then attempting to figure out the start/stop
points seemed endless until I figured out the hexdump -C command
. It
took a scientific calculator to do the hexadecimal calculations to
determine addesses for skip and seek points.

I hope this helps everyone out there. 
I've realized a couple of things while doing this.
1. I do not feel the hard drive needs to be identical. The new drive
needs to have enough space to accommodate the amount of audio data you
wish to copy.
2. The important part is writing an identical header by recording
blank audio with same sample rate, track numbers and enough length to
accommodate the amount of the original recording.
I thought about offering a service to others in need at some cost but
I don't have the time due to my full time job as chief engineer
at a
public radio station. Email me directly and I will try to help more.

Thanks to all,
Karl Fontenot