Anyone got an A4000 with working PAR card?

[slaapliedje]

First zip is the PAR software, second one is the TBC IV software. I believe the manual is on the Big Book of hardware.
Edit, gonna set binwalk to the task!
2nd Edit: Seems it has a lot of 'Uncompressed Adobe Flash SWF File' in the 4gb image.

[mark_k]

Originally Posted by slaapliedje

How did you image these? I'm going to write the image I made back and make sure that's even working.... I did open up the Hex editor that's built into the PAR test software (which I made ADFs of) and it looked like it had some headers and such on the drive. So that'd be a start in figuring out the format.

I think the PAR software "fakes" a header when you copy files from DDR:. Looking in a hex editor the first three bytes will be P, A, R characters. But that header isn't stored on the IDE drive itself.

If you load the big 1GB HDF into a hex editor, you'll see:
- The first 512 sectors are all similar. The first and last 4 bytes of each contain the sector number (little-endian). The rest of the data looks like random garbage (but the same random garbage) in each sector.

After that the data looks all random, there are long runs of "random" bytes followed by $FF bytes padding to the end of a sector from time to time. Each of those runs could correspond to one (compressed) frame. There's no human-readable header there. But it could well be JPEG-like compression, so maybe it'll eventually be possible to extract some data and add appropriate header bytes to convert it to JPEG.

Another thing to note is, data may well be stored byte-swapped on the IDE drive. So for example, data read by the PAR hardware as (hex) 01 23 45 67 would be in the HDF as bytes 23 01 76 45.

[slaapliedje]

So the image that I made from the drive I got with the order of the PAR/TBC IV does have plain readable text in there. The first sector has a mixed up thing that says something like Personal Animation Recorder, but with some missing characters (Like it has A imation). Then if I search for Spiderman (as that is a partial match of SpidermanIntro that the PAR software sees) I can find that within the hex editor.

I'm not seeing any of that on either the 4gb image or the Titanic one.

Edit: I opened these images on my Linux box. The Personal Animation Recorder text I saw was on the hex editor for PARtest though.
But you're right, there is the potential for the byte swapped, though I've mostly only ran into that on Atari ST formatted disks. Seems to me the PAR is just a raw data block though. In fact 'file' command just shows those images as being 'data'. So everyone seems to be correct in that it's just raw data there, and there should be some sort of file headers. Once I get the two 4gb CF cards in, I can try imaging those up (much easier to swap those back and forth between the Amiga and the PC. Getting that 5.25" drive I have, I know I'm eventually going to bend a pin....)

[Starglider 2]

Wow thanks for all the deep diving here guys. All I can add to the conversation is that the Titanic drive was backed up using WinUAE in Administrator mode. I tried several times and it succeeded each time with the same bytes so seemed successful.

The 4GB drive was sent away for recovery and I'm guessing they used ddrescue.

FWIW I have been able to grab the raw video via S-Video. This goes public Saturday but here is early access for you guys. [ Show youtube player ]

[slaapliedje]

Originally Posted by Starglider 2

Wow thanks for all the deep diving here guys. All I can add to the conversation is that the Titanic drive was backed up using WinUAE in Administrator mode. I tried several times and it succeeded each time with the same bytes so seemed successful.

The 4GB drive was sent away for recovery and I'm guessing they used ddrescue.

FWIW I have been able to grab the raw video via S-Video. This goes public Saturday but here is early access for you guys. [ Show youtube player ]

Huh, I tried loading it up under WinUAE and it just kept bitching that it was NDOS, and for some reason wouldn't let me cancel out of that like it normally does. Did the files actually show up on the desktop at all? Maybe I'm not sure how you image something from within WinUAE.

But you're probably right if they did image the drive using ddrescue, that'd make sense. When I attempted the first time of just imaging it over the 8gb one, the file descriptors were still there for PAR, but when I'd go to play the SpidermanIntro file, it'd just show mostly garbage (which makes sense, copying a 4gb over 8gb). Those file descriptors must have been much further in the drive.

Always fun to go sleuthing for old data!

[Starglider 2]

The disk shows as NDOS as it's in a propitiatory format. It will only read when connected to a real PAR card IDE socket. It won't mount in WB by design.

[johnnybarker]

Hello,

You probably know this so sorry if you do. Once the par software is loaded up then the par drive then appears in wb like any other drive. As you said it needs to be connected to a par card. The data can then be copied straight onto the any other drive from there. Presumably you could copy it to your Amiga work drive or whatever then stick it in a PC then access it from UAE?

[Starglider 2]

Hmm I'll look into that...

[johnnybarker]

Amiga-M once in the par software to get back to workbench and the drive should be present.

[johnnybarker]

I have downloaded your images so I'll give it a bash tomorrow if I get a chance.

[johnnybarker]

I also have a funny feeling that the par software has an export function. I have the manual at home but I am away until tomorrow. Will have a look then and report back.

[slaapliedje]

Originally Posted by johnnybarker

Hello,

You probably know this so sorry if you do. Once the par software is loaded up then the par drive then appears in wb like any other drive. As you said it needs to be connected to a par card. The data can then be copied straight onto the any other drive from there. Presumably you could copy it to your Amiga work drive or whatever then stick it in a PC then access it from UAE?

I'll verify that tonight, but I'm pretty sure when I was looking at it last night, it didn't show up at all on workbench. Could be that when I switched back to workbench, I just didn't see it.

I'm hoping my dump of my drive that came with the card and re-imaging it actually worked. I had it fail twice when doing it under Debian (it'd stop at 80.5 percent then all sorts of weird kernel errors appeared, but under Arch Linux (newer kernel) it imaged again, but I was too tired to hook it back up to verify it worked.

I did get the CF cards sitting on my porch, so once I'm home from the office tonight, I can give it another go.

See what quest we've had set upon us??

[slaapliedje]

Originally Posted by johnnybarker

I also have a funny feeling that the par software has an export function. I have the manual at home but I am away until tomorrow. Will have a look then and report back.

It probably does. The issue I've had so far is actually getting the PAR software to unlock the images. But that could be down to me imaging over the 8gb drive that already had a PAR image on it, with the 4gb and then the 1.4gb hdf, (neither of which I managed to have the PAR software see anything on).

[johnnybarker]

Hmm. I was tinkering with mine a couple of months ago and it definitely appeared as a drive on the wb screen. That doesn't sound like everything is playing as smoothly as it should.

I let you know you if I have any luck.

[slaapliedje]

Originally Posted by johnnybarker

Hmm. I was tinkering with mine a couple of months ago and it definitely appeared as a drive on the wb screen. That doesn't sound like everything is playing as smoothly as it should.

I let you know you if I have any luck.

At the stage where I was flipping between screens, it was with the workbench image that wasn't working quite right. I since installed a fresh 3.1.4, installed BestWB1.2, updated to 3.1.4.1, installed the PAR software, installed the TBC-IV software, unplugged the PAR card, took the pictures I uploaded, plugged it back in, reconnected the cables (one to the IDE, one to the TBC-IV (though they're far enough away from each other I did not use the 50pin short cable, and just used one of the SCSI cables from my flyer) and then the software actually worked correctly, unlike before where it'd give 'Board failure' Not sure if I just didn't have it seated correctly or what. But now that I know the software / hardware works correctly, I'll see if the imaging process worked right! Hopefully my IDE -> CF adapter works correctly, then I won't have to back and forth so much with the hard drive that is already old.
But with that old drive, assuming the dd to / from my Linux box worked...

Anyhow, I'll test this out in a few hours once I'm home from the office... if I can resist watching the Witcher :P

[johnnybarker]

Good luck.

I'm looking forward to watching the video!

[LanceT]

Strange File recovered from the Work Partition.
"seldanes.1"

[Starglider 2]

Interesting. Digital Domain did a lot of commercials. Or it could be stock footage from PAR.

[slaapliedje]

Huh, so got a different hex editor that correctly will open that 4GB image (weirdly I swear I saw something similar on the image I made of mine, but now can't seem to get the same thing.

Anyhow, according to what I can see, was the 4GB image created from a SCSI drive?

[Starglider 2]

Ah yes this is something I only realized yesterday, but the second drive is indeed SCSI. However it does have PVR written on it meaning it likely contains video files. Initially I thought it was for the par card and I confess I did not check the port of the drive as it was sent away to someone else for recovery. But anyway, it says that it has video files on it. That's all I know. Sorry! There are about 10 hard drives in total and I must admit I did get a little mixed up about which port was on which drive.

Sent from my SM-F900U using Tapatalk