Re-reading bad sector disks?

All questions about how to use KryoFlux go here.
Post Reply
Rakki
Posts: 747
Joined: Fri Jun 03, 2011 8:29 pm
Location: Jyväskylä, Finland

Re-reading bad sector disks?

Post by Rakki » Fri Jun 03, 2011 8:42 pm

I tried to read couple problematic Amiga disks that already failed with my Catweasel IV+. Today Kryoflux arrived and tested to read couple working disks successfully. After that tried these problematic disks -> now by some reason Kryoflux doesn't try to read bad sectors too many times. Catweasel Imagetool has an option to read 99 times error sector because sometimes re-reading just passes. I have managed to save numerous disks with this tactic.

Is there any possibility to re-read bad sectors multiple times with Kryoflux software?

User avatar
IFW
Posts: 3079
Joined: Mon Nov 08, 2010 2:42 pm

Re: Re-reading bad sector disks?

Post by IFW » Fri Jun 03, 2011 11:44 pm

Yes, you can set the number of retries as well as number of revolutions sampled. Please see the help or manual for details.

tect
Posts: 33
Joined: Thu Nov 04, 2010 7:58 pm

Re: Re-reading bad sector disks?

Post by tect » Sat Jun 04, 2011 1:53 pm

you can set the number of retries up to 999.
But in my case Kryoflux never was successfull reading a bad track after more than 30 retries without re-inserting the disk, repositioning the heads, using another floppy drive or fiddling with the disk in the drive while the disk has been dumped (described here: viewtopic.php?f=4&t=176).
Using this methods disks which were retried before several times :!: by simply using -t999 could be dumped completely.

Rakki
Posts: 747
Joined: Fri Jun 03, 2011 8:29 pm
Location: Jyväskylä, Finland

Re: Re-reading bad sector disks?

Post by Rakki » Sat Jun 04, 2011 6:41 pm

Thanks, I'll try that on later phase and will read the manual + FAQ. Next week is full of other activities so no way to do more tests for now.

Another thing is to order second 3.5" drive just in case. Current one is Sony MPF920.

After 3.5" drives are tested and they work properly I will order this drive for C64 dumping. I have few original titles waiting for preservation :).
http://cgi.ebay.co.uk/Newtronics-D509V- ... 4cf72374b3

User avatar
prowler
Posts: 8
Joined: Fri May 27, 2011 10:14 pm

Re: Re-reading bad sector disks?

Post by prowler » Mon Jun 20, 2011 10:18 pm

I have a couple of Amiga disks, each with a single sector checksum error, which I would like to make valid images of.

The sectors in question are flaky and give different data on each read.

I have tried using the -t999 option while limiting the start and end track for both disks to that which includes the bad sector in an effort to salvage them, but they are so flaky that it is just not possible to get valid captures even after 999 retries.

it seems like the resulting file includes just that data fetched on the 999th retry, when what would be preferable is an integrated or averaged result to help reconstruct the bad sector.

I have found that the best method in this case is to take several captures with minimum retries and compare the results to identify where the unreliable bytes are and to try to recontruct the rest on a maximum likelihood basis.

I realize that such functionality is probably part of the Analyser software which comes with the commercial KryoFlux package, but I would be prepared to pay extra for a feature like this.

Does increasing the revolutions over which the data is sampled provide some integrating effect? If not, what precisely is the difference between retries and revolutions and what advantage is there in increasing the revolutions in a situation involving flaky sectors, please?

User avatar
IFW
Posts: 3079
Joined: Mon Nov 08, 2010 2:42 pm

Re: Re-reading bad sector disks?

Post by IFW » Tue Jun 21, 2011 10:51 am

Revolutions is the number of tracks during consecutive, uninterrupted streaming of flux transitions.
The higher the number, the better for analysis.
Reading a sector image format defaults to a minimum number of revolutions that would cover an entire track regardless of which position of the track the streaming starts.
Selecting preservation format parameters (currently -i2 only) raises the number of revolutions to 5 among other things; this is of course helps the analyser a lot in correctig errors and understanding content.

DTC can use the additional revolutions as well, during the conversion of sector content - and exactly for finding the best match/combination of data in a single try of reading a track.

Changing the number of retries changes how many times the whole track reading and anaysing process gets retried in case of the format conversion signals a hard error. (something that affects data integrity, rather than something that affects how a system treats data, often protection)

Rakki
Posts: 747
Joined: Fri Jun 03, 2011 8:29 pm
Location: Jyväskylä, Finland

Re: Re-reading bad sector disks?

Post by Rakki » Sun Jul 10, 2011 11:29 am

I have one disk that contains very strange error. Already checked if the Saddam or any other virus caused the error but scanning with latest VirusZ was clean.

DTC doesn't notice it by some reason and scanning succeeds always ok (tried 10 times). Image contains an checksum error on one Protracker module on track 41. Copying the module with Direcotory Opus returns zero size file. Any ideas how to save this unique piece of music?

User avatar
IFW
Posts: 3079
Joined: Mon Nov 08, 2010 2:42 pm

Re: Re-reading bad sector disks?

Post by IFW » Sun Jul 10, 2011 11:51 am

Dtc always deals with the physical content of the disk. Os and filesystem faults are a very different thing and by your description that's what you are dealing with. You could try a program called disk salvage - I think - to fix the file system problem; do it under emulation.

Rakki
Posts: 747
Joined: Fri Jun 03, 2011 8:29 pm
Location: Jyväskylä, Finland

Re: Re-reading bad sector disks?

Post by Rakki » Sun Jul 10, 2011 12:14 pm

IFW wrote:Dtc always deals with the physical content of the disk. Os and filesystem faults are a very different thing and by your description that's what you are dealing with. You could try a program called disk salvage - I think - to fix the file system problem; do it under emulation.
Ok, I'll try it under emulation. I have tried it already with Fixdisk 1.8 and Disksalv 4 on real amiga and making all fix trials on copied disk.

Post Reply