I have some floopies I definitely must recover. I used Kryoflux to save the streams and unfortunately I have some bad sectors/places. But miraculously using a different drive I also got a bad sectors but on different places. From the track/dummy view I see I would definitely be able to complete a fully good image if I combine the images somehow.
I wonder is there any good procedure for that... it would take ages just to copy and paste the exact locations. I suppose I could just extract the data from MFM sectors and compare/coplete them in some smart text editor. But I bet some of you may know some better way.
Here are the HFE images: https://www.dropbox.com/scl/fi/rb44981s ... je8a3&dl=0
I would really apreciate any advice. Thank You.
How to make one good image out of 2 broken ones
Re: How to make one good image out of 2 broken ones
if you use kryoflux preservation raw files you can just replace the bad tracks with the good ones (1 raw file = 1 track).
Re: How to make one good image out of 2 broken ones
Yes, but the damage I see on the floppy visualizations are not in the tracks rather by sectors.
THis means I cant replace the tracks, as there is a damage on each track in each image.
THis means I cant replace the tracks, as there is a damage on each track in each image.
Re: How to make one good image out of 2 broken ones
in this case you can use the copy / past tools in the available in the hxc software
But i have just open your images : there is too much damaged sectors to do this manually.
What is needed here is a merge function to recreate one good image from 2 damaged images. i may implement this.
Re: How to make one good image out of 2 broken ones
Yes, I know about the copy and paste, but when I started to do that, I realized I would spend many hours on it with no certain result.
Then I thought, ok, maybe I can export the actual data out of the MFM and compare the two data sets in some developer tool, there I would probably be able to put them together much faster, but then I'm not sure how to convert them back into HFE image or MFM tracks.
Theoretically, if some part of the floppy were readable in one drive and the unreadable parts were readable in second drive (first was on Alps drive, second on NEC) there might be a drive that will manage to read them all. I have tested another two NEC drives but they read basically the same image as first one. Would You have some recommandation for the drive that could read it all? Somewhere I heard the panasonic were the best for the job, but I have hard time of finding one...
Thank You
Then I thought, ok, maybe I can export the actual data out of the MFM and compare the two data sets in some developer tool, there I would probably be able to put them together much faster, but then I'm not sure how to convert them back into HFE image or MFM tracks.
Theoretically, if some part of the floppy were readable in one drive and the unreadable parts were readable in second drive (first was on Alps drive, second on NEC) there might be a drive that will manage to read them all. I have tested another two NEC drives but they read basically the same image as first one. Would You have some recommandation for the drive that could read it all? Somewhere I heard the panasonic were the best for the job, but I have hard time of finding one...
Thank You
Re: How to make one good image out of 2 broken ones
OK, got another brand of drive (Mitsumi), and the result is much better then on previous drives, but still some damaged sectors. But now the coppy and paste may become viable.
The image is shared here: https://www.dropbox.com/scl/fi/zdjkesan ... 7h8xf&dl=0
But questions still stand:
1) Does anyone know about the best drive that has chance to read this floppy correctly
2) Do You know some good way how to complete one image out of two broken (on different places)?
The image is shared here: https://www.dropbox.com/scl/fi/zdjkesan ... 7h8xf&dl=0
But questions still stand:
1) Does anyone know about the best drive that has chance to read this floppy correctly
2) Do You know some good way how to complete one image out of two broken (on different places)?
Re: How to make one good image out of 2 broken ones
Can you be a bit more specific?
I don't even know how to take just the data out of the MFM. When I'm in track analyzer I can see the data in the right window, but I suppose I should not copy from there.
Or do You think I should edit directly the whole HFE image? Or I should save it first in some data only file format like XML or something like that?
And I also could not find HFscan software, can You send the link. I would probably do that in Notepad++ or HXD hex editor or something like that.
Thank You
I don't even know how to take just the data out of the MFM. When I'm in track analyzer I can see the data in the right window, but I suppose I should not copy from there.
Or do You think I should edit directly the whole HFE image? Or I should save it first in some data only file format like XML or something like that?
And I also could not find HFscan software, can You send the link. I would probably do that in Notepad++ or HXD hex editor or something like that.
Thank You
Re: How to make one good image out of 2 broken ones
Ok, with a few more drives and some better reads I managed to get enough good image so I was able to fix it by copy pasting between the two images and it took something like 3 hours.
So here is the final image - maybe it would help for finishing the merging feature of the HXC:
https://www.dropbox.com/scl/fi/efo85zdl ... 8ilhs&dl=0
I have two last questions:
1) I'm not sure I chose the right or easiest way for the copy pasting repair - In track view I found the position of the end of sector cell on the image with good sectors in that position (positioned before the damaged sector of the other image). And as end point I selected the position of sector cell beginning minus one (again positioned after the damaged sectors of the other image). Then I clicked copy, went to data window on bottom, selected all, copied to clipboard. Then moved to other HXC instance with the other image and pasted on the appropriate place (it was not exactly the same place as on the other image, but offsetted according to the neighboring sector positions).
2) I have worked the same on the other floppy disk with program part2. And fixed all the places with data. But then there is a lot of filler (empty data) on that disk which also has a lot of damage. Do You think I must fix that also? Is there some better way how to fill those tracks with the repeating filler pattern much easier/faster?
Here is the other disk for Your consideration: https://www.dropbox.com/scl/fi/42gwmq87 ... 8rr6m&dl=0
As far as I know those programs are not specific to machine serial number, so this images might be a save for anyone having the same problem with Arburg Centex 420 C injection molding machines.
So here is the final image - maybe it would help for finishing the merging feature of the HXC:
https://www.dropbox.com/scl/fi/efo85zdl ... 8ilhs&dl=0
I have two last questions:
1) I'm not sure I chose the right or easiest way for the copy pasting repair - In track view I found the position of the end of sector cell on the image with good sectors in that position (positioned before the damaged sector of the other image). And as end point I selected the position of sector cell beginning minus one (again positioned after the damaged sectors of the other image). Then I clicked copy, went to data window on bottom, selected all, copied to clipboard. Then moved to other HXC instance with the other image and pasted on the appropriate place (it was not exactly the same place as on the other image, but offsetted according to the neighboring sector positions).
2) I have worked the same on the other floppy disk with program part2. And fixed all the places with data. But then there is a lot of filler (empty data) on that disk which also has a lot of damage. Do You think I must fix that also? Is there some better way how to fill those tracks with the repeating filler pattern much easier/faster?
Here is the other disk for Your consideration: https://www.dropbox.com/scl/fi/42gwmq87 ... 8rr6m&dl=0
As far as I know those programs are not specific to machine serial number, so this images might be a save for anyone having the same problem with Arburg Centex 420 C injection molding machines.