You might find it slighly quicker to do the following:

1. Have Nero create the 5 ISO files from your data files.
2. Create the PAR2 files from the ISOs.
3. Burn the ISOs verifying after each burn.
4. Burn the PAR2s (again verifying).
[/quote:3ggz0g2f]

Yes, that will be faster. I didn't know step 1 was possible.

What you do is take the ISO you created from the PAR-DVD and rename the the .ISO file so that it has a .PAR2 file extension instead. You can then open this .PAR2 file with QuickPar. QuickPar will find all of the recovery data and you can use it to repair the 5 data DVDs. Once the data DVDs have been repaired, you can create new undamaged PAR2 files and burn them to a new PAR-DVD.
[/quote:3ggz0g2f]

ok.

This I don't know. I've never attempted it.

If you can only RIP the whole disk, then what you would then need to do make a duplicate copy of the ISO and rename one copy as a PAR2. [/quote:3ggz0g2f]

But we still would need to rip a separate ISO for the data. Anyway, it seems that using a 2nd session for the parity info is not a good idea. Better to store in a separate DVD.

Also, instead of having 11 sessions for the 11 parity "packs", it would be better to store all of them in a single session, in separate directories, isn't it? If a data DVD gets damaged, we simply rip the whole PAR-DVD, rename the resulting .ISO to .PAR, and QuickPAR will only take the parity corresponding to the damaged DVD, ignoring the parity corresponding to the other 10 discs. Am I right?

The time to create PAR2 files is given by the following formula:

Time = (source_size^2 * redundancy) / (block_size * speed)

So, for 4GB of data, a redunancy level of 10%, a block size of 2MB, and a speed of 250MB/s, this gives:

Time = (4096^2 * 0.1) / (2 * 250) = 3355 seconds = 56 minutes.[/quote:3ggz0g2f]

ok, so twice the size of the source, 4 times the time to calculate the parity.
Then it is clear that it is much faster to create parity files separately for each DVD, instead of doing it in sets of 5 DVDs.
If we increase source size 5 times, the time increases not 5 times, but 25 times. Then it is 5 times faster doing it "disc a disc".

and a speed of 250MB/s,[/quote:3ggz0g2f]
what is this speed? the proccessing speed or the disk transfer speed?
is this figure the typical for a desktop PC?

Offline

 

#27 2004-09-11 11:41:53

PeterBClements

Site Admin

From: United Kingdom

Registered: 2004-01-05

Posts: 1529

Website

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

"argab":2ofmlvck wrote:

ok, so twice the size of the source, 4 times the time to calculate the parity.
Then it is clear that it is much faster to create parity files separately for each DVD, instead of doing it in sets of 5 DVDs.
If we increase source size 5 times, the time increases not 5 times, but 25 times. Then it is 5 times faster doing it "disc a disc".

and a speed of 250MB/s,[/quote:2ofmlvck]
what is this speed? the proccessing speed or the disk transfer speed?
is this figure the typical for a desktop PC?[/quote:2ofmlvck]
This is the speed which QuickPar reports on screen during processing. The value will be different for every computer and depends on the speed of the CPU and RAM.

Users on old Celeron based PCs may get speeds as low as 20MB/s, and those with the most up to date CPUs may get speeds of 800MB/s. I get 250MB/s on my PC (with Athlon 1900+ CPU).

Offline

 

#28 2004-09-11 14:25:37

argab

Active

Registered: 2004-09-09

Posts: 28

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

This is the speed which QuickPar reports on screen during processing. The value will be different for every computer and depends on the speed of the CPU and RAM.

Users on old Celeron based PCs may get speeds as low as 20MB/s, and those with the most up to date CPUs may get speeds of 800MB/s. I get 250MB/s on my PC (with Athlon 1900+ CPU).[/quote:tq5ovmmh]

ok, my fastest computer is a Barton 2500+, so I will do it in about 45 minutes. It is a reasonable amount of time.

Offline

 

#29 2004-09-11 18:10:52

argab

Active

Registered: 2004-09-09

Posts: 28

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

About creating the first ISO with Nero, and later with ISOBUSTER, there is a risk, pointed out by tynopik @StorageReview, that you get different ISOs:

if you run isobuster multiple times you will undoubtedly get the same iso each time (unless you have serious corruption issues). My concern was that the iso that it generates will not be the same as the one that nero generates in the first place

i compare it to winzip and pkzip, if you use both to compress the same set of files, they will not generate a bit-for-bit identical zip files. Both adhere to the zip standard, but the standard allows some flexibility. Same thing with different mp3 encoders. Encode the same song with different encoders and both generate a valid mp3, but it isn't exactly the same

the only 100% safe way would be to burn the disc as regular, then use isobuster to read the burned disc and generate an iso, and then use THAT iso as the basis for all parity calculations[/quote:hpfkg2gm]

do you know anything about this risk?

edit:
also, using different versions of ISOBuster, could result in different ISO files?


and oh, I forgot about one important thing:
we are talking about ISO images.... but I burn UDF DVDs.
any problem making ISO images out of UDF DVDs?

Offline

 

#30 2004-09-12 10:03:12

PeterBClements

Site Admin

From: United Kingdom

Registered: 2004-01-05

Posts: 1529

Website

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

"argab":1smy33jq wrote:

About creating the first ISO with Nero, and later with ISOBUSTER, there is a risk, pointed out by tynopik @StorageReview, that you get different ISOs:

if you run isobuster multiple times you will undoubtedly get the same iso each time (unless you have serious corruption issues). My concern was that the iso that it generates will not be the same as the one that nero generates in the first place

the only 100% safe way would be to burn the disc as regular, then use isobuster to read the burned disc and generate an iso, and then use THAT iso as the basis for all parity calculations[/quote:1smy33jq]
do you know anything about this risk?[/quote:1smy33jq]
OK, I would expect that if you use two different programs to create an ISO file from a set of files on your HD, that the ISO files would not be identical.

I was not aware that it was possible for two different ripping programs to create ISO files that were not identical.

edit:
also, using different versions of ISOBuster, could result in different ISO files?

and oh, I forgot about one important thing:
we are talking about ISO images.... but I burn UDF DVDs.
any problem making ISO images out of UDF DVDs?[/quote:1smy33jq]
No idea.

Offline

 

#31 2004-09-12 17:25:19

tgh

Veteran

Registered: 2004-02-03

Posts: 129

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

You can make ISOs of UDF DVDs.

I'm not sure that ISOs will differ if made by different programs (e.g. generating the ISO in Nero/EasyCD, then burning, then ripping to a new ISO using ISOBuster).  I *thought* that ISO files are simply a physical representation of the sectors on the media, in which case "an ISO is an ISO".  Haven't tested that assumption yet.

Offline

 

#32 2004-09-13 01:32:43

argab

Active

Registered: 2004-09-09

Posts: 28

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

You can make ISOs of UDF DVDs.
[/quote:1e31qcxp]

great, thanks

I'm not sure that ISOs will differ if made by different programs (e.g. generating the ISO in Nero/EasyCD, then burning, then ripping to a new ISO using ISOBuster). I *thought* that ISO files are simply a physical representation of the sectors on the media, in which case "an ISO is an ISO". Haven't tested that assumption yet.[/quote:1e31qcxp]

mmh, yes, I also thought that ISO was simply a sequential read of the data on the DVD, so no possible variations. But now I read that ISOs have certain overhead... so they have probably some kind of encapsulation, not only the raw data on the DVD...

well, to be sure, I guess it will be better to extract the ISO, always with the same utility. And it will be interesting to do tests with different utilities to see if the ISOs are different.
Any utility to compare two ISO files?

Offline

 

#33 2004-09-13 05:29:20

argab

Active

Registered: 2004-09-09

Posts: 28

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

Is there any chance that the ISO image couldn't be read from a damaged DVD?
Some people say that it is better to apply PAR2 to actual files in the DVD, instead of applying it to the ISO image. They say that in case of damage to the disc, it would be easier to read (with ISOBUSTER) those actual files instead of the whole ISO. Is that true?

In that line, it is suggested to create a big ZIP (or better, a Virtual Drive), and burn it to the DVD. The PAR2 parity would be created on the big ZIP.

(The idea has been suggested at the SR forum, by tynopik.)

Offline

 

#34 2004-09-13 11:35:46

PeterBClements

Site Admin

From: United Kingdom

Registered: 2004-01-05

Posts: 1529

Website

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

If the PAR2 files were created from the files, then it makes no difference to QuickPar whether you get it to scan a ripped ISO of the whole disk, or ripped copies of the individual files.

Obviously if the PAR2 files were created from a ripped ISO of the whole disk, then that is another matter.

I'm not familiar with the "SR forum".

Offline

 

#35 2004-09-13 16:00:44

argab

Active

Registered: 2004-09-09

Posts: 28

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

If the PAR2 files were created from the files, then it makes no difference to QuickPar whether you get it to scan a ripped ISO of the whole disk, or ripped copies of the individual files.

Obviously if the PAR2 files were created from a ripped ISO of the whole disk, then that is another matter. [/quote:1dk5qs90]

what they mean is that it is safer to rip files instead of ripping an ISO, in a damaged disc.

I am going to copy here the pertinent excerpts from the thread at Storage review Forum:

<!-- m --><a class="postlink" href="http://forums.storagereview.net/index.php?showtopic=16925&st=0&#entry180557">http://forums.storagereview.net/index.p ... ntry180557</a><!-- m -->

TYNOPIC WROTE:

1. use a 'virtual drive' program that lets you mount a file as a drive. As bonus, many such programs (like the one from jetico) support encryption. This will solve lots of problems with long pathnames and funky filenames as you're actually only burning one file to the dvd. And since you're dealing with a real hard file, there are no issues with trying to regenerate an iso and file recovery utilities would work better (you can use something like badcopypro to copy of a part of the file, but i haven't seen anything that lets you generate part of an iso)
[/quote:1dk5qs90]

ARGA WROTE:

I am a beginner here too. I didn't know ISO recovery was difficult. In fact I thought that it was very easy. Do you mean that, if a DVD has errors (and this is the case we are dealing with), I will have problems to extract the ISO? And maybe that extracted ISO won't be recognized by Quickpar to do the repair task?
[/quote:1dk5qs90]

TYNOPIC WROTE:

it's not something i have tried nor would i want to try. who knows, it might work fine, but i feel more comfortable working with an actual file[/quote:1dk5qs90]

-------

Another poster, "tolyngee" has another approach:
he creates a big collection of RARs, 20MB each, and stores them in the DVD. He creates PAR files out of this collection of RARs. Advantage: in case of damage to one or several RARs, he only has to RIP and repair those damaged RARs, and not the entire disc.

Offline

 

#36 2004-09-13 16:55:24

PeterBClements

Site Admin

From: United Kingdom

Registered: 2004-01-05

Posts: 1529

Website

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

"argab":5yfr0vnx wrote:

what they mean is that it is safer to rip files instead of ripping an ISO, in a damaged disc.[/quote:5yfr0vnx]
I suppose it depends on how bad the disk is. I know trying to rip a bad disk takes a long time, but other that the time taken I would not expect there to be any other problems associated with the attempt.

I am going to copy here the pertinent excerpts from the thread at Storage review Forum:

<!-- m --><a class="postlink" href="http://forums.storagereview.net/index.php?showtopic=16925&st=0&#entry180557">http://forums.storagereview.net/index.p ... ntry180557</a><!-- m -->[/quote:5yfr0vnx]
Thanks, I'll have a look.

"TYNOPIC ":5yfr0vnx wrote:

1. use a 'virtual drive' program that lets you mount a file as a drive. As bonus, many such programs (like the one from jetico) support encryption. This will solve lots of problems with long pathnames and funky filenames as you're actually only burning one file to the dvd. And since you're dealing with a real hard file, there are no issues with trying to regenerate an iso and file recovery utilities would work better (you can use something like badcopypro to copy of a part of the file, but i haven't seen anything that lets you generate part of an iso)
[/quote:5yfr0vnx]
Virtual drives can be very usefull. I use them myself. And writing the virtual drive file to DVD for backup is a great idea.

"ARGA":5yfr0vnx wrote:

I am a beginner here too. I didn't know ISO recovery was difficult. In fact I thought that it was very easy. Do you mean that, if a DVD has errors (and this is the case we are dealing with), I will have problems to extract the ISO? And maybe that extracted ISO won't be recognized by Quickpar to do the repair task?
[/quote:5yfr0vnx]
I would have assumed that you could always extract an ISO from a DVD, and that the parts of the ISO that correspond with unreadable parts of the disk would just be filled with NULLs. QuickPar would be perfectly happy with this.

Another poster, "tolyngee" has another approach:
he creates a big collection of RARs, 20MB each, and stores them in the DVD. He creates PAR files out of this collection of RARs. Advantage: in case of damage to one or several RARs, he only has to RIP and repair those damaged RARs, and not the entire disc.[/quote:5yfr0vnx]
Actually, you would still need to copy or rip all of the files from the DVD to HD, as QuickPar will get confused when it tries to do the repair if you try to get it to use the files that are still on the DVD.

Offline

 

#37 2004-09-13 18:48:42

argab

Active

Registered: 2004-09-09

Posts: 28

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

I suppose it depends on how bad the disk is. I know trying to rip a bad disk takes a long time, but other that the time taken I would not expect there to be any other problems associated with the attempt.
...
I would have assumed that you could always extract an ISO from a DVD, and that the parts of the ISO that correspond with unreadable parts of the disk would just be filled with NULLs. QuickPar would be perfectly happy with this.
[/quote:7y7ghbwq]

alright. seems fair enough.
This was my main concern right now.

Virtual drives can be very usefull. I use them myself. And writing the virtual drive file to DVD for backup is a great idea.
[/quote:7y7ghbwq]

I will try this too then. But first I will use ISOs, as the discs can be read without any special Virtual Drive software.

Actually, you would still need to copy or rip all of the files from the DVD to HD, as QuickPar will get confused when it tries to do the repair if you try to get it to use the files that are still on the DVD.[/quote:7y7ghbwq]

I see...
besides, this RAR method has the drawback of having to depack everything before watching.
I will have this method in mind, but I will try the others first.

Offline

 

#38 2004-09-14 00:22:33

PeterBClements

Site Admin

From: United Kingdom

Registered: 2004-01-05

Posts: 1529

Website

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

"argab":fx0t64ul wrote:

Actually, you would still need to copy or rip all of the files from the DVD to HD, as QuickPar will get confused when it tries to do the repair if you try to get it to use the files that are still on the DVD.[/quote:fx0t64ul]

I see...[/quote:fx0t64ul]
Yeah. QuickPar does not currently understand "read only" volumes.

If you open a PAR2 file directly from the CD/DVD, and attempt a repair, QuickPar will attempt to write the repaired files to the CD/DVD.

If you open a PAR2 file from the HD, and then "Add" files from the CD/DVD, then QuickPar will attempt to move the files from the CD/DVD to the HD by renaming them.

Both of these cases will fail.

Offline

 

#39 2004-09-14 03:18:17

argab

Active

Registered: 2004-09-09

Posts: 28

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

@PeterBClements

Yeah. QuickPar does not currently understand "read only" volumes.

If you open a PAR2 file directly from the CD/DVD, and attempt a repair, QuickPar will attempt to write the repaired files to the CD/DVD.

If you open a PAR2 file from the HD, and then "Add" files from the CD/DVD, then QuickPar will attempt to move the files from the CD/DVD to the HD by renaming them.

Both of these cases will fail.[/quote:2yt1udrp]

ok, no problem. I will do everything in the hard disk. In fact, it looks the most natural way of doing it.

Lets see if tomorrow I can start my first tests. First I have to make a little room in my hard disks for the ISO rips    <!-- s --><img src="{SMILIES_PATH}/icon_wink.gif" alt="" title="Wink" /><!-- s --> 

Ah, by the way, probably you know this, but anyway, I tell you: reading speed for recordable media is normally limited to 8X on DVD readers.
But some readers (e.g. LiteOn) have  patched firmwares available that remove these limits. With these firmwares, you can read at full speed (16X) if the disc is of good quality. If the disc is of bad quality, you can still try to read at 16X, but the reader will find errors, and it will need to slow down, and re-read the bad parts.

This website publishes patched firmwares:
<!-- m --><a class="postlink" href="http://codeguys.rpc1.org/firmwares.html">http://codeguys.rpc1.org/firmwares.html</a><!-- m -->

in fact, I am using them in my LiteOn 163D, and LiteOn 167T.

Offline

 

#40 2004-10-26 18:10:15

brett

Member

From: Columbus OH

Registered: 2004-03-10

Posts: 4

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

"argab":1khplpsu wrote:

what they mean is that it is safer to rip files instead of ripping an ISO, in a damaged disc.

I am going to copy here the pertinent excerpts from the thread at Storage review Forum:

<!-- m --><a class="postlink" href="http://forums.storagereview.net/index.php?showtopic=16925&st=0&#entry180557">http://forums.storagereview.net/index.p ... ntry180557</a><!-- m -->
[/quote:1khplpsu]

I love StorageReview, but those guys you're quoting ... well ... they're utterly clueless.

If you wanted to make a million tiny 20MB RAR files, why not just use RAR's built in Recovery Volume functionality?  It's basically the same as PAR files, and WinRAR recovers them automatically.  The recovery volumes only work on a 1:1 basis, though -- each RAR is marked as either damaged or not, and if your total number of damaged RARs exceeds your number of recovery volumes YOU LOSE.

Anyway, about ISO's:

ISOs aren't supposed to have any overhead.  An ISO is the raw image of what's on a CD/DVD.  The only confusion comes with whether or not it is a raw 2352-byte/sector image or a cooked 2048-byte/sector image.  This is all irrelevant for DVDs, because DVDs only have 2048 bytes per sector, so every DVD ISO is going to be a 2048-byte/sector raw image.

So, as far as a DVD-R is concerned, any program you use to extract an ISO should give you the exact same file.  If it's not the exact same file and/or it has some sort of header on it, the program you're using is doing it wrong.

It is absolutely more reliable to make your PAR2 data from the ISOs, not from the actual files.  If you're trying to read the filesystem of a DVD, the sectors with the filesystem itself are extremely important.  If those sectors are damaged, you can't even find the files to try and repair them.  If you make PAR2 data of the ISO, though, then no sector is any more important than any other one.  I have done tests actually drilling holes in discs and using sandpaper on parts of them, and having PAR2 data of the ISO is by far the best way to recover them.

IsoBuster has made this less of an issue, because it does an amazing job of finding files even when the directory structure is damaged.  Still, it's more work to recover the files than to just rip the ISO and fix it with the PAR2.

Seeing as a 200GB hard drive is $100, it's pretty easy to deal with a ton of 4.3GB disc images.

Say you have 8 DVDs of stuff you want to archive and you're going to have 2 DVDs of recovery data.  Absolutely the most reliable way to do it is to make two DVDs containing a single set of PAR2 data for all 8 ISOs.  It's a bit inconvenient, and it takes 45 GB of hard drive space, but that way you can totally lose one DVD and beat the hell out of all the other ones and you'll still probably recover it all.  If you think that's too much of a pain, though, you'll still probably be just fine with 1GB of individual PAR2 data for each DVD, then just burn the two recovery DVDs with 8 different PAR2 sets.

Being able to recover a totally destroyed DVD is very nice when it comes to totally irreplacable data.  For most things, though, it may be overkill.  I backed up 6 DVDs of MP3s, and I made a single CD-R with 116 MB of PAR2 data (116 * 1MB blocks) for each DVD ISO.  I don't want to lose them, but it's not like I can't re-rip or re-download them.

Offline

 

#41 2004-10-27 09:54:36

brett

Member

From: Columbus OH

Registered: 2004-03-10

Posts: 4

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

I guess that was pretty harsh for me to say they're utterly clueless... but they're going about things completely wrong.  If you have one huge RAR file on a DVD-R, then there are bad sectors in that file, you can have a lot of trouble copying the individual file.  Having many small files is a partial solution if you have enough PAR2 data to recover several files, but if the directory structure becomes damaged, things get very difficult.

They're talking about using IsoBuster to extract the individual files instead of using it to extract the whole ISO.  That's what causes problems.  If you were to rip the whole DVD to an ISO and recover the ISO with PAR2 data made from the ISO image, everything would work fine.

If it doesn't matter that much, you probably might as well just make RAR Recovery Volumes, since that whole process is very simple.

Offline

 

#42 2004-11-15 23:51:19

tynopik

Member

Registered: 2004-09-24

Posts: 8

Re: Isit possible to create a parity DVD-R from a set of DVD-Rs?

"brett":3lbjzew1 wrote:

ISOs aren't supposed to have any overhead.  An ISO is the raw image of what's on a CD/DVD.  The only confusion comes with whether or not it is a raw 2352-byte/sector image or a cooked 2048-byte/sector image.  This is all irrelevant for DVDs, because DVDs only have 2048 bytes per sector, so every DVD ISO is going to be a 2048-byte/sector raw image.

So, as far as a DVD-R is concerned, any program you use to extract an ISO should give you the exact same file.  If it's not the exact same file and/or it has some sort of header on it, the program you're using is doing it wrong.[/quote:3lbjzew1]

ok, thanks for answering that question

i didn't know if variation was possible within an iso, but it was a concern to me

since it apparently is not, then by all means, basing parity data off the iso image would be the way to go

Offline

 

 

Pages: 1 2

• Index
•  » Data Backup
•  » Isit possible to create a parity DVD-R from a set of DVD-Rs?

Board footer

Jump to
Announcements UseNet Data Backup Technical support Wishlist Technical support Wishlist

Powered by PunBB
© Copyright 2002–2008 PunBB

QuickPar © 2002–2008 Peter B. Clements