[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: How to "blank" / "zero out" / "nullify" a dvd-rw?


j t:
> just found http://fy.chalmers.se/~appro/linux/DVD+RW/-RW/ which
> explains the same subject, I think)

Yep. This puts some emphasis on readability by
entertainment hardware. Video players, game consoles.

Cool, i did not know the paragraph about 
DVD-R Dual Layer yet. Gives me something to
search for in the MMC specs.

We got a zoo of media and many of them can be
written in more than one way. DVD-RW has three
ways to write (DAO, Incremental, Overwrite) and
four media states (fully blank, blanked fast,
appendable, overwriteable) which allow one or
two of those ways.

Joerg Schilling:
> You should try to use cdrecord blank=all or cdrecord blank=all on the medium
> in order to bring it back into usability if it was used by growisofs before.
> Later, when using cdrecord for writing, it is sufficient to use 
> cdrecord blank=fast to make it usable again.
> ...
> A full blank is only needed if the media was written in packet mode
> before or if it was formatted.

Are you sure that blank=all is really needed for
deformatting ? Since cdrecord writes in DAO mode
it should suffice to blank fast. Even if the
media was formatted before.

Contemporary growisofs does not format DVD-RW
by default.
It leaves sequential DVD-R[W] appendable by
default, which makes them unusable for DAO.
But a simple blank=fast should suffice.
If option -dvd-compat is given, then the media
get closed and thus unusable for any writing
until they get blanked. If -dvd-compat is given
on blank media then DAO is used.
  cdrecord -sao
does about the same as
  growisofs -dvd-compat -Z

The erase-my-bytes job with cdrecord would look like
  cdrecord -v dev=ATA:1,0,0 blank=fast
  dd if=/dev/zero bs=2048 count=$size | \
  cdrecord -v dev=ATA:1,0,0 -sao tsize="$size"s -
Possibly it is wise not to use the full announced
capacity but a few sectors less. Like:

I just tested with an overwriteable DVD-RW and
cdrecord-2.01.01a23 . Only failure is that i cannot
combine the blank run with the write run:
  cdrecord -v dev=0,0,0 blank=fast -sao tsize="$size"s -
  WARNING: Phys disk size 2146400 differs from rzone size 0! Prerecorded disk?
  WARNING: Phys start: 196608 Phys end 2343007
  cdrecord: Data will not fit on any disk.
  cdrecord: Cannot write CD's >= 100 minutes.

If i do blanking and writing in two runs then it works:
  Blanking time:   46.793s
  Writing  time:  842.177s

Same result with an appendable sequential DVD-RW.

You should make some new tests. 

> Well, cdrecord will soon have it's 10th DVD writing anniversary ;-)

Congrats :)
Since one year we are ageing at the same speed.

> dev=ATA:1,0,0 will not work.

ATA:1,1,0 as /dev/hdd  works fine for me on a SuSE 9.3
with four scannable drives:
  /dev/s[rg][01] , /dev/hd[cd]
of which two are burners.

> > unformatted state. -blank without "=full" does the same as
> > cdrecord blank=fast. 
Joerg Schilling:
> It is most unlikely that dvd+rw-format -blank=full does the same
> as cdrecord blank=fast. This would be counter-intuitive.

I wrote:  _without_ "=full". 
dvd+rw-format -blank=full is like cdrecord blank=all
dvd+rw-format -blank      is like cdrecord blank=fast

> DVD-RWs last a very long time if you use the right software,
> e.g. cdrecord.

Not with my DVD drives. 4x DVD-RW die with all programs
and if they are dead then nothing can revive them.
The software is not decisive. I went nuts when i developed
libburn's sequential DVD-RW support and lost two media.
So i made expensive tests and lost one media each with
growisofs and with cdrecord by exactly the same sympthoms.
That soothed me.

Have a nice day :)


Reply to: