Re: DVD+, buffer underrun, write speed and verification issues
Thanks for the tips Volker, Greg and Bill. ;)
Volker Kuhlmann wrote:
> DVDs are much more precise in finding the spot again from which to
> continue burning, especially DVD+ which I understand is similar to
> DVD-RAM. DVD-RAM effectively goes on nothing but underruns. You
> needn't be concerned about engaging underruns with DVDs.
You are right, but I just want to elaborate some more on this as I
found some more sources of information on the subject.
My question directly plays out in the conversation here:
http://club.cdfreaks.com/lite/t-91137.html
The point I want to emphasize is the second post which was
by "Wesociety":
: AFAIK losslinking and buffer underrun protection are 2 separate
: features and are both used when burning a DVD+RW.
To which "Henjie" replied:
:: I'm kinda under the impression that buffer underrun protection is
:: an inherent part of the DVD+R/RW format though.
The key point is that the lossless linking feature of DVD+ allows data
to be recorded at the packet level without having any data degradation
as a result of buffer underruns. According to Andy's docs, the DVD+
"splice point" lies within sync patterns, not user data. In effect,
from a practical standpoint, buffer underruns are of no consequence
with DVD+ formats because the DVD+ system does not require the buffer
to be continuously filled during the write process. This is why DVD+RW
can be thought of, and used much like (with the few caveats Andy
mentioned in his docs) a slow hard drive - no special writing
software is even required. It is the same with DVD+R, except that
each sector/packet can only be written once. IMHO, DVD+ is obviously
superior to DVD- for this reason.
A good source for technical papers on the subject can be found
here: http://www.dvdplusrw.org/Article.asp?mid=0&sid=4&aid=6
However, it is important to point out that this is NOT true
with DVD- formats. Interrupting the data stream during a write
will have consequences as it will result in several K bytes of linking
data added and that several bytes of user data will be corrupted
(and this will be true even on the best "burnfree" DVD- drives,
without burnfree this event = coaster). The ECC system is capable of
correcting these minor errors and so the data should appear error free
when read back. However, the errors do reduce what I call "ECC margin"
in that any existing (pre-corrected) errors on the media do reduce the
capacity of the ECC system to correct for *future* errors (due to dust,
scratches, etc.) Thus, a DVD- media that was subjected to underruns
will be less (how much so I don't know) reliable than if the buffer
had been kept full during the recording.
Being able to view the DVD+ format confidently as a slow hard drive
is probably why the DVD+ developers did not feel the need to provide
for simulation/dummy writes - they viewed the simulation mode only
as means by which users can detect potential buffer underrun conditions.
However, they really should have realized that we users often use the
simulation mode for more than just testing for underruns. Heck, its
those unexpected pesky SCSI errors we sometimes are dealing with. ;)
If it's write-once, it needs a simulation mode, IMHO.
Which brings me back to another issue. Cdrecord-ProDVD lists possible
buffer underruns after a burn. Perhaps this message should be qualified
(as relatively unimportant) for DVD+ media.
On the other hand, growisofs never reports possible underrun conditions
- at least according to my cursory examination of its code. Perhaps it
should do so - especially if it is being used with DVD- media - even if
for no other reason than that we can use this information to evaluate
the tuning of our systems.
Cheers,
Mike Shell
Reply to: