Amazed: DVD polish worked!

[The Natural Philosopher]

...trying to play some DVDs of my sisters, and one wouldn't play.

Half an hour with Brasso* and a soft rag absolutely removed some appalling scratches, and it played end to end.

*All she had...

[Andy Burns]

Yeah, have done it with scratched CDs, polish getting everywhere did manage to remove the silk-screen printing on the other side too ...

[Dan S. MacAbre]

I use T-Cut, which is suppose is similar.

[The Natural Philosopher]

Good info.

I would have, if there had been any handy.

[Dan S. MacAbre]

I've also found that toothpaste works with very light scratches, and sometimes rubbing on a bit of 3-in-1, then wiping most of it off again (otherwise it gets sprayed around your dvd drive). Maybe something to do with refractive indexes (indices?). I used to get a lot of scratched disks from the Amazon Lovefilm rental library :-) They're much better now.

[Malcolm Race]

Surprised it needed that. When CDs were introduced I saw a TV program where the vsawed 4 radial cuts into the CD at 90 dgree intervals. The CD played almost as good as the original.. The purpose was to show tge redundacy of the data recorded.

Malcolm

[Jethro_uk]

The cynic in me suggests that's the only damage the redundancy would cope with, having had CDs become unplayable with almost imperceptible scuff marks.

OP: The first thing to do once it played was to rip it to file ...

[Chris Bartram]

Radial scratches are tolerated better, the more they go around the disc, the worse.

In order of preference for polishing, I've found Mer car polish the best, it's a bit gentler.

[jkn]

Those sort of demos were later admitted to be fakes, for the purposes of marketing. A bit like the original BSB squarial, which was waved around as though it was a product, when in fact it had not been developed yet.

(not that there isn't redundancy in CD information etc, but not, ahem, 'perfect sound forever')

Jon N

[Mike Tomlinson]

En el artículo , Andy Burns escribió:

... and the upper side is actually the important one. Scratch that side and it's sayonara.

[Andy Burns]

Radial cuts and scratches will tend to cause errors in non adjacent sectors (which is why you're encouraged to clean them radially, rather than circularly)

[alan_m]

I think it was an open university program broadcast on BBC2, probably about 20 years ago.

[The Natural Philosopher]

the BIG scratch was an arc..would have wiped out a huge section of a track

[The Natural Philosopher]

That is what I WAS doing...

[Mike Clarke]

Isn't that just for re-writable ones?

[Andrew Gabriel]

It's a lot on an audio CD - for each 2048 byte sector, there's an additional

176 bytes of redundant error correction code written, which is designed to be able to recover from a large chunk of a sector being wiped out. It uses a pair of interleaved Reed Soloman error corrections, interleaved to better cope with a contiguous chunk of a sector being destroyed (as would be done by a scratch). It can reconstruct up to 500 bytes of lost data (a scratch about 2.5mm wide measured as an arc along the track, or more strictly totalling no more than 2.5mm of scratches across all the sectors covered by that correction code).

If a scratch is radial, it's unlikely to be 2.5mm wide, so that will be completely recovered. However, as the angle of a scratch changes to be more alighed with the track, the effective arc length increases and will exceed 2.5mm of track. The other problem is the head maintaining tracking across such a scratch, and we've all heard CD's jumping when this fails, even emulating an LP record stuck in the grove.

[Mike Tomlinson]

En el artículo , Mike Clarke escribió:

No, the reflective surface with the data pits and lands (representing binary 0 and 1) is on the top. It is only protected by a thin layer of lacquer and the label or screenprint. When you put the disc in a player, it is facing down and the laser shines upwards.

It's the top surface that is more delicate and prone to damage. The bottom is actually far more robust, which ic why you can polish out scratches with stuff like toothpaste and get away with it.

If you scratch the top (labelled) surface, hold the disc up to a light and can see light through the scratch, the data in that spot is now unreadable as you've removed the reflective layer that the laser uses to detect pits and lands.

This pic illustrates it:

[Brian Gaff]

That is right, though not quite so bad on dvds that are multi layer generally. the top is the silvered side on most moderncds. What you do not want to do is get air into the area which is silvered as they tend to oxidize and go a funny brown colour at which point CDs seldome play reliably any more. I remember in the early days they use to produce a tst cd with a small hole through the disc at certain places, and these generally played OK when new but once the rot set in around the hole no more play.

I always remember the old spread jam on it wash it under warm water and it still plays. What they failed to say was that finger prints and scratches make the discs unplayyable apparently due to scatter effects. Brian

[John Rumm]

More the virtue of the forward error correction (Reed Solomon encoding)... Very good at fixing burst errors, but not so capable with long runs of corrupted bits. Probably would have been ideal on a caddy based CD system ;-)

..and another 10+ years on that ;-) Tomorrow's world showed the demo as well ISTR.

[pamela]

For even lighter scratches a squirt of Pledge furniture polish can help. The silicone seem to fill in some of those tiny scratches.