Helmut Jarausch wrote:
I am mostly concerned with the ability to read all sectors in reasonable time, less so with the plotting. I have somewhere here a program I wrote decades ago to smooth data, and it allows graphing all forms of average, and of high and low values to put an envelope around the average. I do find the evaluation useful, but on ISO media it also evaluates the runout (-pad) sectors which I don't care about. So many disks which I think are fine are rated as trash.On 26 Dec, Bill Davidsen wrote:I found this link: http://swaj.net/unix/cdck/cdck-0.6.0.tar.gz which is a program to read a CD or DVD and report on the goodness of the burn, in terms of current and potential errors. I can't swear that the results are accurate in terms of predicting future failures, but they are consistent, in that a DVD identified as marginal, good, or excellent on one machine will have the same rating on at least two other machines using other brands of DVD reader hardware.I have been checking my critical backups and reburning any rated less than good. Only rarely do I see the excellent rating, which may reflect my use of good but not premium media.I've tried it with the '-p' option on a DVD. It produces a file of 15 Mb containing 1.5 million lines. That is too much for my gnuplot (probably it needs more than 1.5 Gb memory, or some other limitations) Has anybody had more luck?
I am looking at the source, to see what it would take to add an option to stop testing at the last meaningful sector. The obvious solution appears to fail with multisession media, so I have to rethink it. Joerg was kind enough to point out the meshpoints feature of the current readcd, which may be useful, although it reads in speed. Unfortunately DVD is recorded differently than CD, so c2scan doesn't work to find the sectors which are "really right" on the media.
I did some work with fire code in the 80's, it's not clear just how bad a media that would let you recover.
And I am working my way through some old tapes I made around 1990, so far not an error in the bunch, 60 tapes, 3-6 files per tape. They are small, but robust, I guess.
-- bill davidsen <davidsen@tmr.com> CTO TMR Associates, Inc Doing interesting things with small computers since 1979