Posted by Rick Merrill on 07/13/06 19:10

PTravel wrote:

> "Richard Crowley" <richard.7.crowley@intel.com> wrote in message
> news:e95tan$4bh$1@news01.intel.com...
>
>>"PTravel" wrote ...
>>
>>>"Toby" wrote ...
>>>
>>>>>More sloppy typing by me -- that should read "gross drop out errors
>>>>>that CAN'T be fixed by conventional ECC . . . "
>>>>
>>>>You mean "can't be fixed LOSELESSLY by conventional ECC"? I would bet
>>>>that almost all ECC in DV is lossy.
>>>
>>>Yes, that's what I meant. As for your bet, I'm not interested in
>>>gambling. I'm more interested in hard data which, so far, no one has been
>>>able to provide.
>>
>>I suspect that if there *IS* any data of your specified hardness,
>>it is considered proprietary by the manufacturers. As long as we
>>consumers (both amateur and professional) are satisfied with the
>>performance, it would not be in their best interest to reveal the
>>actual numbers. This principle likely applies to breakfast cereal
>>and laundry detergent as much as it does to tape and video/audio
>>equipment.
>
>
> I would think something like "mitigated" error correction rate is something
> that would be a function of equipment manufacturers rather than tape
> providers. I know this kind of information is available for other kinds of
> hardware, e.g. hard drives.
>
>
>>Can one force a corporate entity to reveal proprietary informaion
>>in discovery? in testimony? Just curious.
>
>
> Yes. It's produced subject to a protective order, usually stipulated, that
> the information is either "confidential," i.e. not to be shared or exploited
> outside the litigation, or "confidential -- attorneys eyes only," i.e. not
> to be shown to the client, but available to the lawyers to develop and
> present their case.
>

Excuse me for reminiscing, but the early IBM drives had something like 9
read heads - the extra track was a parity track - and analog reads on
each track could trigger the parity correction if one track was too much
worse/lower a signal level than all the others. However, if that track
were consistently bad it would literally be ignore for the whole tape -
the consequence was that any drop outs on other tracks would be ignored
completely!

[Back to original message]