Sorry, but I just don't buy it. A certain amount of bent bits in the
middle of an image file could very likely go unnoticed. Enough bent
bits to be noticeable in the image data itself would be, well...
noticeable. But if there are enough bent bits in the image data to be
noticeable there are likely an equivalent proportion of bent bits in the
non image data as well.
What that means is that visible corruption on the card's image data will
eventually show as bad bits in the EXIF and the FAT as well. It might
take a while to notice the EXIF errors but the FAT data errors won't
take long to become noticeable in a very unfriendly way.
If you have funny looking images you need to look elsewhere for the
cause. I'll bet you can read the images off that card into your hard
drive, write them back and do a bit for bit compare and do that until
the cows come home and you won't find a single bit error on any of the
cards you use.
Chuck Norcutt
AG Schnozz wrote:
>> You jumped the gun on the date, AG! This was intended to be posted
>> on April 1, no?
>
> It was April 1st in half the world when I wrote it. I didn't want our
> New Zealand friends to be disappointed.
>
>> If you're serious, I am reserving disbelief.
>
> The basis of my report is in fact. In the particular industry that
> pays my insurance, I'm involved in R&D. One thing we encountered was
> equipment that uses CF cards. This isn't any old equipment, but that
> stuff of which we absolutely depend on and costs millions of dollars.
> Well, the manufacturer specifically recommended a particular
> brand/model of CF card because of internal shielding, and long-term
> reliability. Shielding? Yes. It turns out that some cards are more
> susceptable to radiation and RF interference than others.
>
> One thing I am required to do in my job is to "never assume
> anything". If a mission-critical piece of equipment were to fail at
> an inopportune time, people can die. Thanks to laws inacted since
> 9-11, we are now held to a much higher standard than ever before and
> must have contingency plans for every possibility.
>
> This was a question I was asked by high-level management on how
> reliable these cards are. This is for equipment which will most
> likely be in place for up to 25 years.
>
> It turns out that there ARE differences in the design and manufacture
> of the various CF cards. It turns out that good-old magnetic media
> as found in the Microdrive is the most reliable in a high-RF
> environment. Also, the driver chipsets in cameras and these
> dedicated devices ARE optimized to certain performance specs.
>
> We assume that because it's "digital" that it's always going to be
> 100% correct. How do we really know that? Experience? Just because
> we've never seen it YET, doesn't mean that the problem doesn't exist.
>
> I was very specific in my "report". I identified three different
> cards and identified just ONE camera as a potential source of
> incompatibility. Do incompatibilities exist in other digital
> components? Of course they do. Why wouldn't cameras be any
> different? As an example, RAW files from E-system cameras will tend
> to have artifacts in orange/yellow highlights when processed with
> some raw converters, but not others.
>
> Back in a previous employment, my company used high-performance
> hard-drives. We had, by far, the most demanding requirements of any
> application on hard-drives. We always had to stress-test every new
> make and model and even manufacturing run of the drives because of
> incremental firmware changes which would alter timing, response or
> other issues. Thermal-Recalc anyone? If the hard-drive failed to
> communicate a "not-ready" state to the device on the other end of the
> ribbon-cable, we would actually get data dumping out onto the floor.
> (so to speak). It is highly plausable that these kinds of errors or
> design-faults exist in our cameras and storage media. Usually the CRC
> wouldn't catch it because the failure occured ahead of that point.
>
> I am reminded that when the EOS-1Ds Mark II came out that it had
> imcompatibilities with several types of storage cards. Sometimes it
> would result in lockups, othertimes it resulted in corrupted data.
> These types of failures are OBVIOUS. But what about the subtle
> bit-bending that can occur within a file that we may not always be
> aware of?
>
> AG
>
>
>
> ____________________________________________________________________________________
> Get your own web address.
> Have a HUGE year through Yahoo! Small Business.
> http://smallbusiness.yahoo.com/domains/?p=BESTDEAL
>
> ==============================================
> List usage info: http://www.zuikoholic.com
> List nannies: olympusadmin@xxxxxxxxxx
> ==============================================
>
>
>
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|