I'm afraid I spoke prematurely and incorrectly based on a single
corrupted file instance. I assumed it was corrupted by my shutting the
power switch off but I've just demonstrated to myself several times that
I can't cause file corruption by turning off the power switch or by
opening the CF door for that matter.
The power switch is a soft switch just as I would have expected it to
be. You can turn it off but, as long as data is being written from the
buffer, the red I/O light is on and the camera does not power down.
The corrupted file I had yesterday was, as near as I can tell, the file
that would have been being written when I turned off the power.
However, try as I may, I can not replicate that problem. Chalk it up to
cosmic rays I guess.
Opening the CF door does exhibit different behavior from the power
switch but I think it's clearly by design. Opening the CF door
activates another soft switch that stops writing as soon as the current
file or file set (raw + JPEG) is written. It definitely does not stop
writing immediately and I could not force a corrupted file. However,
anything beyond one file or file set is discarded.
I say I can see that this is by design since I think the logic is thus:
"Uh, oh! You opened the damned door and I'm still writing. If you
hit that eject button it's possible that I might be writing the FAT at
that time and if the FAT gets trashed you're in the hurts... and there's
nothing I can do now to stop you if you suddenly yank out the CF card.
However, I'm fast enough that I can probably compete the current file
write before you can get your thumb on that eject button. I'm gonna
finish this one but then I'm gonna have to shut down so you don't
completely screw yourself."
Or something like that...
Chuck Norcutt
AG Schnozz wrote:
> Chuck Norcutt wrote:
>
>>Well, I didn't have to wait long. While out shooting today
>>with the camera set for both raw and large JPEG's I shut the
>>camera off prematurely. I thought the OFF switch was a soft
>>switch and that the camera wouldn't power down until any I/O
>>activity was completed. Apparently not so for the 5D.
>
>
> Does anybody else find this to be rather disturbing? Was this
> just a one-time quirk of the 5D or is this a normal
> design-flaw--you know, like prematuraly opening the CF door
> while it is still writing?
>
> I shudder to think that it would be this easy to corrupt a CF
> card. No Canon-bashing, just highly concerned as I prematuraly
> shut-off my cameras or open the CF doors while file writing
> maybe once a month.
>
> AG
>
>
>
> ____________________________________________________________________________________
> Everyone is raving about the all-new Yahoo! Mail
> (http://advision.webevents.yahoo.com/mailbeta/)
>
>
> ==============================================
> List usage info: http://www.zuikoholic.com
> List nannies: olympusadmin@xxxxxxxxxx
> ==============================================
>
>
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|