Datacolor User to User Support Group.

>I'm running XP Pro SP2, PFP 1.11 and started by profiling Hahnemuhle
>Photo Rag on my r2400.
>
>I've done both 729/225 patch readings and in both sessions going back
>to re-read mistakes during the session has caused the reading to
>suddenly go very light and off color (cool). Once in this state, all
>further readings are identical.

Identical, meaning: no matter what patch you try to measure, you get the
same (or roughly the same) numbers out of it...? (Just making sure)

Obviously: something's wrong here.

>I've tried the following without success:
>
>Recalibrate the spectro

That was definitely the right thing to try first.

>Close & reopen PFP
>Remove v1.11 & install v1.0

Good idea, but no. I wouldn't go back to 1.0, it shouldn't help.

>disconnect/reconnect the spectro at the USB port

Also a good idea to try (not that this should be necessary)

>remove/re-install driver

Good idea, again. Removing a USB driver in Windows is a terrible thing
to attempt. It's almost impossible to get all of the "hooks" out of the
registry.

The driver that the 1.1.1 installer puts in is named starting with
"dcsc" (rather than starting with "cm3", as in the 1.0 software), and is
a driver that is "signed" by Microsoft.

>
>the only solution seems to be to close and reboot XP.

And after you did that: it came up and started measuring correctly again?

>I've not nailed down exactly where it goes south... seems after X
>number of readings (perhaps 100 or more) on a new session. It's going
>to take a lot more time & reading sessions to narrow it down.

Hmmmm..........

Try this:

Start a new PFP session, and instead of measuring the target, go into the
Tools menu; calibrate it there; and just start taking measurements; leave
it on the same patch and just keep measuring by pressing the Enter key, so
you can (fairly quickly) get through a lot of measurements without worrying
about moving the spectro or paying much attention. Does it fail, if you run
it this way? (Maybe set the Measure command to write the measurements to a file
and then you can see, if it fails, how many measurements it took to get there
from the start of the sequence).

We haven't heard of this before, as far as I know.

We'll do some similar tests to try to replicate this; it just hasn't been
reported by anyone else to date.

>Other than this major irritant,

(I certainly understand that; we'll be on this first thing on Wednesday,
after the holiday, and I'll spend some time looking at it myself tonight, to
see if I can at least repeat this, or (hopefully) not).

>the corrected profiles so far are
>excellent. Any reports of something like this before, or am I unique
>(grin)

(see above: so far, you seem to be unique, and that, in this case, is a
good thing...:-)

Don't worry: whatever's going on here, we'll figure it out and fix it for
you.


Best regards,


-- 
David Miller
Senior Software Developer, Digital Color Solutions
ColorVision