[argyllcms] Re: SPAM: The Eye One! It works!

  • From: Graeme Gill <graeme@xxxxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Fri, 10 Aug 2007 15:31:32 +1000

Ben Goren wrote:

Next,   I  think   it  would   be  reasonable   to  make   the  -H
high-resolution mode the default. I'd  think that the overwhelming
majority of  people would rather use  it than not. If so,  why not
make  the  common choice  the  default,  rather than  requiring  a

Firstly, it doesn't actually add much to the reading of reflective
samples, since the nature of them is smooth. The only noticeable effect
is on samples with sharp transitions, such as displays.

The main reason I don't want to make it a default, is that it's
accuracy is unproved. I've simply interpolated/extrapolated things
like instrument response and wavelength calibration curves, as well as
formulating finer resampling filters, and while the results look
like they're reasonable, I don't have access to reference laboratory
grade color references and instruments to verify the accuracy.

In a  similar vein...why not  make the inclusion of  spectral data
with  dispread be  the default? Even  if not  as many  people will
specify a different  white point for a monitor  profile, surely it
doesn't do any harm to include the data in the file?

It's a speed thing. It slows down some instruments to read
the spectral data, and it's not as often used for display
profiles. Since there is no illuminant or FWA to worry
about, the only possible use is for a different observer
model. I could make a different default for different instruments,
but maybe that would be confusing.

With  printread, I  like the  beep confirming  that it's  ready to
scan. But I'd  *really* like audible  feedback for the  success or
failure of  scanning. Perhaps one  beep on  success, and  three on
failure? That way,  I wouldn't have to  look up to make  sure that
the strip was read okay.

I thought about something like that, but there are cross platform
issues. Linux simply doesn't have a reasonable beep capability.
I've literally got to send a bell character to the console
to get it do something ! OS X is only slightly better. There are
other paths, but they are not simple, and create dependency and
configuration complexity. (On my test Linux box I can't even
get it to beep. I can get the screen to flash, but no beep
from Linux. I know the HW is OK because the BIOS happily
beeps on startup).

    Updating the calibration and log parameters to EEProm failed

I'm not  worried about it,  but I  suspect that this  is something
you'd want to fix / suppress / etc.

It's currently turned off, but I don't see the above message when
I run it, so I'm not sure why it's generated.

P.S. I created  a 16K patch target  for my monitor and  let it run
overnight -- something unimaginable  to me before. The improvement
in contrast, shadow detail, etc.,  is rather amazing. Now I'm just
worried that I'm going to have  to spend a couple hours scanning a
similarly-sized print target for my favorite paper.... b&

Some people do this sort of thing, but they are generally using
more automated instruments such as the DTP70 or iSis.

Glad you've had some success.

Graeme Gill.

Other related posts: