[argyllcms] Re: New to color management
- From: Leonard Evens <len@xxxxxxxxxxxxxxxxxxxxx>
- To: argyllcms@xxxxxxxxxxxxx
- Date: Thu, 24 Jan 2008 12:18:27 -0600
On Thu, 2008-01-24 at 15:51 +0100, Geert Janssens wrote:
> On Thursday 24 January 2008, Lars Tore Gustavsen wrote:
> > On 1/24/08, Geert Janssens <in> wrote:
> > > Hi,
> > >
> > > Let me start by saying I am new to Color Management. I would like to
> > > start profiling my work environment. I am working on linux, and my quest
> > > for decent profiling tools on that platform brought me to Argyll CMS.
> >
> > First what do you mean by work environment. Are you indicating
> > printer, monitor, camera,scanner and the ambient light in the room?
> >
> My work environment would be all of the above. I have
> * at least two, and possibly four computers I would like to calibrate, so
> that
> makes four monitors.
> * one networked Epson inkjet photo printer, used by all four the computers,
> * one Epson flatbed scanner attached to one pc, but useable by all other pcs
> over the network
> * one Canon EOS 400D camera
I use Vuescan (www.harmick.com) under Linux with my Epson scanner. You
can use the argyll programs to create a scanner profile, but I found
that the built-in Vuescan procedures for building profiles worked well.
That required IT8 targets which I got from www.targets.coloraid.de I
made a scanner profile using the scanner IT8 target, and I also
photographed the camera IT8 target using color negative film and use the
Vuescan procedure for profiling film. That is not supposed to work, but
it appears that it did. If you go this route, there are some confusing
aspects, so let me know and I may be able to save you some time.
>
> I have no idea (yet) how important ambient light is, and how to compensate
> for
> it. The primary PC is relatively far away from any direct outside light, and
> there is always artificial light in that room. The other PC's are more
> exposed. One is a laptop, and moves, so it will be hard to compensate for
> ambient light all of the time for it.
>
> My workflow is mainly consists of:
> * Input: different kind of scans, like images, drawings, and rather
> particularly in our case: real 3d jewellry. Additional inputs will be
> pictures taken with the Canon 400D
> * Processing in ICM-aware programs like Gimp
Just what gimp does with color management is still something of a
mystery. I haven't been able to find any good advice about it anywhere.
The gimp documentation is not too helpful. I can share some of my hard
won insights, and I would be happy to learn anything that you've managed
to figure out.
> * Output: partly for web, partly for print. Ideally our own printer should be
> useable for reasonably natural color representation. This is especially
> important for the jewellry. In the current configuration, all subtle color
> variations in for example gold are lost (it's currently not possible to
> distinguish a jewel in red gold from a jewel in yellow gold when printed).
>
> > > I am not sure what profiling device (of spectometer ?) I should buy. None
> > > are officially supported under linux by their respective
> > > vendors/manufacturers. I found Argyll CMS does support a number of them,
> > > so that's why I am here. I still have a few questions though before I go
> > > out and buy one. If I am on the wrong list with these questions, you can
> > > just point me to the correct one.
> >
> > I think you mix up colorimeter and spectrometers. You can read more
> > about the different instruments in chapter 4 of the book Understanding
> > Color Management by
> > Abhay Sharma online either at amazon.com or at books.google.com.
> > I also think you are on the perfectly right list. There are a lot of
> > pages on the net with buying recommendations and tests. The problem
> > with this is of course that they normally write about the bundled
> > software and the accuracy of the profiles created with that software.
> > I guess you are interested in the quality of the hardware.
> >
> I am indeed mixing up colorimeters and spectrometers. I think I read about
> the
> difference even in the Argyll CMS documentation, but I didn't realize it was
> that important.
>
> What I am mainly interested in is what tool do I need to get at least
> reasonable results. At first I thought I would require a spectrometer in
> order to accurately profile my printer. The Argyll CMS documentation also
> explains how to do this with a scanner, and explains the limitations of this
> method. But I have no idea how much accuracy I would loose with this method.
> What are others people's experiences here ?
I tried it with my scanner, but it didn't work very well. I was at an
early stage in the learning process, so perhaps I might do better now.
If you can afford it, I strongly recommend a measuring instrument such
as the Eye One Pro.
>
> > > * Is there a lot of difference in profiling quality between the different
> > > spectrometers ? I mean, given that all of them would be used with Argyll
> > > CMS, would one spectrometer result in a better (more acurate) color
> > > profile than another ?
> >
> > Someone else may answer this.
> >
> > > * If that is the case, what can you recommend ? What's your experience
> > > with different spectrometers ?
> > > * Reading through the Argyll CMS documentation, I read that I would
> > > require some test charts to profile printers and scanners. Are such test
> > > charts provided with the spectrometers or do I have to buy them separatly
> > > ? If so, where could I buy these ?
> >
> > The short answer are, it is possible to create printer charts with
> > argyllcms. You need a spectrometer or a scanner to "read" the chart.
> >
> Ok, thanks.
>
> > I think the odds are close to zero for using the spyder3 with argyllcms.
> > The spyder2 should work fine but you have use firmware stripped from
> > a dll file. Normally people on this list recommend the eye-one display
> > 2 or the cheaper huey, because of the firmware workaround. All this
> > devices are colorimeters. If you are looking for a spectrophotometer
> > look for a package called eye-one photo LT.
> >
> I suppose you mean that the firmware workaround required for the spyder2
> makes
> it less recommended than the eye-one display or the huey ?
>
> I looked around a bit for the two recommended models, and it seems I may find
> them. I also found one site that sells an eye-one photo package and was
> surprised of the difference in price: it's about 5 to 20 times more expensive
> than the colorimeters. I can begin to understand why the latter are popular !
>
> Anyway, thanks for the explanations.
>
> Geert
>
>
Other related posts:
