Re: New EyeOne ruler and soft case
Re: New EyeOne ruler and soft case
- Subject: Re: New EyeOne ruler and soft case
- From: Steve Upton <email@hidden>
- Date: Wed, 2 Nov 2005 10:29:41 -0800
At 2:43 PM +0000 11/2/05, Steve Kale wrote:
>Unless I am mistaken the steps involved in profile creation are rather
>limited and straight forward. There are only a few things to screw up. One
>just has to be careful to do those few things correctly.
Overall, I can't disagree with your statement. The steps, in general, are limited and can be straight forward.
The problem lies in a few different areas:
- the system cannot be profiled using the standard steps - in other words, you follow the standard steps and you still don't like the results. OR
- the evaluation of the resulting color is flawed.
These two points actually encompass a large number of potential problems.
I'll take your printer profiling point as an example:
>
>Printer profiling is more or less the same common sense. Make sure you have
>good nozzle checks, make sure you don't colour manage the target when
>printing it, make sure it has dried adequately before you measure it and
>take care to ensure it isn't damaged beforehand, and make sure the
>spectrophotometer settings are set correctly (for most users these can not
>be changed).
all good things to do.
> The only other variable is deciding which target to print
>which again for most users is determined by the profiling package.
yes, this can be the case. Target selection though, can be a very important factor in profiling success and larger is not always better.
> The
>biggest issue with printer profiling is the inability in many instances to
>linearize the printer prior to profiling it so that sample stimulus-response
>behaviour has a much better chance of accurately estimating the entire
>behaviour. It would seem to me that the best way to improve printer profile
>creation is to demand linearization ability from the manufacturers.
I agree here as well.
If you involved a RIP in the printing process and you gain control over all ink channels, then you add at least the following:
- ink limiting (per channel)
- ink limiting (total)
- linearization or curving
- channel mixing (eg light cyan / dark cyan)
- screening - halftone or stochastic?
- resolution
If you are not using a RIP there is still a fair amount of experimentation you can do with different print settings to find the one that gives the most shadow detail, smoothest gradients, etc.
Then you can add:
- 3rd party ink choices
- OEM and 3rd party media choices
- printing environment
- head alignment
- ink settling in cartridges (a bigger factor than many realize)
- post-printing finishing such as lamination, coating, etc
- media manufacturing consistency
- ink and media storage and transportation issues
- printing consistency and whether multiple targets should be printed and averaged
- how long should the print sit and dry before measuring or viewing? is it susceptible to reactions with the atmosphere? (eg ozone)
And that's just getting a print you might be happy with. Then we get into the measurement & physics side of things:
- measurement consistency and whether single targets should be measured multiple times & averaged
- instrument filtration - which is a huge potential topic including:
- UV filter or not
- Polarization or not
- let the profiling software counteract paper brighteners in software or use a filter?
- paper backing - should a black or white backing be used. Which white? paper white?
- spectral or Lab?
- what is the role of metamerism failure in the ink set?
- are some of the inks themselves fluorescing?
Then we get into illuminants and lighting and viewing
- which lighting should be used to view the print?
- which illuminant should be used to calculate the colors, should it be the same as the viewing light source? (can it be?)
- how is the print viewed? Is the viewer far enough from the print? It's best in a light booth, etc.
- to what degree is the viewer color blind?
- how old is the viewer (the older the viewer the more corneal yellowing affects blue perception)
- basically, how close is the viewer to the standard human observer curves established in 1931?
Then we get into the math, which I honestly don't want to totally enumerate here. But let's mention a few:
- which rendering intent was chosen to create the print?
- was black point compensation used?
- which CMM did this magic?
- if gamut mapping was used (and it always is in some way) then how was it done? Would a profile from a different manufacturer give more pleasing results?
And then finally we get to the pure human factors:
- what is the desire of the viewer?
- what is the cultural background of the viewer? (which can significantly affect their tastes and expectations)
- did they make any mistakes along the way?
- did they eat a lot of peanut butter for lunch? (just checking if you have read this far - although foods can affect color perception)
The sad thing is I am sure I'm leaving a bunch of things out.
Why do I bother listing all these things? Basically because any of these could cause the basic steps to fail. If the user finds themselves in that situation they can either evaluate many of the applicable factors listed above, correct the problem if possible, and then reprofile OR break out the editor and correct the problem by hand.
Regards,
Steve
________________________________________________________________________
o Steve Upton CHROMiX www.chromix.com
o (hueman) 866.CHROMiX
o email@hidden 206.985.6837
o ColorGear ColorThink ColorValet ColorSmarts ProfileCentral
________________________________________________________________________
--
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Colorsync-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden