Pages: [1]   Go Down

Author Topic: I1Profiler Custom Illuminants and M0/1/2 - Suggest using only M2  (Read 271 times)

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1961
I1Profiler Custom Illuminants and M0/1/2 - Suggest using only M2
« on: February 05, 2020, 05:33:17 pm »

Summary. Whenever you use a custom illuminant to make a printer profile, use M2 Measurements or OBA free paper.

I1Profiler can take measurements that are based on M0,M1, and M2 where M1 has full D50 uV levels and M0 has about 2/3rds of the uV that M1 has. When one of the M illuminants is selected to measure a chart, the measured data includes whatever the M1/M0 uV part bumped the visible blue part of the measured spectrum. For instance, 440nm is often bumped to 120% reflectance! This is then factored in as if the paper reflected these levels of light from the custom illuminant. Whether the custom illuminant has a lot or little uV is ignored. For that matter, custom illuminant spectra only goes to 380nm. While some of this would be converted to visible light I1profiler has no way to know and simply treats it as regular visible light (with little effect at these wavelengths).

Since most custom illuminants have relatively small amounts of uV, measurements made with M0 or M1 should not be used on any paper with significant OBAs as they will significantly distort the profile.

Another reason to avoid OBA paper.

Added:
This is an example of how M1 measurements distorts I1Profiler profiles. I used a custom LED illuminant with no uV (very little light below 420nm and less than .01 below 410nm). The LED doesn't activate OBAs.  Using high OBA paper and this LED custom illuminant there should be no difference between M1 and M2 profiles. But making a profile using the M1 measurements I get a white point that has a b* of -14, a pretty blush white and it soft proofs with a strong bluish tint using show paper color. With M2 (uV cut) and the same LED illuminant, the white point is near neutral and the soft proof is no longer shifted to the blue.
« Last Edit: February 05, 2020, 10:00:26 pm by Doug Gray »
Logged

GWGill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 553
  • Author of ArgyllCMS & ArgyllPRO ColorMeter
    • ArgyllCMS
Re: I1Profiler Custom Illuminants and M0/1/2 - Suggest using only M2
« Reply #1 on: February 06, 2020, 07:07:19 am »

Whether the custom illuminant has a lot or little uV is ignored. For that matter, custom illuminant spectra only goes to 380nm. While some of this would be converted to visible light I1profiler has no way to know and simply treats it as regular visible light (with little effect at these wavelengths).
Note that ArgyllCMS's FWA compensation will model a custom illuminants effect on fluorescent paper whitener appropriately. Yes it needs the custom spectrum values down to 300nm to do this.
Logged

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4007
    • A Goldhammer Photography
Re: I1Profiler Custom Illuminants and M0/1/2 - Suggest using only M2
« Reply #2 on: February 06, 2020, 07:54:54 am »

I remember when I first profiled Ilford Gold Fiber Silk which only had modest OBA content, I was fooling around with different illuminant values using Argyll.  When I created a profile and then went to soft proof, I got a significant blue cast with the profile.  It's been a while and I cannot remember the exact setting but it did point out that with even small amounts of OBAs, things can go wrong.
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1961
Re: I1Profiler Custom Illuminants and M0/1/2 - Suggest using only M2
« Reply #3 on: February 06, 2020, 03:56:01 pm »

I remember when I first profiled Ilford Gold Fiber Silk which only had modest OBA content, I was fooling around with different illuminant values using Argyll.  When I created a profile and then went to soft proof, I got a significant blue cast with the profile.  It's been a while and I cannot remember the exact setting but it did point out that with even small amounts of OBAs, things can go wrong.
Basically, almost all illuminants that aren't D50 cause much lower fluorescence. So M2 measurements should be what one uses when selecting the custom illuminant. CFL and regular fluorescent tubes have a bit of uV bu still much less than D50 levels. So even those should use M2 measurements. Even M0 has much too high a uV component.

There is one way to adjust this. Take M1 and M2 measurements and subtract them. This produces the visible fluorescent response. By taking a small fraction of this, around 20%, and adding it back to the M2 patch target measurements then using these response spectra, one gets a good profile that includes the small amount of uV the illuminant has. This has worked well for some F8 lamps but I can't say for other types.

Alternately, XRites' OBC process should do a reasonable job with custom illuminants but I haven't tested it.

The whole thing's a bit of an outlier and can be completely avoided by using OBA free paper.
Logged
Pages: [1]   Go Up