Black Width (WAS: "Multiple Master" Approach to Profile Conversion)
Black Width (WAS: "Multiple Master" Approach to Profile Conversion)
- Subject: Black Width (WAS: "Multiple Master" Approach to Profile Conversion)
- From: Rick Gordon <email@hidden>
- Date: Thu, 1 Jun 2006 23:29:22 -0700
>The biggest issue you will discover is the total ink limit.
>
>Use a software such as ColorShopX to look at the Neutrals to see the shape of the Black curve, as well as checking the Black Max and total ink limit.
>
>This is interesting, or maybe you should read that as disturbing.
If I create two identical profiles except for black width (one set to 0%, the other 100%), and separate the same file once with each profile, I find the following:
1) Areas not containing black are identical when converted through either profile.
2) Maximum blacks are trivially different, and suprisingly to me, the one separated with 0% black width has a very slightly higher TIL.
BUT...
3) Areas containing some black have a great deal less saturation and are somewhat lighter when separated (using RelCol rendering) with 0% black width than with 100% black width. I might have suspected that the lower amount of black would be mitigated by sufficiently higher amounts of CMY to achieve the same luminosity and might perhaps have had more saturation -- but it's not the case, at least with ProfileMaker 5.0.5b.
I am therefore left to conclude that near maximum black width will produce better separations. I might have thought that being able to shape the black channel with width controls would have more advantages than I seem to be able to demonstrate to myself.
Can someone suggest what circumstances would make reducing black width advantageous, in the light of decreased saturation in GCR-range midtones?
Thanks,
Rick Gordon
------------------
On 6/1/06 at 2:36 PM -0400, Roy Bohnen wrote in a message entitled
"Profile Testing":
>Rick
>
>I do much the same process as you have outlined.
>
>I will do a 0% start, 10% start, 25% start, 50% start, and a 75% start for Black. I will then do a 90% Max, 95% Max, and 100% Max. The differing amounts are used in an attempt to locate the profile which has the best Dmax, or lowest L value, especially on an Epson printer with UltraChrome inks.
>
>The Black Width is easily tested by performing a conversion on CMYKRGB tone scales from 0% to 100% to see where and how the black channel will transform the secondary colors and where in the tone range.
>
>The biggest issue you will discover is the total ink limit.
>
>Use a software such as ColorShopX to look at the Neutrals to see the shape of the Black curve, as well as checking the Black Max and total ink limit.
>
>This is interesting, or maybe you should read that as disturbing.
>
>Make sure you click the Balance button each time you make a change and watch the curves change, then create the profiles.
>
>You can use Photoshop for the test I defined within ColorShopX. ColorShopX is from XRite and runs in demo for two weeks.
>
>Roy
>
>--------
>
>Date: Thu, 1 Jun 2006 01:00:37 -0700
>From: Rick Gordon <email@hidden>
>Subject: "Multiple Master" Approach to Profile Conversion
>To: ColorSync List <email@hidden>
>
>Have any of you explored an approach to profile conversion that is roughly analogous to the way multiple master fonts operate? For instance, using ProfileMaker methodology, if one were to create the following profiles from a single data set, and conforming to a single total ink limit:
>
>GCR1, 0% black start, 95% max black, 100% black width
>GCR1, 0% black start, 95% max black, 0% black width
>GCR1, 50% black start, 95% max black, 100% black width
>GCR1, 50% black start, 95% max black, 0% black width
>
>GCR4, 0% black start, 95% max black, 100% black width
>GCR4, 0% black start, 95% max black, 0% black width
>GCR4, 50% black start, 95% max black, 100% black width
>GCR4, 50% black start, 95% max black, 0% black width
>
>...and possible some other intermediate specifications as well, or possible also create identical profiles set to different gamut mapping schemes (such as LOGO Classic -> LOGO Colorful) one could make two separations set onto separate layers, and use layer opacity to effectively interpolate between two choices.
>
>Initial tests for me are proving quite interesting. Could anyone comment on their experiences with such an approach, or caveats to be considered?
>
>Thanks.
--
___________________________________________________
RICK GORDON
EMERALD VALLEY GRAPHICS AND CONSULTING
___________________________________________________
WWW: http://www.shelterpub.com
_______________________________________________
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