Re: Theory versus practice
Re: Theory versus practice
- Subject: Re: Theory versus practice
- From: Steve Kale <email@hidden>
- Date: Wed, 05 Oct 2005 18:48:55 +0100
- Thread-topic: Theory versus practice
Title: Re: Theory versus practice
The driver simply tells the printer what inks to fire and how much for each possible 8 bit pixel value. In your greyscale example it received a pixel value of 0. Pixel value 0 is printed at maximum K ink load which happens to produce L*14 on the paper you used. So on and so forth until no ink is laid down. Pixel value 0 will always correspond with defined maximum ink lay down and 255 with paper white or no ink lay down. What counts is the “shape” of the response in between.
Black Only printing is used by some people as an easy way to avoid the introduction of colour shifts/metamerism that is evident in many colour ink workflows – it simply and crudely prints with just the K ink. The issue with Same as Source, incl. “Black Only”, printing is that you have not profiled the shape of that stimulus-response behaviour. You get what you are given and there is no management of luminance between the two. In the case of Black Only it may not even be linear (an even progression in L* from dMin to dMax). It may not have a gamma anywhere near that which you would prefer or observe on screen.
Black Only and other Same as Source workflows can be improved significantly by profiling the luminance of its greyscale output and then using colour management to manage the compression from file (0-100, GG2.2 for example) to the print “space” (14-96, unknown gamma). That is to say, moving away from a Same as Source workflow. Most profiling software is geared for colour work only and is not helpful in this regard because it tries to manage hue as well. However, Roy Harrington has written a cool little application to profile the luminance of greyscale output (using an Eye-One Photo spectrophotometer). One profiled you can convert to this profile prior to leaving PS so that there is an appropriate compression of the image luminance in the narrower (in dynamic range terms) Black Only on matte paper print space. Luminance compression is managed but there is no attempt to correct hue which is determined by ink selection and paper colour. The application is a module of Quadtone RIP and is called QTR Create ICC (QTR is a shareware product and costs $50.) Very cool.
Steve
From: eugene appert <email@hidden>
Date: Wed, 5 Oct 2005 11:03:19 -0400
To: <email@hidden>
Subject: Theory versus practice
If I weren’t holding the target sample print in my hand I wouldn’t have had the confidence to ask this question.
I overheard a technician suggest to a student who was printing greyscale on an Epson 7500, to send the file “ same as source” and print with black ink only “to get the best results” Of course the technician was focusing on the colour inconsistency problem with pigmented inks, we generally refer to as “metamerism”
I decided to be prudent and actually test the theory before I bloated out about how preposterous it would be to send a file in its workspace, with a dynamic range of L*0 to L*100, without any conversion, compensation or compression to a printer with a maximum shadow at about L*14.
I sent a target which is basically a series of pairs of black patches with values of L*1 difference that I use to discern maximum shadow and minimum highlight. For example the first patch is L*1 in L*0, the second L*2 in L*1 etc. When I send this target using relative with no black point compensation to the 7500, the first line appears in the L*14- L*15 patch. Using perceptual, depending on the profile, the line appears in the first or the second patch.
The Greyscale 2.2 to “same as source” test print produced lines in every patch.
In all other circumstances, converting a file to its output profile “en route” as Photshop allows us to do, or actually converting the file before hand and using the “same as source” option, produces identical results. To me this means the “same as source” option is performing exactly as it is meant to and leaving the file as is, transferring the data without any modification.
So how, or what, is translating L*0 to L*14? and subsequently compressing all the other values up the curve????
Can any one help me understand what is happening?
Eugene
_______________________________________________
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