RE: New Topic
RE: New Topic
- Subject: RE: New Topic
- From: Steve Upton <email@hidden>
- Date: Fri, 8 Feb 2008 11:52:45 -0800
At 10:41 AM -0500 2/8/08, Mark Rice wrote:
>I think my main point is being missed. I don't really care of the
>linearization is neutral (although it is nice for making B&W prints with
>full ink density) but I want REPEATABILITY. Example - I once made a
>beautiful print of Halle Berry, with glowing skin tones. Since then, I have
>recalibrated -- and I haven't been able to get as good a print since then.
>A RIP system that recalculates some unknown curve value buried in binary
>data is never going to achieve repeatablilty.
agreed.
>
>What's this "so twentieth century thinking" business. I thought only valley
>girls talked like that. Density is an absolute measurement of light
>reflectivity or transmission. LAB is a relative measure that approximates
>human perception. One cannot be substituted for the other. BTW, Status E,A,M
>and T are not related to America or Europe - they are sets of filters with
>specified bandpass capabilities for denistometric measurements.
Strictly speaking you are correct (mostly). The problem is that densitometry tends to be more dependent on the instrument than colorimetry. Thankfully many instruments are now using spectro measurements to calculate density so this is more reliable and repeatable. But we are still left with multiple standards where they are not required and there's precious little benefit for the confusion.
Densitometry is still about a filter's response to a measurement (either physical or math filter). This means that differences in density can depend somewhat on the match between the filter colors and ink colors. (remind anyone of metamerism?) As a result, density numbers mean different things for different inks. A 0.05 difference on one ink set may be an acceptable shift but on another may be unacceptable. As we are ultimately talking about acceptability (by us and our customers) it makes sense to use a measurement system based on acceptability.
That doesn't mean that densitometry is useless. In a classic closed-loop production system it is still an effective quality control metric. But we now have instruments that gather much better data than we did in the past. It seems a shame to 'dumb down' the resulting calculations.
Also, while Lab has relative components, it's not entirely relative like, for instance, relative density is (not knocking density on this point, just using as an example). If it were truly relative then all paper white measurements would be Lab=100,0,0. In fact, Lab will tell us much more about a paper white than a density reading will, for instance.
A few points to ponder:
- Lab is really the most widely used, most absolute measurement of color that I know of (next to perhaps XYZ but that tends to be used more for emissive work).
- Lab values can be exchanged between people much more easily than density.
- Lab and XYZ values can be used for calibration quite effectively AND because they are colorimetric, the same measurements can extend the calibration to deal with gray balance and other things.
- Lab (and LCH) values can help diagnose hue-hooking problems which plague inkjets. It could be argued that without detection of hue hooking and ink limiting to avoid it's worst effects, that effective calibration is not possible on inkjets. Densitometry cannot help at all in this department except to suggest halting ink levels when density increases cease. A crude tool at best.
- If desired, alternate colorimetric numbers can be calculated from the spectral source measurements to evaluate ink mixtures for stability under differing lighting conditions. This can affect black generation algorithms.
Some of these things can be done with densitometry but the reasons to stick with densitometry are disappearing. It starts to become a "yeah you could, but why bother" proposition.
Just my 2 bits. I agree with your desires about RIP behavior. I'm not sure I've ever met a RIP I really liked. Just ones I disliked less than others. To be fair, it's a complicated and confusing process and they have come a long way...
It does raise a question worthy of another thread and perfect for this list:
In the age of spectral measurements and colorimetric (XYZ or Lab) interpretation, with an eye on the future, is there a need for densitometry any longer?
regards,
Steve
________________________________________________________________________
o Steve Upton CHROMiX www.chromix.com
o (hueman) 866.CHROMiX
________________________________________________________________________
--
_______________________________________________
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