Re: Substrate Relativity Calculator
Re: Substrate Relativity Calculator
- Subject: Re: Substrate Relativity Calculator
- From: Claas Bickeböller <email@hidden>
- Date: Fri, 19 Oct 2012 14:54:51 +0200
Am 19.10.2012 um 14:33 schrieb Michael Eddington:
> Class brings up some good points, particularly points #2 & 3. With
> regards to point #1, I think real world testing and application have
> shown this method to have legitimacy.
The prediction is nearer to the real world than assuming that the printed colours (with constant density) will not be affected by the paper colour, that's right.
But even for the original application (backing conversion), there are many models that outperform this XYZ-scaling (or call it ICC relative colorimetry because in the end that is what it is) a lot when it comes to non-solid colours.
So the real world testing only proofs that colour is affected by the paper colour and thus digital proofs do not match a print with constant density on a paper with a different white point.
These tests do not proof that XYZ-scaling is a good method to predict the changes in an accurate way.
Best regards
Claas
>
>> On Thu, Oct 18, 2012 at 12:40 PM, Claas Bickeböller
>> <email@hidden> wrote:
>>> Dear Mike,
>>>
>>> the underlying calculation method was originally developed and tested to predict the colour-deviation of a solid ink when the measurement backing is changed. BTW: For this purpose (solid inks and change of backing) it works very well.
>>> Knowing this it is quite easy to understand what you will get from the calculation.
>>>
>>> If you have a colorimetric description of a print on paper 1, you feed this along with a whitepoint of paper 2 (or the same paper on a different backing, see above) to the calculation method
>>> the model will predict the outcome of a print with constant density (remember, originally it predicts backing differences for the same print) on the new paper.
>>>
>>> So if you used this new dataset for imageseparation and proofing and if the printer printed with the same densities that he would have used for paper 1
>>> you will have a match……… if the method worked.
>>>
>>> Things to consider:
>>>
>>> 1) The underlying model makes the assumption that the influence of the paperwhite on the resulting colour is linear from paperwhite to the blackpoint.
>>> This assumption is simply wrong. So the predicted colours for everything but the solids is at least inaccurate if not completely off (depending on the inks, the screening…)
>>>
>>> 2) Is it a good practice to keep the densities that were found to fit perfect for one paper for another paper?
>>>
>>> 3) With increasing deviation between reference data-set and real-world paperwhite the prediction get's worse.
>>>
>>>
>>> Regarding your question "acc. GraCol":
>>> If you ask, if it was according to ISO 12647-7 (this is the international standard for digital proofing), the answer is yes, if the proof is in tolerance
>>> when you compare reference printing condition (in this case the result of your calculation) and simulated colour on the proof.
>>>
>>> But what do you win, if you simulate something in tolerance that does not reflect reality…..?
>>>
>>> Best regards
>>>
>>> Claas
>>>
>>>
>>>
>>>
>>>
>>> Am 17.10.2012 um 19:39 schrieb Mike Stewart:
>>>
>>>> With papers changing the way they are; most do not fall within the
>>>> tolerances specified in 12647-2. I like this idea of the Substrate
>>>> Calculator. My question is this. Let's say we proof to GRACoL tolerances
>>>> on a paper that is within the specification. We now get a #1 or #2 paper
>>>> that is out of spec. I use the calculator, create a new CGATS file and
>>>> create a new ICC Profile. I use CGS ORIS Color Tuner and hone in my
>>>> proofing device (Indigo press) to within tolerance and proof my
>>>> files/images on the "new" paper. What have I achieved. I have created a
>>>> proof that the printer may have an easier time "matching"; but can I say
>>>> that this proof adheres to GRACoL specifications? If you were to read the
>>>> 12647 color bar it would pass only if I have the proper dataset loaded to
>>>> compare to using software such as ORIS Certified Proof. Someone else
>>>> checking my proof to GRACoL would say it "fails" because they would be
>>>> comparing to the "original GRACoL". What exactly have I accomplished and
>>>> how can it be classified as a specification.
>>>>
>>>> Mike Stewart
>>>>
>>>> _______________________________________________
>>>> 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
>>>
>>>
>>> _______________________________________________
>>> 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
>
> _______________________________________________
> 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
_______________________________________________
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