Applying plate curves (Offset)

[eighthmonkey]

I have been having a discussion at work about whether the color bar should be curved when a curve is applied to the rest of the graphics on a plate. (Using a CTP)

Our customers are savvy enough to measure TVI and LAB values for quality purposes.

I am saying that we should curve the color bar when a curve is applied to the rest of the image. By doing this, we would know that we are hitting the SWOP targets of 25% - 16.4, 50% - 20.2, 75% - 14.4. In addition, I feel that if TVI measured on the color bar does not match the rest of the image, we would get confusion from the customer and some of our press operators would miss that correlation (in addition to needing to know what the curve is at).

Am I wrong in this? Should the color bars remain linear regardless of the curve applied to the rest of the image?

 

[meddington]

I am of the opinion that the colorbar should be curved in kind to the live image area, using the same screening, etc.

I am NOT of the opinion that there should be any color transform applied, for CMYK-CMYK or application of GCR., or at least there should be control elements that are free of such.

 

[Lukas Engqvist]

Colour conversion should not be applied, but the TVI curves are to "normalise" the dotgain. I would apply curve to the entire ripped data, so that dot gain and target values can be verified, what functions would the controlstip have if it were'nt so?

 

[eighthmonkey]

The co workers that I am having this discussion with argue that the color bar should not be curved so that we can know actual press TVI.

I agree that we would not know if we are hitting SWOP TVI targets if the color bar were not curved.

Please bear in mind that we are also collecting data with a press side spectro into a colormetric database so we can statistically process the data to improve process capability. I think the idea for them is to collect actual TVI for data collection purposes.

 

[meddington]

I understand their point, I guess I just don't care. Knowing the actual TVI of a linear tone scale on press would be interesting, but not really of value when running a job where plate curves have been applied. Where I would potentially want a "uncompensated" tone scale is on the plate, to verify that the imager is consistent.

 

[gordo]

The co workers that I am having this discussion with argue that the color bar should not be curved so that we can know actual press TVI.

I agree that we would not know if we are hitting SWOP TVI targets if the color bar were not curved.

Please bear in mind that we are also collecting data with a press side spectro into a colormetric database so we can statistically process the data to improve process capability. I think the idea for them is to collect actual TVI for data collection purposes.

This is a common point of confusion in the industry.

Because it changes with the print condition and screening - TVI is not the target. i.e. you are not trying to achieve 18% dot gain at 50%. The target is/should be the final tonality on press. i.e. at the 50% requested tone you want to measure a 68% tone. Although that sounds like saying the same thing it is conceptually very different.

For example, if you are using 133 lpi AM and 20 micron FM then they will need very different dot gains in order to deliver the same final target 68% tone on press. The same thing will happen if you use the same screening but papers of different absorbency.

The color bar in a curved plate will allow you to measure the final tone (e.g. 68%) on press at the requested 50% value. You then track that 68% tone value to make sure that it stays within your variation limits. E.g. if you allow plus or minus 2% then the measured tone value at the requested 50% can range between 66% and 70% and still be considered "in spec."

Tracking the variation of that target tone value over time gives you a process control metric that is more meaningful and useful than TVI.

The so-called "actual press TVI" is interesting but not particularly of any value because it is not an absolute - like "where I request 50% I expect to measure a 68% tone. Instead, TVI is a variable that changes according to conditions like screening, press condition, and substrate.

You can still build a colormetric database to statistically process the data to improve process capability with a curved color bar.

I agree with the others on this thread. The curve that's applied to the live image area should also be applied to the color bar.

best, gordon p

PS I somewhat slightly disagree with Mike when he wrote: "Where I would potentially want a "uncompensated" tone scale is on the plate, to verify that the imager is consistent." You don't really need to do that. If you request a 50% in the file and the applied curved results in a 45% tone value on plate then there's your target for verifying that the imager is consistent. I.e. when I measure the 50% patch on plate it should read 45%. If you have multiple curves then it would probably be easier if you had uncurved prescreened patches in the non-printing gripper margin area that you could measure for consistency. Those test patches would be applied to every plate imaged and because they have the same values irrespective of the curve being applied you could measure them to check plate imaging consistency.

 

[Bloodsaler]

Where you get 25% - 16.4, 50% - 20.2, 75% - 14.4 this SWOP target value?
Actually in our web,75% dot gain a little big bigger than 25% part,but the gray balance control well.

 

[Bloodsaler]

And there got some color strip that can contain linear part and the curve part(Such us AGFA plate control and Heidelberg plate control),So this not a problem~You can get all the information from this control strip~

 

[meddington]

PS I somewhat slightly disagree with Mike when he wrote: "Where I would potentially want a "uncompensated" tone scale is on the plate, to verify that the imager is consistent." You don't really need to do that. .

Totally agree, and in fact I am not currently using any uncompensated tone scale on plates. As you went on to state, for those using many different plate curves though, its a means to verify the imager independent of tone scale adjustment.

 

[rich apollo]

In the end, it really doesn't matter, as long as the procedure is decided upon up front and remains consistent. Consistency and repeatability are more important than accuracy.

Any gray patches WILL HAVE to be constructed in accordance to the decision made up front. For example, in G7 methodology, middle gray has been defined as 50, 40, 40. BUT, after calibration, you may find that your system actually achieves middle gray at 51, 38, 37. If you allow the colorbar to be curved with the rest of the imagery, then you can leave your gray patches at 50, 40, 40 and the platesetter will take care of the rest. If you leave the colorbars uncompensated you must either a) build the middle gray patch as 51, 38, 37; or b) run middle gray patches in your colorbars that are off color/density.