ChartThrob + Printer Adjustment Curve Question

[NoSuchNick]

Hi everyone,

I'm working on getting my digital film curves under control for Photogravure. I've already made a successful printer adjustment curve for my workflow. I ran a ChartThrob grid through my whole process, scanned the print, and generated a final curve using ChartThrob's Photoshop script. Here's where I run into a gap in advice online. For all my films going forward am I, 1 - ditching the printer adjustment curve and only printing using my new curve generated by ChartThrob, 2 - using both curves but ordering my curve layers with the printer curve on top of the ChartThrob curve in my layers panel, or 3 - using both curves and ordering my curve layers with the printer curve below the ChartThrob curve in my layers panel. All three radically change the appearance of my films. I was curious if anyone had an answer, or if they noticed I could be doing things a little differently. I would appreciate any advice!

<3

 

[koraks]

Did you apply a curve to the original charthrob calibration chart? If so, then you'll have to do the same with the other images.
I've never used chartthrob, but I think you're not supposed to run an adjustment curve over the original chart you use as the input for linearization. Technically, it'll work, but it complicates matters.

 

[nmp]

What do you mean by "printer adjustment curve?"

 

[NoSuchNick]

My printer adjustment curve creates more even ink densities across my film. It compensates for uneven ink densities that one's printer might create on factory settings. A 10% gray in Photoshop may print as 5% gray read through a densitometer. The curve ensures that 10% gray on Photoshop prints as a 10% gray read through a densitometer.

The only curve I applied to the original ChartThrob was this Printer Adjustment Curve to make its ink densities more suitable for exposure. If I hadn't I would have been getting an extremely contrasty film, which the ChartThrob can't really handle well.

 

[koraks]

The process you describe w.r.t. ink densities is what we call 'linearization'. That's what tools like Chartthrob are designed for.

If I hadn't I would have been getting an extremely contrasty film, which the ChartThrob can't really handle well.

In this case, you can indeed apply a pre-linearization curve the way you're doing, and then put another curve on top. It's not unheard of; Calvin Grier does something similar, although sort of the other way around: he overlays a fine-tuning curve on top of a crude linearization curve. The principle is pretty much the same.

It'll work, but ensure that you're working in 16bit pixel depth to avoid posterization.

 

[NoSuchNick]

The process you describe w.r.t. ink densities is what we call 'linearization'. That's what tools like Chartthrob are designed for.



In this case, you can indeed apply a pre-linearization curve the way you're doing, and then put another curve on top. It's not unheard of; Calvin Grier does something similar, although sort of the other way around: he overlays a fine-tuning curve on top of a crude linearization curve. The principle is pretty much the same.

It'll work, but ensure that you're working in 16bit pixel depth to avoid posterization.

Overall a pretty minor question but your answer saved me three test plates and a few days of work. Thank you so much!

 

[koraks]

You're welcome! Hope it helps. Looking forward to seeing some prints when you get to that point! It's a beautiful process, photopolymer intaglio.

 

[nmp]

I will disagree with @koraks here and recommend that the curve generated by Chartthrob should be below the printer adjustment curve (if I understand it correctly.) Check out this discussion that we had some time ago:

Using 2 Correction Curves to make Digital Negatives

I have had this question for a while: Can you use a second correction curve as a way to fine tune the the first one if the latter didn't quite give a good enough linearity in the print? My intuition was that you should be able to. A couple of times I tried this, the results were not consistent...

www.photrio.com

 

[koraks]

@nmp I remain skeptical.

 

[NoSuchNick]

I suppose I'll try both if both appear to work for people. I'll just pick which yields the most interesting results. @nmp, that thread seems to have some compelling info on merging curves too. Thanks again to you both.

 

[nmp]

@nmp I remain skeptical.

You can look at it this way: ChartThrobe maps the original digital file vs measured values, not what happens to that data such as inversion, flipping, colorizing, as well as application of a printer profile (which is what OP is in essence doing by adding a printer adjustment layer.) So the ensuing correction curve applies to the original data not converted data. All the adjustments prior to sending to the printer become a part of black box which should be repeated the same way for the image file, i.e. image layer + ChartThrobe correction curve + Printer adjustment layer + everything else. You can simulate this for a given Charthrobe input/output relationship and see which scenario gives one a diametric straight line which is what I did to satisfy myself that indeed that was true.

:Niranjan.