One application, or three, or five?

Here everybody can post his problems with PhotoLine
Martin Huber
Entwickler
Entwickler
Posts: 4186
Joined: Tue 19 Nov 2002 15:49

Re: One application, or three, or five?

Post by Martin Huber »

mwenz wrote: Fri 13 Sep 2019 19:32 Yes, but it is not the same thing. Placing an overprinted layer atop other layers does work as in other applications that can fake a duotone. But do try Photoshop or Corel's PhotoPaint and make an actual duo- or tri-tone image. Adjust their transfer curves. But use an actual photo.
What's the difference between the transfer curve in Photoshop and assigning a gradient in PhotoLine? IMHO by using the appropriate colors, the gradient has the same effect as a transfer curve.

Martin
mwenz
Mitglied
Posts: 122
Joined: Fri 13 May 2011 23:50

Re: One application, or three, or five?

Post by mwenz »

First, I am not saying that a gradient cannot not be used for some images. Tis a little more fiddly for some otherwise continuous tone images turned gray, but for some, sure. If a tri-tone is called for in the budget, multiple gradient layers with one or two or three spots that are going to be overprinted by the overprint attribute or say a multiply blend mode cannot even be screen proofed accurately. And there are likely going to be issues at print by laying down too much ink.

The main difference of gradients vs. transfer curves is akin to trying to adjust tonal curves in an image with only using a gradient with a blend mode. It can work on some images, not so well for most. It takes using curves and/or a histogram adjustment to get the most clear dynamic range because one can target lows, mids and highlights or clip either highs or lows while boosting mids, etc.

That is what the transfer curves can do over the tonal range when adjusting the two spots. Sometimes one needs that second spot to enrich the lows, hit the mids and without the highs getting colorant. Or enrich the lows and mids with a touch of colorant in the highs across the whole of the image wherever the highs are on the image. One may be able to fake that somewhat with different gradient types and multiple stops. But why? Why not just use an editor capable of doing that by adjusting the spot's curve shape?

I am pretty sure someone can articulate these and other issues far better than I can.
User avatar
Herbert123
Mitglied
Posts: 2202
Joined: Sat 12 May 2012 21:38

Re: One application, or three, or five?

Post by Herbert123 »

Martin Huber wrote: Fri 13 Sep 2019 21:13
mwenz wrote: Fri 13 Sep 2019 19:32 Yes, but it is not the same thing. Placing an overprinted layer atop other layers does work as in other applications that can fake a duotone. But do try Photoshop or Corel's PhotoPaint and make an actual duo- or tri-tone image. Adjust their transfer curves. But use an actual photo.
What's the difference between the transfer curve in Photoshop and assigning a gradient in PhotoLine? IMHO by using the appropriate colors, the gradient has the same effect as a transfer curve.

Martin
Martin Huber wrote: Fri 13 Sep 2019 12:25
mwenz wrote: Thu 12 Sep 2019 19:27 But I need real duotone images using certain (generally) Pantone colors which will separate to spot color plates. I do not see this as a possibility. At least when used in conjunction with layout software.
Maybe I misunderstand your intent and I don't know about that layout software part, but on export the attached document creates a PDF with a separate spot color plate.
The spot color itself has been created using the Document Color List (View > Lists > Document Color List).

Martin

SpotColor.pld
This process is called a "Duo-graph" or "fake duotone", isn't it? Printing a single colour with a one-colour halftone over it. It will lose much of the contrast of a photo compared to a regular "real" duotone, as far as I am aware, due to the lack of custom transfer curve control which duotones allow for. A duotone, tritone, etc. expands the range of possible values on paper. (Which is what mwenz also wrote).
/*---------------------------------------------*/
System: Win10 64bit - i7 920@3.6Ghz, p6t Deluxe v1, 48gb (6x8gb RipjawsX), Nvidia GTX1080 8GB, Revodrive X2 240gb, e-mu 1820, 2XSamsung SA850 (2560*1440) and 1XHP2408H 1920*1200 portrait
Martin Huber
Entwickler
Entwickler
Posts: 4186
Joined: Tue 19 Nov 2002 15:49

Re: One application, or three, or five?

Post by Martin Huber »

mwenz wrote: Sat 14 Sep 2019 05:18 First, I am not saying that a gradient cannot not be used for some images. Tis a little more fiddly for some otherwise continuous tone images turned gray, but for some, sure. If a tri-tone is called for in the budget, multiple gradient layers with one or two or three spots that are going to be overprinted by the overprint attribute or say a multiply blend mode cannot even be screen proofed accurately. And there are likely going to be issues at print by laying down too much ink.
I understand that a transfer curve is easier to edit than a gradient, but aside from that I don't see a difference, neither in on-screen display nor in printing. On-screen, spot colors have to simulated the same way for both methods. And why should there be a difference in printing?

Martin
mwenz
Mitglied
Posts: 122
Joined: Fri 13 May 2011 23:50

Re: One application, or three, or five?

Post by mwenz »

Martin Huber wrote: Mon 16 Sep 2019 09:27I understand that a transfer curve is easier to edit than a gradient, but aside from that I don't see a difference, neither in on-screen display nor in printing. On-screen, spot colors have to simulated the same way for both methods. And why should there be a difference in printing?
You are overlooking, or dismissing, this part of my post:
mwenz wrote: Sat 14 Sep 2019 05:18...The main difference of gradients vs. transfer curves is akin to trying to adjust tonal curves in an image with only using a gradient with a blend mode. It can work on some images, not so well for most. It takes using curves and/or a histogram adjustment to get the most clear dynamic range because one can target lows, mids and highlights or clip either highs or lows while boosting mids, etc.

That is what the transfer curves can do over the tonal range when adjusting the two spots. Sometimes one needs that second spot to enrich the lows, hit the mids and without the highs getting colorant. Or enrich the lows and mids with a touch of colorant in the highs across the whole of the image wherever the highs are on the image. One may be able to fake that somewhat with different gradient types and multiple stops. But why? Why not just use an editor capable of doing that by adjusting the spot's curve shape? ...
One cannot ever use a gradient for a photograph and achieve anything more than an approximate of what an image is capable of using true duotones.

This resistance to at least research, to try and understand the subject/issue, is why I haven't brought this up in several years. If you do not desire to understand this issue, fine. I have other applications I can use.

Mike
maxwell
Mitglied
Posts: 222
Joined: Mon 01 Jul 2013 20:53

Re: One application, or three, or five?

Post by maxwell »

A good base for discussing duotone and PL's "fake" realization is given in

https://creativepro.com/extending-tonal ... -duotones/

I formerly learned to delete transfer curves in PDF/X, because they are strongly print machine dependent. There are cases, where transfer curves make sense as can be seen in the above discussion. Thus, you can learn something new every day.

Thanks for discussion

Josef
Martin Huber
Entwickler
Entwickler
Posts: 4186
Joined: Tue 19 Nov 2002 15:49

Re: One application, or three, or five?

Post by Martin Huber »

mwenz wrote: Mon 16 Sep 2019 15:20This resistance to at least research, to try and understand the subject/issue, is why I haven't brought this up in several years. If you do not desire to understand this issue, fine.
I am really interested in understanding this problem, but I simply can't imagine a type of dutone image which can't be simulated using gradients and overprinting.
If there is none, and the problem is the more difficult handling of gradients compared to transfer curves, then I already understood the problem.
If theres are any, maybe you can create a small sample pdf containing one.

Martin
mwenz
Mitglied
Posts: 122
Joined: Fri 13 May 2011 23:50

Re: One application, or three, or five?

Post by mwenz »

Martin Huber wrote: Thu 19 Sep 2019 15:13I am really interested in understanding this problem, but I simply can't imagine a type of dutone image which can't be simulated using gradients and overprinting.
If there is none, and the problem is the more difficult handling of gradients compared to transfer curves, then I already understood the problem.
If theres are any, maybe you can create a small sample pdf containing one.

Martin
Hello Martin.

I have linked to a 7zip file that contains the jpg I used, the duotone dcs-type eps file, and a resultant pdf. The pdf is how the eps file appears on-screen in ID/QXP.

https://www.dropbox.com/s/yuocmo9bsl6lj ... _6.7z?dl=0

Thank you, Mike
maxwell
Mitglied
Posts: 222
Joined: Mon 01 Jul 2013 20:53

Re: One application, or three, or five?

Post by maxwell »

With respect to Pdf/X-3 embedded transfer functions are forbidden. This is not the case with eps files in prepress. Thus, eps duo tone are not really of interest in this discussion.
Using transfer functions in the context of pdf, they have to be preprocessed while building the pdf file. Studying literature I have to support Martin, where and in which cases trandfer functions cannot be replaced by gradation curves.
mwenz
Mitglied
Posts: 122
Joined: Fri 13 May 2011 23:50

Re: One application, or three, or five?

Post by mwenz »

maxwell, I have no idea of the "why" of your post.

yeah, pdf doesn't support a whole lot of things that are native constructs in an application, so what? I was asked for a pdf, I supplied one. I also supplied the dcs-eps so it can be inspected--in PS or another image editor thus capable. That's all. I also included the image so Martin, or whomever wants to, can play with the same continuous-tone image.

I often use .psd or pdf files versus dcs-eps files when placing into the layout application, but the dcs-eps files are typically smaller. As well, all I care about is putting these duo- tri-tone images into the layout application intact--so all that matters is that the colorants separate properly when the final pdf is generated. Therefore it doesn't matter what format is output from the duo-tone generating application is. I am not requesting dcs-eps support be added to PL. It's pdf output is fine with me because the spot survives and the final pdf is correct.

However, first PL needs to handle spot colors properly when used for the various types of creative functions they can be used for. And it needs to be able to create them quickly for their intended purpose. I myself don't even care if PL can ever even open them when originated in another application, though such .psd file compatibility would enhance the value to others I am sure.

Mike