Contact the printer. There’s nobody here who is half as qualified to answer your questions as the one who will be working with the files you make.
You can use Custom CMYK to make a color conversion that has no black plate generation. It’s not the same as just pasting the RGB channels into the corresponding CMYK channels, which you can also do. It sounds like you’re going to want to experiment with whatever it takes to get the final image you’re after. You can also use Apply Image or Channel Mixer to blend channels into other channels.
The colors in the individual RGB channels will look totally different than the CMY channels.
Peter and Jason, thanks for the ideas. This will take lots of trials, since there are so many variables: the qualities of the original image, the channel separation, exposing and developing the plates, the color of the inks, and how much ink is applied, the type of paper and the pressure of the press.
Lots of play I’d say.
Best, David.
Creating a custom colour profile would also be a solution, but you’d probably have to make one for every ink set and press settings you use
I don’t know if this helps but when I first got into the printing industry 50 years ago, process colors meant 3 colors, CMY. Black would be made from these 3. Most presses were only single color so forgoing black was a 25% savings. The reality was, however, that black text was required so a black printer became economical. The black was to give definition and sharpen up the image. The dark velvet look was precisely what we wanted to eliminate.
Just removing the black now days probably won’t get that same effect because the images are scanned assuming the black printer will be present but your printer might know of a way to get this effect.
—
Hank Henniger
Dellas Graphics
835 Canal Street
Syracuse, NY 13210
315-424-7861 ext.19 (direct)
315-474-4641 (main phone)
315-474-4650 (fax)
wrote in message
I’m converting to CMYK to create a separate image from each channel, which
I then use to make etching plates. I’m trying four-plate images with CMYK.
First, my color setting for CMYK is U.S. Web Coated (SWOP) v2. Is this the
correct setting for what I’m doing?
And, I read about a printer who was creating etchings using 3 plates, one
each for C, M and Y, but no black. The images end up with "dark velvet" colors but no true blacks. Does anyone have a suggestion for how I’d create these three plates. I’m thinking maybe trying plates from RGB or us CMYK, but somehow overlay the black channel onto each of the other three, so each gets a bit of the info. from this channel.
Best, David.
And just as an FYI, I’m not looking for exact match on colors as you would with offset or inkjet printing. The variables are part of the beauty here.
d.
Scott, I am the printer. I’m doing photo etchings from my photographs.
This is handprinting, not offset.
David.
I understand a bit more what you’re doing now.
I would still bother making a custom profile. The idea being that it adds some predictability to the colours. And nothing prevents you from tweaking the colours at the print stage.
In my mind, for what you do, the idea behind profiling is more to get you close to what you want so you can spend your time fine tuning.
So, they’re hand-pulled by you on an etching press?
Sorry for a bit of confusion. Yes, these are hand-pulled prints on an etching press. I’m learning a photo-polymer process where I create a plate or plates of my image using ImagOn film mounted on plexiglass. Then ink these up and run through an etching press. So far I’ve only done one-color/ one-plate prints based on B/W images.
Nomad, I’ve had profiles made for my inkjet papers, and basically understand what’s involved, but I’m not sure how I’d do this for a hand-pulled print.
Best, David.
The thing is I think a full profile would only work if you were to use process colours (or something close to them).
You may want to look into creating a custom contrast curve instead since those could be applied no matter what colour ink you use.
It’s a somewhat lenghty process and I’d do it becasue I’m a bit of a CM geek.
Thanks c-nomad.
I use CM in my inkjet printing.
However, I think the photo-polymer etching process has too many variables for a custom profile. Besides, the inks I’m using don’t have "process" colors, only approximations, and they also need to be mixed with an ink "base" to make them more transparent for multi-plate printing, so another variable.
Still, I appreciate this discussion and input and have gleened some useful ideas from it.
Best, David.
However, I think the photo-polymer etching process has too many variables for a custom profile. Besides, the inks I’m using don’t have "process" colors, only approximations, and they also need to be mixed with an ink "base" to make them more transparent for multi-plate printing, so another variable.
I kind of got that earlier (hence my suggestion of a contrast curve instead)
That being said, for even the contrast curve to be useful, you have to control the entire process fairly tightly. This may or may not be desirable.
I did the contrast curve thing with digital outputs to fiber based B&W photo paper (hand processed) needless to say we spend a lot of time establishing making sure everything (chemistries, processing times, etc…) was as constant as possible. Some may not have found it worthwhile, but I did.
C-Nomad: I know what a curve is and what contrast is, but I’m not sure what you mean by the phrase "contrast curve"?
In making the transparency positive for exposing the film, you need to apply an adjustment curve, but it is sort of to flatten the highlights a bit — exactly the opposite of what I assumed you’d need to do.
Best, David.
Very few devices (none AFAIK) are linearized, all of them need to have a contrast curve applied to the file in order to be so. This may be done automatically by the RIP, or manually (which is extremely rare).
Now the device used to output your plates is probably linearized (at least is should be), but you may lose this linearization in subsequent steps of your particular printing process.
If I were you (and bear in mind, I’m a calibration and CM fan here) I would output myself a bunch of grey patches (I suggest using the values to create a custom dot gain profile) ranging from white to black) bring these to final output (ie press) then measure them with a densitometer.
You can then use these values to see how the output density (ie the measured density) compares with the input density (ie the numbers in PS) and use this to create a curve in order to have a linearized output.
Of course for the curve to be valid, you have to follow pretty much the same procedure every time you print.
(Hope I haven’t confused you too much here)
Edit: in my first paragraph, when I say "Very few devices (none AFAIK) are linearized" I mean of course uncalibrated, since calibration linearizes an output device.
Cybernetic Nomad,
But, if I have understood David all right, he’s going to print with inks mixed and prepared on his own for handprinting.
If so, I doubt he’s going to reach a really predictable output in consistency terms to be worth making a profile. (er.. Sort of trying to standardize (spelling?) moonshine?).
Just wondering myself.
Very few devices (none AFAIK) are linearized, all of them need to have a contrast curve applied to the file in order to be so. This may be done automatically by the RIP, or manually (which is extremely rare). . . . (Hope I haven’t confused you too much here)
C-Nomad, I don’t have clue what you’re getting to here. But that’s okay.
David.
C-Nomad, I don’t have clue what you’re getting to here. But that’s okay.
I had a feeling my explanations were getting confusing, oh well 😉
If so, I doubt he’s going to reach a really predictable output in consistency terms to be worth making a profile. (er.. Sort of trying to standardize (spelling?) moonshine?).
Gustavo, I sort of agree with you on this. Which is one reason I’ve mentioned being a CM geek/fan/enthousiast. I would do it because it’s the kind of thing that interests me.
That said, it may also be an interesting exercise in order to learn more about how the process reacts under various conditions.
PS: Isn’t standardized moonshine better known as Jack Daniel’s?
it may also be an interesting exercise in order to learn more about how the process reacts under various conditions
Yep. No doubt. I’d like to be there to see it (with a bottle or two of JD) ;P
David,
Cy-No is saying that no printing device behaves in such a way that when you feed it ink in a linear increasing amout (1,2,3,4…) prints a result that is really linear when measured (1,10,12,22…).
That’s is why you need to introduce a correction curve in the platemaking/printing proccess to force the machine to spit linear printed results (converting (1:1, 10:2, 12:3, …).
Linearization is sort of like an equalizer on your home or car stereo. It’s a compensation curve.
Thanks to all.
I don’t mean to be persnickity about this, but we’re not talking machine or offset lithography here — and the making of the "plate" (a thin light-sensitive film that I adhere to plexiglass) is more akin to photographic printing in the darkroom, not a metal plate for an offset press.
I’m first making a transparency positive of a photo–made either on an inkjet printer or a copy machine. This is used to expose or "burn" the plate (its a positive image as opposed to a negative that you’d use in photograpy), which is then developed in water and soda ash (vinegar stops development). The lighter areas are hardened by exposure to light, making a surface that doesn’t hold ink; the darker areas create an ‘etched’ textured surface that holds ink. The final plate is hand-inked and run through an etching press producing an image on dampened printmaking paper.
And as I noted before, there are many variables: the original image, the transparency, how long the film is exposed, how long its developed, how much ink is applied to the plate, what color/opaqueness/consistency the ink is, the pressure of the press, the color and surface of the paper and how damp it is. No RIPs or ink cartridges here.
Yes, I may have to apply a curve in Photoshop when making my transparency, and I may do as suggested and produce a grayscale strip for testing my plate-making and development, but this is again more akin to what I’m doing with my photography.
I don’t really understand the discussions of linearization or "splitting linear results," though I appreciate peoples’ interest and willingness to share.
Best, David.
PS – No JD in the house, only some Scotch.
