Making halftone screen negatives at home: possible?

Alright, this is a long shot. A very long one.

I've been messing around with carbon transfer lately, particularly color carbon. For black & white I like to use with camera negatives, but for color, I don't see this as a realistic option (i.e. making color separations using an enlarger). So it's going to have to be digital negatives.

So far I've done the usual thing with inkjet (quasi) continuous tone negatives, but it's an inherently flawed approach. Basically, a decent (high resolution) halftone screen approach would be far superior in several ways. But...there are limitations. For instance, I'm looking to do this myself. I'm not considering using a commercial service that I send files to and they send negatives back. I know that it's a possibility and also know a few very respectable/appropriate providers, but I also know how I explore printing processes and the turnaround time for a commercial service would make it a non-starter for me.

I've been wrecking my brain on this one, but all I can come up with is dusting off an old & beat up imagesetter scavenged from the dumpster of a print shop somewhere. And I'm frankly not sure if I want to go there. Realistically it's a €10k startup cost to get one in working order and then it's probably going to take far more religious ritual & shamanic efforts than I can muster to keep such a beast going for any useful period of time.

What are your thoughts on the matter? Any miracles up your sleeves?

You should be able to output color separations from InDesign or Illustrator. I even think you can define the line screen and output as a negative. Haven't done anything like tha in a while, but I remember the capability. Adobe allows you to download trial versions of the software, just be sure to cancel the subscription before the end of the rial period. Unless, of course you will want to use it regularly.

Yes, but output to what kind of device? As I said, inkjet halftone screens will have poor halftone dot geometry and as a result poor resolution as well as poor tonal transitions. Laser isn't going to be all that much better, either.

@sasah zib please feel free to share your darkroom-oriented approach as well! I posted this in the digital section because I guessed the chances of any solution being (partly) digital would be the biggest. If, for instance, there is a good way to 'convert' a continuous tone film/transparancy-based image into either a frequency- or amplitude-modulated halftone image, I'd be very interested in this as well.

I don’t quite understand your query, but Polaroid sold kit that allowed halftone negatives to be made for publication with P/N film. Would something like that help you? I’m not sure if I still have the kit.

It might; it sounds interesting in any case.
In essence, my question is pretty simple: I'd ideally like to make good-quality halftone negatives for contact printing, particularly carbon transfer. I'm currently exploring the various possibilities for doing so, so my question should be seen in that light.

koraks said:

Yes, but output to what kind of device? As I said, inkjet halftone screens will have poor halftone dot geometry and as a result poor resolution as well as poor tonal transitions. Laser isn't going to be all that much better, either.

Click to expand...

Try contacting a printer (as in printing press) or pre-press firm and see if they can or will output your files to litho film.

Yes, I considered that, but as I wrote in the first post, I'm looking for in-house possibilities for now. I'm aware of a couple of places I could send my files to and get imagesetter negatives back, but the turnaround time is something that bothers me especially in a testing/exploration phase.

when I used to say 'screened' negative, it meant that "dot" thing. They can become very complex. There was a trade group, many specialty schools, etc. GATF was on US association with texts, etc. I'm assuming that you know of some of this since you use "dot gain" in reference to an error you get.

Kodak and Agfa made screens of different sizes, dot shapes, density, etc. These things were tossed by the pallet in the early 90s. Kodak had a booklet Q-21 Kodak Contact Screens with product specifics. You would probably want the 'grey' not the magenta screen since you would be doing post-masking, if any.

other than Kodak, Agfa, Caprock, there were other firms. Caprock still has a website.

the other approach I suggested is "dust box" method used in aquatint. Takach still lists their ready-made tint screen:

These aqua tint screens are custom-made random pattern stochastic tint screens. Their random dot pattern eliminates the mechanical patterns of halftone screens. Similar to traditional aquatint on copper plate which utilizes powdered rosin as an acid resist, the aquatint screen masks areas of the photosensitive material from ultraviolet light creating a non-mechanical structure of intimately spaced ink wells in areas that would otherwise "open bite". The resulting tonal ranges appear organic.


https://www.takachpress.com/access/screen.htm

Click to expand...

somewhere in Europe some school must still teach etching, they would have a library and someone able to demonstrate the use of such method as AT.

hope this doesn't take you off-path --

ahh, one other thing: the finer dot methods will require "fine line lith" procedures (2 dev baths; different dilutions). That isn't a simple answer. (in this instance i use "lith" meaning lithfilm processing, not lith-printing )

sasah zib said:

when I used to say 'screened' negative, it meant that "dot" thing. They can become very complex. There was a trade group, many specialty schools, etc. GATF was on US association with texts, etc. I'm assuming that you know of some of this since you use "dot gain" in reference to an error you get.

Kodak and Agfa made screens of different sizes, dot shapes, density, etc. These things were tossed by the pallet in the early 90s. Kodak had a booklet Q-21 Kodak Contact Screens with product specifics. You would probably want the 'grey' not the magenta screen since you would be doing post-masking, if any.

other than Kodak, Agfa, Caprock, there were other firms. Caprock still has a website.

the other approach I suggested is "dust box" method used in aquatint. Takach still lists their ready-made tint screen:



somewhere in Europe some school must still teach etching, they would have a library and someone able to demonstrate the use of such method as AT.

hope this doesn't take you off-path --

ahh, one other thing: the finer dot methods will require "fine line lith" procedures (2 dev baths; different dilutions). That isn't a simple answer. (in this instance i use "lith" meaning lithfilm processing, not lith-printing )

Click to expand...

In traditional 4-color CMYK press printing, there are dedicated screens for each color, making a distinctive "rosette" pattern if you look at the printed image with a loupe. So I assume that whatever method you end up with, you will need 4 screens with the corresponding angles or if going stochastic there are similar requirements so the colors don't overlap.

Pieter12 said:

So I assume that whatever method you end up with, you will need 4 screens with the corresponding angles

Click to expand...

That's right. Or...just rotate the same screen, provided it's big enough. But you still have a point of course.

sasah zib said:

when I used to say 'screened' negative, it meant that "dot" thing.

Click to expand...

Yeah, we're on the same page. And I'm indeed aware of the basics, but with little to no hands-on experience.

As to the etching thing: I did photopolymer for a brief while. I used the inkjet dot 'screen' (black only), and it was...well, kind of poor in quality. I considered proper aquatinting with a rosin box, but decided I was giving my lungs enough abuse already, so I decided against it. I did do some experiments with enlarging film grain (Delta 3200 I think) onto x-ray film and develop to high contrast to try and make a stochastic screen that way. Evidently, this sucked...badly

So long story short, I've pondered a couple of things, tried one or two over the years, but the challenge is kind of different now and I've learned a bit in the meantime as well.

sasah zib said:

Takach still lists their ready-made tint screen:

Click to expand...

I think there's several; at least I came across a few when I did intaglio. I considered them for this as well, and it's still a possibility that I might try. I'll have to think about it for a bit.
I was kind of hoping there's some magic one could work on an existing inkjet printer; I looked into QuadTone RIP (I've used it on occasion throughout the years for other purposes) but it doesn't really RIP in this sense, it seems. Then there's GhostScript & GhostView, but the documentation is so...Unix-y that it's hard to even figure out what it does and doesn't do.

koraks said:

I did do some experiments with enlarging film grain (Delta 3200 I think) onto x-ray film and develop to high contrast to try and make a stochastic screen that way.

Click to expand...

Litho film rather than X-ray - and you'll need to get a hard dot for etching/ polymer gravure use. Etched glass of various sorts and some patterns of diffusion material can also be co-opted into acting as halftone screens (again, it depends on which process you are going for as to whether you'll need to harden the dot).

When we were using halftone screens for silk screen prints, we bought them from art supply stores. Try this below, not sure if they would work for your purposes though.

koraks said:

Alright, this is a long shot. A very long one.

I've been messing around with carbon transfer lately, particularly color carbon. For black & white I like to use with camera negatives, but for color, I don't see this as a realistic option (i.e. making color separations using an enlarger). So it's going to have to be digital negatives.

So far I've done the usual thing with inkjet (quasi) continuous tone negatives, but it's an inherently flawed approach. Basically, a decent (high resolution) halftone screen approach would be far superior in several ways. But...there are limitations. For instance, I'm looking to do this myself. I'm not considering using a commercial service that I send files to and they send negatives back. I know that it's a possibility and also know a few very respectable/appropriate providers, but I also know how I explore printing processes and the turnaround time for a commercial service would make it a non-starter for me.

I've been wrecking my brain on this one, but all I can come up with is dusting off an old & beat up imagesetter scavenged from the dumpster of a print shop somewhere. And I'm frankly not sure if I want to go there. Realistically it's a €10k startup cost to get one in working order and then it's probably going to take far more religious ritual & shamanic efforts than I can muster to keep such a beast going for any useful period of time.

What are your thoughts on the matter? Any miracles up your sleeves?

Click to expand...

I am just beginning along a very similar journey to find a solution that can work with inkjet digital negatives. First initial idea is generating multiple negatives that are representative of luminosity tones. For the highlights I am thinking layering a generated digital negative and 80% aquatint screen.

I have sample screens in hand, just need to get time to some testing to see how viable this approach is.

Thanks for responding @DPVisions!

I've given the aquatint screen some more thought. It won't work for me, I think. Aquatint is nice because of what it was intended for initially: creating a pattern of bumps in solid black areas so that there's something to trap ink on a printing plate. My application at least is different, since carbon transfer has no trouble printing solid black patches - or any other tone, except very light ones! Think about it: what do you get when you overlay an aquatint screen, or rather, a stochastic screen over a continuous tone negative: that's right, continuous tone dots and blank areas in-between (depending a bit on the average coverage of your screen). For carbon printing, or any other process that doesn't happen to be based on some form of intaglio, this won't do any good.

If you happen to be doing intaglio printing, then of course a regular stochastic screen will be fine.

However, you're onto something:

DPVisions said:

generating multiple negatives that are representative of luminosity tones

Click to expand...

That's indeed what tonal separations do. If you read Calvin Grier's eBook on calibration, he goes into this in detail in part 3. While there are some obvious drawbacks, tonal separations will 'fake' additional tonal resolution, so a set of inkjet negatives might just look a little less horrible than a single halftone screen inkjet negative. I have half a mind of trying this, but I'm not too hopeful. Mostly because the halftone dots of an inkjet printer are downright abysmal in terms of dot quality and even with e.g. 5 separations (which is already a truckload of work) the result will be quite poor. So I don't have all that much hope for this working particularly well.

Another route I've explored at least in theory is using laser printed halftones. With a good laserprinter with PostScript support, it might be possible to make something somewhat decent. There are laserprinters specifically intended for this kind of work; they're used for e.g. screen printing applications. Perhaps combined with tonal separations this is a feasible route - but it does require a suitable laserprinter; not just any type will do. I haven't looked into availability in the second hand market yet, but I expect it might still be a costly experiment. Theoretically a 1200dpi laserprinter should be able to produce 21u dots, which is imagesetter territory. UV blocking power will likely be limited by comparison, although there are workarounds for this (opacity enhancing sprays).

In all the searching and reading I've done, so far my conclusion is that inkjet works fairly well for halftoning, but only for applications where a relatively coarse screen is acceptable. For instance screen printing t-shirt/textiles. Anyone who's doing photographic reproduction is either using continuous tone inkjet or imagesetter negatives. Continuous tone inkjet really is a fundamentally flawed approach for pigment processes; I've been doing quite some experimentation lately and the more I learn, the more I realize it's just a very poor solution. Might be OK-ish for B&W, but for color, it's just a mess.

koraks said:

Alright, this is a long shot. A very long one.

I've been messing around with carbon transfer lately, particularly color carbon. For black & white I like to use with camera negatives, but for color, I don't see this as a realistic option (i.e. making color separations using an enlarger). So it's going to have to be digital negatives.

So far I've done the usual thing with inkjet (quasi) continuous tone negatives, but it's an inherently flawed approach. Basically, a decent (high resolution) halftone screen approach would be far superior in several ways. But...there are limitations. For instance, I'm looking to do this myself. I'm not considering using a commercial service that I send files to and they send negatives back. I know that it's a possibility and also know a few very respectable/appropriate providers, but I also know how I explore printing processes and the turnaround time for a commercial service would make it a non-starter for me.

I've been wrecking my brain on this one, but all I can come up with is dusting off an old & beat up imagesetter scavenged from the dumpster of a print shop somewhere. And I'm frankly not sure if I want to go there. Realistically it's a €10k startup cost to get one in working order and then it's probably going to take far more religious ritual & shamanic efforts than I can muster to keep such a beast going for any useful period of time.

What are your thoughts on the matter? Any miracles up your sleeves?

Click to expand...

I've not been succesful to ever get a decent halftone negative from an inkjet(lack of Dmax) but a halftone neg from an imagesetter is a dream,athough inconvenient. For me the solution in the end was direct printing (such as White Wall) to photographic paper. The quality is high and the cost within limits. I find the turn-around time acceptable and worth the wait. The results can be sold as-is.

RalphLambrecht said:

I've not been succesful to ever get a decent halftone negative from an inkjet(lack of Dmax)

Click to expand...

Well, dmax actually isn't my worry. I'm using an Epson 3880 with InkjetMall (formerly ConeColor) K3 inks and they provide good UV blocking power. More than enough, in any case. I just did a quick test with a trial version of AccuRip to see if the halftone approach is indeed promising for color carbon:

This is a 5" x 2" 21-step without linearization applied, using matte black ink from the K3 ink set onto generic screen printing film (the cheap kind that looks a lot like Fixxons and Pictorico). This is with 16% sensitizer, 0.5ml per 4.5x6", which is in the same range I would use for B&W in-camera negatives and also in the ballpark of what people like Sandy King generally recommend. It's by far enough to give lots of relief (in fact a bit too much for color work), so in short, dmax of the negative is not a problem at all.
What is a problem, of course, is:
1: The lpi resolution of this approach. I can't get higher than 88lpi; AccuRip goes up to 92, but at that point my printer doesn't print the smallest dots anymore. 88lpi is the upper limit of what it seems to be capable of. Even with tonal separations that's lower than I'd ideally want to end up with, but as a 'proof of concept', it's a useful experience.
2: There are evident issues with dot gain. There's inkjet dot gain that blocks up the 5% patch; conceivably a higher quality film would make a difference, but I think it's just inkjet limitations. Linearization would probably solve this fairly well, although I think it could be better. Then there's some dot gain in the high density patches; this is optical dot gain in the carbon process. Part of it is halation within the carbon tissue, so I'd need to make the tissue thinner. Not a bad idea anyway; these are already about 40% thinner than my B&W tissues and for halftoning I could go much further. The light source is also not optimized for halftone dots; I think I can work my way out of that one as well.
3: There's also a dot loss in the highlights; the 5% patch has some dots, so perhaps there could have been more, but they didn't adhere. Looks like a combination of tissue thickness and processing; this was a quick & dirty job and with some more care I think I can get smaller and more dispersed dots to stick better.

Long story short, it's very promising, but a printer like the 3880 is a hard limitation that I'd need to get around, so in that sense I'm still firmly at square one

RalphLambrecht said:

For me the solution in the end was direct printing (such as White Wall) to photographic paper.

Click to expand...

Yeah, that's nice for sure. Honestly, if I wanted to do this the easy way, frankly I'd just order a $1k pigment printer with an extended gamut ink set (what's that these days, 10 or 12 colors?) and print onto decent baryta paper. I could unpack it in the morning, profile in an afternoon and be printing in the evening. If I wanted to do B&W really well in addition, I'd just convert the 3880 to all black inks and dedicate it to B&W work.
But that's not the point...I'm doing this carbon thing not because I think it's fundamentally superior or because I think it'll produce the best prints. I'm actually 100% sure that I could make far superior prints with far less effort the inkjet way. I do it this way because the carbon process fascinates me; it has a tactile quality that I really like. To be honest, halftoning already takes part of the magic away from it, but it has so many advantages that I would consider that direction. Ideally I'd be doing this continuous tone only, but the inherent problems of that approach are kind of prohibitive.

Try PS & print max resolution this screen grab 300dpi you may may need to build a curve to keep the highlight dots.
The dots are superimposed back onto the greyscale for visual. Might just be that stochastic screens are better (plug in to PS or standalone).

Check this out: https://www.facebook.com/

Thanks @ced. In a very similar vein to your suggestion I've been experimenting a bit with AccuRIP; see example above (the cyan print.) This works quite nicely in principle, but resolution is insufficient with my inkjet. Theoretically a 1440dpi inkjet will never go beyond 90 lpi anyway at 8 bit greyscale; the example above was indeed 88 lpi after which things break down. I considered laser but this is also practically limited to 1200dpi = 75lpi max for halftones. I guesstimate that anything beyond 200lpi would look decent also in a smaller print (4x5" to 8x10") unless looked at from too closely... That's beyond desktop printer territory as far as I can tell. I've seen mention of inkjets being stretched to 125lpi with debatable dot quality; I'm skeptical of even this because a printer like e.g. the SC-P700 doesn't go beyond 1440dpi in the longitudinal axis.

Btw I don't have Photoshop but I understand GIMP also has a halftone screen filter. I might experiment with it after the AccuRIP trial runs out, but as outlined above, hardware limitations are prohibitive of a real solution in this direction at this point.

Looks like the inkjet & laser options are great for screen printing on t-shirts and whatnot, but I don't see anyone choose this direction for more demanding applications.

koraks said:

Thanks @ced. In a very similar vein to your suggestion I've been experimenting a bit with AccuRIP; see example above (the cyan print.) This works quite nicely in principle, but resolution is insufficient with my inkjet. Theoretically a 1440dpi inkjet will never go beyond 90 lpi anyway at 8 bit greyscale; the example above was indeed 88 lpi after which things break down. I considered laser but this is also practically limited to 1200dpi = 75lpi max for halftones. I guesstimate that anything beyond 200lpi would look decent also in a smaller print (4x5" to 8x10") unless looked at from too closely... That's beyond desktop printer territory as far as I can tell. I've seen mention of inkjets being stretched to 125lpi with debatable dot quality; I'm skeptical of even this because a printer like e.g. the SC-P700 doesn't go beyond 1440dpi in the longitudinal axis.

Btw I don't have Photoshop but I understand GIMP also has a halftone screen filter. I might experiment with it after the AccuRIP trial runs out, but as outlined above, hardware limitations are prohibitive of a real solution in this direction at this point.

Looks like the inkjet & laser options are great for screen printing on t-shirts and whatnot, but I don't see anyone choose this direction for more demanding applications.

Click to expand...

In the end you're the boss ;-)