I complained earlier about digital negatives. They’re my least favorite part of the color carbon process so far, and I have a feeling it’s going to stay that way. Truth be told – I still think digital inkjet negatives suck a$&. Seriously. They’re work of the devil. Not that I haven’t made any progress on this front. To the contrary, apart from being away from home for a few days, I actually did a truckload of work involving digital negatives and they have improved. Read on for more info on the crucial improvement(s) that I made.
Before we start, I need to tell you that I just wasted about 3 US-Letter sized sheets of Fixxons transparency material to print ONE set of 4×5″ color separations. So I’m not in the best of moods, let’s put it that way. Moreover, I notice that the color carbon project currently revolves a lot around spending time behind a computer monitor and coaxing an inkjet printer into doing something it wasn’t made for (i.e., printing digital negatives). That’s not what appealed to me in darkroom photography, so it’s something I’ll have to periodically evaluate and see if I want to go down this road. I’m tech-savvy enough, but the appeal of photography for me is that it’s essentially a non-digital process. Computers are great for other things and can be fun or interesting, too. But for me, they’re not a pleasant (albeit a very effective) tool for photography.
Software – f^ck you Adobe, grumpy GIMP and indolent IrfanView
So what have I been running into on the digital negatives front? Almost too much to mention, but let me give it a go. Starting on the software end, I’ve been trying to put together a workflow that allows me to make color separations and print them as negatives on my old Epson 3880 printer.
Part of the challenge is to work around Adobe and to use shareware or freeware (or open source) software to do this. Why? Well, Adobe, if you’re reading this: your licensing schemes are downright RIDICULOUS for amateur photographers like me. For what I do, I basically need Photoshop, and that is currently a $20/month plan, or $240 on an annual basis. I’m sorry, but I can’t justify this given the fact I also have operational costs in chemistry, paper, ink, other tools, etc. etc. The photo editing suite is just one building block in the whole chain, and while Photoshop is certainly a very high-quality product, I’m just not going to throw close to €250 at it per year and not even own the g*ddamn product! Unless you revise this practice of pure idiocy, I’ll categorically refuse using Photoshop. Period.
So I’m in the hands of GIMP, and while it can be an annoyance in itself, it has come a long way. Very importantly, it can do CMYK separations out of the box – it’s actually a plugin that packages with the default distribution, and this plugin in previous versions didn’t even work properly, but it does now. So at the press of a button, I get my cyan, magenta, yellow and key/black layers in a neat little document at the required color resolution. Neat!
There might be a bit of an issue with these separations, though, because I have some serious doubts about how the black/key layer works. In my current color checker torture test, I have lots of color (duh!) and also a couple of pure greyscales. It occurs to me that the greyscales split out into just the key/black layer (this is good), and the color patches also involve a bit of black (also good in principle, evidently). What is not so good is that the black/key component of the color patches seems to be very dense (in the final print) compared to the pure greyscale component. The outcome of this is that if I have a black tissue for the key layer, it can either render a full greyscale ranging from black to white, but also yielding very dark color patches, or I can have color patches with decent overall density, but the greyscale comes out washed out and ranging only from white to a light middle grey. This is not something I can fix with adjustment curves, so I might need to revise the whole color separation process and do it manually – at least for the key layer. Not really looking forward to that, as you can imagine.
One thing GIMP still does not do, is print decently. Any paper size that’s not within its very restricted default set is just not going to materialize, so you’re on your own there. I rooted around a bit and ended up installing IrfanView, which actually has a decent print engine. It’s not without its qualms, though. For one thing, I initially tried saving the separation layers as .PNG files since they are in principle not lossy and allow 16 bit color depth while still being more efficient storage-wise than TIFF. Well, in theory that’s nice, but IrfanView actually does not understand .PNG the way it’s meant to be, and simply downscales 16 bit .PNG’s to 8 bit. Useless for color separations, so we’re back at good old harddisk-consuming TIFF. Ah well, could be worse.
The opaque world of transparencies
Then there’s the negatives themselves. The essence of a negative is that it’s semi-opaque: it blocks light where you want white, and passes light where you want tone in the final image. This needs to be continuous, at least in the way I’m currently working now. An alternative to continuous tone would be a dot screen approach, and I’ve done some preliminary experiments with that, but so far I have not yet migrated to it. Mind you, I just might. More on this later.
One of the essential parts of the negative is, well, the negative itself – the material it consists of. Evidently, for the opaque bits I’m using inkjet ink, but it needs to sit on something. Enter the different kinds of transparency film. Over the past years, whenever I needed (or thought I did…) continuous tone transparencies, I just grabbed some sheets of Esselte brand inkjet transparencies (sorry, link only in Dutch) and get cracking. This worked OK-ish for things like cyanotype and I even got away with it when doing photopolymer intaglio. However, for carbon transfer and just about any other alternative printing process, these sheets are just garbage. Seriously, don’t waste money on this. While cheap, they really have nothing to offer and in fact, I wonder why this product even exists in the first place. I’m not joking! I really think Esselte would be far better off just relabeling one of the alternatives and selling it under their own brand.
By the way, not all of the alternatives are much better; over the years I tried several similar inkjet transparency sheets, and they were all horrible. Just like the Esselte ones, they simply don’t hold much ink, and even at very low ink loads, the ink actually reticulates as it dries. Well, once it dries, eventually, because this takes literally ages and the sheets remain tacky even when ‘dry’. A hairdryer helps some, but doesn’t help the cracking/reticulation issue. Not that it matters, because there appears to be no way to prevent this reticulation problem in the first place.
This makes it virtually impossible to create negatives with good blocking power, since there’s always a truckload of light literally slipping through the cracks. One workaround I did employ for a brief while was to print two sets of transparencies and then carefully align them on top of each other. This actually works sort of OK-ish, but of course there is degradation of fine detail in the prints because one of the transparencies just isn’t in contact with the carbon tissue (or whatever other print material is being used). I’m sure I’ll find a good use for these sheets, but making digital negatives is certainly not one of them.
Recently I was pointed on Photrio by insanely helpful member Andrew O’Neill to screen printing transparencies, which he argued will hold a lot of ink. Sounds good, so I picked up a pack of A3-sized generic screen printing sheets and gave it a go. You know, if I were Esselte, I would do the same, stick a big Esselte brand label on the box and sell them instead of the pitiful product they carry now! This material really is far better than the other stuff and it sure will hold a lot of ink. In fact, in the Epson print dialog, there’s a setting that allows you to increase the ink density up to 50%, and these generic screen printing sheets will happily hold even this ridiculous amount of ink and allow it to dry (quickly!) to a fairly even surface. And this stuff is quite cheap as well! It’s less than €0.50 per A3-sized sheet in a pack of 50 sheets, if memory serves. Nice!
Is it ideal? Well…no. Sadly not. Firstly, there’s an odd halation issue that creates density differences in areas of high contrast. Let me explain. Print a light grey patch and a black character (e.g. a number) inside it. For some reason, the ink dries up darker along the edges of the pitch black character than in the rest of the patch, resulting in a lighter halo around the text on the final print. Why, or how, this happens? Beats me, but it’s a kind of inkjet-based edge effect that is absolutely undesirable. Now, it doesn’t seem to happen consistently, so I imagine I can work around it, so it’s not a deal-breaker – yet.
Then there’s another ‘minor’ issue. While these transparencies dry within a few minutes, they do prove to be incompatible with sensitized carbon tissue. I imagine this has to do with trace amounts of water remaining in the sensitized tissue, even though I quite thoroughly dry these tissues currently. The effect is that the entire ink layer of the transparency mates with the carbon tissue and peels off the transparency material base. It’s, uhm, kind of ugly. Here, look:
Now, the problem can be prevented easily and effectively by simply placing a thin sheet of siliconated mylar/boPET between the negative and the carbon tissue. I’ve been doing this for most of the continuous tone tests I’ve done over the past week, and it works OK in this scenario. However, it does not work at all for halftone screen negatives. Even though the mylar is thin (around 60um if I’m not mistaken), and even though I’ve been using my semi-collimated LED lights for these test exposures, the intermediate sheet still degrades the definition of the image more than sufficiently to turn the fairly sharp inkjet dots into a layer of soft mush. So if I want to try halftone inkjet screens some more, and I think I’ll end up doing this at some point, the screen printing transparencies aren’t the way to go. Too bad, because they’re really nice overall.
Earlier I also mentioned once that I ordered some Fixxons 8.5×11″ negative film. Economically it’s not nearly as attractive as the screen printing material, because it needs to be imported from the US and that makes it all a bit more expensive than it appears at first glance – and the material is a bit more costly to begin with. But still not as pricey as the competing and allegedly very good Pictorico OHP stuff. This we can get quite easily in mainland Europe, but I consider it a last resort as it’s a bit costly for the kind of experimentation I’m doing with it. I do print through rather impressive stacks of sheets quite quickly, I find, and I don’t like to worry about material costs when doing so.
Well, turns out the Fixxons looks suspiciously like the screen printing film. Actually, when I unpacked it, I thought it was essentially the same stuff, only thinner. And I do think that conceptually it is sort of the same – it’s probably a clear polyester base with a thin film adhered to it that takes the ink. While the Esselte (&co) use something that feels and behaves like a gelatin coating as the ink recipient, the screen printing and Fixxons films use what looks like a microporous polymer film with undoubtedly some nifty additives to make it take ink and easily release moisture. I really don’t know what’s on there, but it works quite well. Apart from the peeling issue that the screen printing film suffers from, and interestingly, this problem does not appear to affect the Fixxons! That’s excellent news, because an initial test suggests that I could safely print with the Fixxons negatives in direct contact with the carbon tissue. Oh, it also holds the same amount of ink.
Is it perfect, then? Nope, no cigar. Three problems and one concern. Starting with the latter: I’ve also seen similar uneveness in the drying negatives as I observed on the screen printing film, and even some additional blotchiness. I hope this remains invisible in the actual prints, but I’m concerned about this. Then there are some actual problems. Firstly, the Epson pizza wheel marks show up very clearly on the Fixxons film. Every high-density area is adorned with longitudinal tracks of tiny dots, and these do show up on the final prints. After about 20 years of engineering, Epson apparently still hasn’t removed this major problem from their latest printers, so upgrading my 3880 likely won’t fix this issue. If I stick with digital negatives, I’ll have to use one of the several workarounds that have been developed to circumvent or even entirely remove these pizza wheels from the transport chain.
Then there’s an issue with the transparency bit – while the Esselte and the screen printing films are pretty much transparent to UV, it seems the Fixxons actually blocks it a bit more. So exposures are a little longer on this film, by about 10-15% as an initial estimate. It’s not a major issue and I can live with it, but it’s a pity still.
Finally, because the Fixxon’s is a relatively thin film, although not as flimsy as the Esselte crap, and it’s a laminate, it tends to curve inward a bit. This results in head strikes on the edges of the film – yeah, not only do those ugly pizza wheels leave nasty marks, they aren’t effective to boot! In initial testing I’ve seen that I need to reprint around 20% of the small sheets due to alignment issues because the print head simply knocks the entire sheet slightly out of alignment, resulting in prints showing double images and basically being rejects. I don’t like waste, let alone for stupid reasons like this one. Ugh.
Conclusion – no conclusion…
So, where does this leave me, currently? Well, I have one type of transparency film that sort of works, and that will probably be good enough to proceed with. The Fixxons pretty much (though not entirely) does what I hoped it would, so that’s a bonus. I mean, come on, I can actually print negatives with sufficient blocking power in a single pass, and that’s something that was simply impossible before. It’s actually a major step forward!
And in terms of the software chain, it seems that the combination of GIMP and IrfanView is quite effective in making separations, making adjustment curves and committing all this to
paper plastic. No, it’s not the elegant effectiveness of Photoshop, but it also doesn’t come with an unjustifiable ripoff licensing scheme.
In conclusion, despite a good dash of discomfort, annoyance and frustration along the way, I’m in a position to do some testing. Which is to say, I’m knee-deep in the whole curve adjustment/linearization mess, which is a story all unto itself, and I haven’t even touched the color profiling or soft-proofing issue yet. It sure is a very, very deep hole.