DAS right! – Solving the teething problems of DAS carbon transfer

Previously I wrote about trouble I was running into with DAS-incorporated carbon tissue. I just couldn’t get those tissues to transfer properly. Well, looks like I figured out why. In the process, I received some useful advice to boot, and perhaps there are some myths being established as well. Are you starting out with DAS and running into issues? Maybe some of the info in this blog can be of help.

Issues with the tissues! More specifically, an odd problem with highlights that won’t transfer, and millions of tiny bubbles. The bubbles emerge upon soaking an exposed DAS-incorporated tissue and mostly/only pop up in areas that have received UV exposure. They are really tiny; a fraction of a millimeter, and they pop up by the millions. The transfer issue only appears later, once the actual transfer has been made to the final or temporary support, and the print is being ‘developed’ in the warm water bath. As the unhardened gelatin washes away, at some point the highlights also start to lift off the support, frill, and float away. The problem gets progressively worse into higher densities. So, not good.

What got me into this position? In other words: materials, workflow, etc.? Here’s the glop recipe I started out with, inspired by King, Nelson & Lockhart’s “Carbon Transfer Printing” book (recipe #2, page 135):

Ingredient	grams (in 100ml glop volume):	% of weight gelatin:
Gelatin	10.00	100.0%
Kremer Pbk7 pigment	0.10	1.0%
Sugar	3.00	30.0%
Glycerin 86% solution	0.25	2.5%
DAS	0.40	4.0%

Note I included the ratio of each ingredient to the weight of the gelatin, since this is especially indicative of the difference performance aspects of the dried tissue.

The pigment was added to the glop as a 10% dispersion (home-made). Once the glop was bubble free, the DAS was added in a separate solution with water, and the volume of the glop made up to 100ml. I’ve since received a suggestion by expert printer Kees Brandenburg that weak solutions of DAS (around 3% w/v) keep well (away from UV light), which makes for a convenient way of dosing it when making tissues. I now keep a brown bottle of pre-made 2.5% w/v DAS at hand. From the point the DAS is added to the glop, the glop is sensitive to UV light and pouring tissues and drying them must happen under safelight conditions or in the dark.

I exposed the first batch of tissues in various ways – using step wedges and a number of negatives, using the 400nm LED UV light source I also use for dichromate-sensitized tissues. Further processing was also along the same lines as I usually do with dichromate, with a number of variations to try and solve the adhesion tissue. Let’s have a look at these.

Cold water soak

After exposure, the tissue is soaked in cold water. Advice on how long this soak needs to be is pretty much all over the place. Sandy King mostly recommends a 2 to 3 minute soak, but I’ve also read longer and (mostly) shorter times. More important than the actual durations being mentioned, are the associated effects I’ve come across, either in my own experimentation or through hearsay:

Too short a transfer may result in adhesion problems. I’ve personally never been able to verify this, and some practitioners seem to ‘get away’ with very short soak times. E.g. Kees Brandenburg told me that long soak times often aren’t necessary (with a little caveat) and that significantly shorter than 1 minute actually works, and I’ve seen similar approaches by e.g. Tod Gangler. If you watch the first few minutes of this particular video, you’ll see that the total soak time the black tissue gets is only around 90 seconds or so, which partly seems so ‘long’ because Tod explains what he’s doing.

The caveat with short soak times is that when working with transfer media that are heavily curled, they need some time to soak up water and ‘relax’ so that the transfer can successfully made. But this is a soak of the support the transfer is done to, and not of the tissue itself. Indeed, in my experience, dichromate tissues transfer just fine to an albumen-sized polymer sheet after a soak of only a couple of seconds – just enough to get the surface wet. In fact, with some tissues, particularly with PR122 pigment, I found that longer soak times (> 1 minute) can create problems when transferring to Yupo. The image would not adhere well and slide off entirely during warm water development. This only occurred with specific pigment loads, however. In general, a soak time of one or two minutes works just fine.

There is one aspect that makes me prefer a soak of around 2 minutes: tear marks. These are striations that run in the direction the tissue and support are pulled apart in, and show up as parallel, wiggly lines of added density. They mostly wash out during warm water development, but not always entirely. I found that this effect tends to show up in tissues that have been soaked very briefly.

Certainly, I tried soaking longer or shorter with the DAS tissues, but noted no difference in transfer success – or failure. With dichromate tissues, the pre-transfer soak has not been as critical as I assumed based on what people have written about it. There seems to be some leeway here, to put it mildly.

Waiting period between mating and development

Many sources indicate that there needs to be a significant wait time between mating the tissue and the support, and warm water development. I recall mentions varying from 10 minutes to 30 minutes minimum, with several hours still being OK as long as the sandwich is prevented from drying out from the edges inward.

I have never done rigorous testing with this, but especially when transferring to non-absorbent materials (Yupo, polymer sheets such as acetate or PE), this wait time doesn’t seem to be relevant. Usually I mate the tissue & support, squeegee them together and hit the water cooker immediately after it to boil some water for the development bath. The sandwich typically hits the warm water a minute or perhaps two after mating – the time it takes me to prepare the warm water bath.

Perhaps with gelatin-sized art papers, there is a minimum. I usually err to the side of caution and let the sandwich rest for maybe 10 minutes, but generally not longer. I’ve not yet observed a transfer failure that I could reliably attribute to too brief a resting period. I’m at a point where I’m inclined to conclude, after thousands of prints, that it’s probably not a hugely important factor overall.

Yes, I did observe this (in my opinion, of questionable importance) resting period in my DAS experiments, because who knows, right? But it didn’t seem to matter much.

Support choices – what to transfer to

Despite I hadn’t read anything about DAS-tissues being particularly difficult to transfer to some materials, I figured that maybe this would make a difference. I have done most of my testing so far with Yupo because it requires no preparation to transfer to, in general. So I tried a number of alternatives, all of which work great with dichromate tissue:

• Gelatin-sized art paper. I size with gelatin with some chrome alum added to it, as this is ready to go as soon as the paper is fully dry. This is in contrast to e.g. formalin hardening, which needs some time to complete. And it has a poor reputation due to the vapors it lets off, too. I’ve done many single transfers over the past months (and during earlier carbon episodes before that, too), and my sizing procedure for Schut Laurier paper seems to be quite reliable. But with DAS, it didn’t do any better than Yupo.
• How about acrylic-sized paper? I actually had some bad luck with this before, but at some point I read that this also needs some time to cure. I trialed it with dichromate tissue and found it actually worked quite well, so I gave it a try with DAS as well. No luck; if anything, it was even worse than Yupo.
• The one thing DAS tissues really need to transfer to, in my workflow, is albumen-sized transparencies. Again, with dichromate tissues, these work great. I size laserprinter transparencies with a 2% albumen solution with a little chrome alum added to it – again, this makes it ready to go as soon as the sheets are dry. Virtually 100% success rate with dichromate…but not with DAS.
• Then a person called Eric D. on the carbon transfer group suggested glass. Well, I had transferred dichromate tissue to gelatin-sized glass before, so I know that works. I never tried unsized glass, but gave it a try exactly as Eric describes in the linked post. The effect was the same as with the other supports.

The similarity between the different supports was actually striking – here, have a look:

The image is a composition of clouds, and the higher tones (where the sun strikes the clouds) are all washed away. The extent of the problem seems to vary a bit, but this is mostly due to the moment I decided to stop development and give up on a particular print. I was a bit more patient with some than with others.

Warm water development bath

There’s not all that much to say about this. Well, one or two things, perhaps.

Temperature is of course critical, to some extent. Too cool a development bath and it’s impossible to separate the tissue and the support. I’ve had this on occasion, where I really applied considerable force to try and separate the two. Interestingly, I never did any damage to the actual print, and just adding some hot water and trying again after a minute made things easier – with no damage done. I probably lucked out there, but the lesson learned, is that it’s easy enough to spot if the bath is too cold. It doesn’t take a thermometer.

On the hot side, there are bigger risks, I think. Gelatin gets softer as it gets hotter, and that’s true for hardened gelatin just as well. Besides, trapped air (bubbles) expands more the hotter it is, so I can imagine how bubbles will be a bigger problem if the transfer bath is too hot. However…my conclusion so far is that a transfer bath that’s on the hot side may exacerbate problems that are already there (e.g. bubbles, poor hardening of gelatin sizing) and not so much present a problem by itself. Still, I aim for a temperature between 42C and 47C or so.

And I did just that with my DAS tissues, whipping out the thermometer to verify I was on track, and…this wasn’t the problem either.

Is my DAS bad?

Alright, so something weird going on, for sure. Let’s have a look at the actual DAS. I know it works, at least to an extent. After all, it produces a print-out image – quite strongly so. Here’s an example from one of my firsts tests:

Moreover, some hardening was definitely going on, as evidenced by the illustration of the failures shown above. However, I did start to suspect the DAS based on its color. Various sources describe the proper color of dry DAS as an off-white to yellowish powder. Mine isn’t exactly that – I took a small bit and photographed it under warm-white light:

But is it bad enough to be ‘bunk’? Well, out of curiosity, I exposed the powder above for a few minutes under my exposure unit. Afterwards, it looked like this:

This suggests that the ‘fresh’ DAS I have probably has degraded somewhat, but that there’s still significant more room for additional exposure. And again, I did see hardening action, so could it be that the DAS is still active, but not entirely so? It’s something I couldn’t quite explain chemically (but I’m no chemist) – so I had to try it experimentally.

So I increased the DAS content of the glop – initially increasing it to 0.6% of the dry gelatin weight, and ultimately 1%. You guessed it – no difference! Well, there’s a difference for sure, since increasing the DAS content seems to reduce contrast due to some very heavy self-masking behavior. I might come back to this in a later blog, but I’m for sure not the first one to notice this.

Shedding light onto the problem

In a belated moment of clarity, it dawned upon me that there’s one factor I had been overlooking in my experiment: the excellent value-for-money UV LED light source I had been using. Could it be…? Well, I still had my old bank of Philips BL tubes read to fire up, so that’s what I did. Guess what? A perfect transfer on the first try.

So what’s going on here? I’m not entirely sure, I must admit. Obviously, there’s the matter of wavelength. Let’s see what a LED puts out, going by a fairly randomly selected spectral distribution chart of Lumileds Luxeon UV FC LEDs. Note that these are not the ones used in the Chinese floodlights I’ve been using, but emission spectra of LEDs are pretty similar overall.

The red lines is of interest here, since this will correspond closely to the LEDs I’m using. EVidently there’s a peak around 405nm, with a rapid and symmetrical falloff on both sides. Power output is nearly zero at around 375nm and 430nm.

Now let’s look at the Philips Actinic BL tubes I was using before. This is their spectrum according to the current datasheet of that product line, which again will be similar or identical to the actual product I’ve got right here:

The difference is clear, and we can in fact make it a little clearer still. Hold on, let’s overlay those spectra for more clarity:

That’s pretty different alright, but how does it relate to the sensitivity of DAS? Well, Sandy King back in 2019 uploaded the following absorption spectrum for a DAS solution of unknown concentration:

The caveat in this image is that it’s not clear what the tapering bit on the right-hand side at around 400nm means exactly. Is the absorption here near-zero, but not quite, or is it already a firm zero? Also, the end of the plot is around 420nm, but does that mean there’s no absorption beyond that wavelength, or was this not part of the measurement? I don’t know, but I’m going to take a stab at it and assume there is some marginal absorption beyond 400nm, but not enough to work with practically.

If we overlay the DAS absorption chart on the spectral curves of both light sources, we get this:

This shows that the emission of a typical 405nm LED is a nearly completely disjunct from the absorption spectrum of DAS. In other words: I’ve got lots of UV LED light, but the DAS is effectively blind to it.

Well, not quite, because I do get quite a bit of gelatin hardening with it. If you look at the failed transfers, there’s actually pretty decent hardening action occurring – perhaps it’s not exactly efficient in terms of how many Watts of electrical power I’m throwing at the tissue, but the hardening does happen. What’s still a mystery to me, is why the highlights in particular wouldn’t harden sufficiently to form a transferrable relief, while they do harden OK with the somewhat deeper UV light that the BL tubes put out. Wavelength and thus photon energy undoubtedly has something to do with this, but how exactly, I don’t know.

Trying to exploit the fact that the LED light source also gives some action, I tried combining exposures with both light sources. I’ve done some informal experiments, and an exposure of 2-3 minutes followed by on of 8-10 minutes under the LEDs gives an image where the highlights still transfer OK. What appears to happen is that the brief tube exposure serves to harden the highlights, but most of the density in the shadows and midtones is built using the LED light.

I also have the impression that the 405nm LED source gives a higher contrast image (at the cost of highlights that fail to transfer properly) than the Actinic BL tubes. I still need to do more systematic testing to confirm this. This could, however, allow for some contrast control by combining exposures of different wavelength. Yes, I’m aware that e.g. Calvin Grier uses a combination of 365nm and 400nm LEDs in his custom-designed light source, and that he uses the 365nm LEDs for highlight exposures and the 400nm ones for shadows. Looks like I ‘discovered’ why!

DAS concentration, contrast and printing speed

In terms of contrast, what works with dichromate, also works (at least to an extent) with DAS: a higher sensitizer concentration reduces contrast. But other than with dichromate, I don’t observe much gain in printing speed as the DAS concentration increases. For instance, the following two prints were done with a 1% DAS to gelatin ratio and a 0.4% DAS to gelatin ratio, respectively. Otherwise, the glop formulations were identical, and exposure was also in the same ballpark.

It seems that there’s a lot of self-masking going on that reduces the contrast, and subjectively I would say that it’s a far stronger effect than I see with dichromate sensitized tissues. I have yet to test a seriously ‘lean’ tissue with 0nly 0.2% DAS to see what the limits of the envelope are. They’re presently drying, so I’ll find out pretty soon. I’m also planning on boosting the contrast by increasing the pigment load on these tissues. Given the strong self-masking behavior, it seems that a thinner tissue with less gelatin and more pigment would just work better with DAS, although highlight rendition is a bit of a concern.

And I’ll probably also need to revisit the whole UV exposure unit business. I’m eyeballing some UV LEDs and am about to place an order for some stuff to play with. The Actinic BL tubes work, but I’m still not happy with the diffuse nature of the light. I’d also like to bring down the exposure times. With the BL tubes, I’m presently in the 10-20 minute range, and that’s way beyond what I find comfortable. I’d like to push that down by a stop or two. LED seems just a better option overall, so let’s see if we can get that show on the road on a shoestring budget as well.