Aching, flaking – A persistent problem with carbon DAS highlights

The problem that just won’t go away: disappearing highlights on DAS carbon prints. Twice so far I’ve declared this problem solved on my blog, only to have to revoke my ‘eureka’ shortly afterwards. Over the past few weeks, I’ve been systematically testing the process to find a cure. So far, to very little avail, I’m afraid. But perhaps that this writeup will (1) help me re-think the problem and (2) it’ll give some hints to anyone else struggling with DAS carbon problems by listing factors that can/should be evaluated.

The nature of the problem

What does the problem look like, again? Take these prints from part of an 8×10 in-camera negative (Fomapan 100):

DAS print to the left, dichromate to the right

To the right is a dichromate-sensitized print, spirit sensitized with 0.5ml of 16% ammonium dichromate on a 5×7″ approx. tissue, applied with a foam roller. The print is made to favor the highlights, since that’s where the problems with my DAS prints generally appear the first. Note that the dichromate print renders the highlights just fine, also seen in a closeup:

Section of dichromate-sensitized print. Note delicate highlight especially in the two diagonals and the single vertical 2/3 from the left of the image. There’s tone here, maybe zone VIII or IX-ish.

With DAS, the problem looks something like the left print of the first illustration above – in close-up, this gives a more detailed impression:

DAS carbon print with severe problems. Lighter greys are partially gone, delicate highlights missing at all, and signs of flagging.

These prints are made with a very high pigment load, so an inherently very contrasty tissue. This deliberately exacerbates any highlight problems, since rendering delicate highlights with a highly pigmented tissue will involve super thin layers of gelatin that need to withstand warm water development and adhere reliably to the support material. I’ve been testing this way lately, and mostly with this particular negative, because it quite consistently replicates the problems.

Problem severity and variations

The severity of the problem varies a lot, depending on many factors. Here’s a quite randomly picked example of the range of the problems, taken from a stack of test prints on my desk:

Typical variation in the extent of the problem

The severity ranges from only minor problems with the lightest highlights breaking down into paper white, to major parts of the image missing entirely. The problem always emerges during warm water development. It manifests itself as pieces (tiny, or larger) drifting off of the support as development progresses. The problem tends to get progressively worse as development proceeds. Often it’s visible from the moment the tissue support is peeled away, sometimes the problem only becomes apparent halfway through development.

The one thing that is very clear to me, is that it’s an adhesion problem. The highlights do seem to harden just fine during exposure, and initially develop fine as well. They just don’t stick reliably to the support.

Not negative-related

Several people have suggested that the problem traces to the negative: trying to print a negative with too long a tonal scale (too much density in the highlights), resulting in loss of highlights. However, the problem here has nothing to do with a negative of any kind. This is demonstrated by replication of the problem without any negative involved at all:

DAS carbon gradient, made without negative

The above image shows a dark grey vertical control band, followed by a gradient that runs from paper white to middle grey. Well, it’s supposed to be. The gradient was made by sliding a piece of cardboard over the tissue as it was being exposed, in order to gradually uncover a growing band of the tissue to the light source. The problem is clearly visible here as well, especially if we zoom in a little:

Close-up of gradient shown above

The transition of visible tone into paper white isn’t gradual, but quite abrupt. It looks a little gradual, but this is only because at the interface between no visible tone and the first visible grey, the highlights break down somewhat erratically over a narrow zone.

In Calvin Grier’s terms, what is shown here is a rather severe ‘tonal threshold’: an abrupt transition of no tone to the first visible grey. Consistent with the earlier example shown above, the first visible tone rendered is a light, middle grey. Anything more subtle / less dense is simply missing from the print. Since the problem can be demonstrated without the use of any negative or step wedge tablet, it must be related to the printing process itself.

Contributing factors

So what’s causing this? Frankly, I don’t have a clue. All I know is that the problem occurs in my DAS tissues, but tissues of otherwise identical formulation and processed the same way (save for sensitizing) with dichromate work just fine as they always have. So it’s somehow DAS-related, but how?

That the severity of the problem varies, might trigger the response “so you know that certain factors make a difference”. Indeed, I do. But put simply: I know several things that make the problem worse. I have not found anything that makes it better. There’s sort of a baseline severity of the problem that corresponds to a ‘best practice’ – if I do everything as well as I can, with the most conservative/safest approach along every step of the way, the problem remains limited to missing highlights. By varying process parameters, I can make the problem worse, and introduce new ones. But I have not found anything that takes the problem away. Of the many hundreds of tests I’ve done, not a single one (on critical inspection) was free of the problem. It’s highly consistent in this sense.

Reporting on all the (hundreds!) of experiments I’ve done in the same way as I did for the role of the negative above would probably result in not a blog entry, but a modestly sized book. So instead, I’ll list the different factors involved in making a DAS carbon print that I’ve tested/varied to find a cause for this DAS carbon problem. To make the list manageable, I’ve divided it into sections in accordance with the printing process.

Tissue materials

FactorNotes / tests
Gelatin4-5 different gelatins tried, both porcine type A and bovine type B, different bloom strengths.
SugarDifferent amounts of sugar (including none at all), and sugar replaced with honey
DASDAS of two different sources, both sources being used successfully by other practitioners. DAS added either as a pre-made 2.5% stock concentrate in water (filtered solution to remove undissolved sediment), or weighed powder immediately before use and then dissolved in small amount of warm water and added to glop (unfiltered). Tried different DAS:gelatin ratios ranging from 3% to 5%, but mostly kept stable at 4%.
PigmentDifferent dry pigments (mostly Kremer Pbk7 and Pb15:3) as well as Talens India Ink (recommended ink by Calvin Grier for carbon transfer). Also different dispersion agents for dry pigments, including gum, honey, Orotan 731K and gelatin.

Tested mostly with pigment concentration of 0.25g dry Pbk7 per 10g gelatin (2.5% w/w gelatin) and between 1ml and 2ml Talens India ink per 10g gelatin (10-20% v/w).
WaterOnly tried tap water so far for glop
SubstrateMostly tried Yupo, either virgin but also reused, and albumen-, and gelatin-subbed polyethylene
pH modifiersTried small amounts of sodium hydroxide and hydrochloric acid to raise resp. lower pH of the glop. The sodium hydroxide seems to react with the DAS, creating large bubbles and an ammonia smell. Oddly, these tissues still appeared to work. The hydrochloride tissues didn’t work well at all.
ThicknessTissue thickness can either be controlled by adjusting the gelatin (and all other dry components) concentration of the glop, and (to an extent) by how much glop volume is coated per surface area. I mostly work with a 10% gelatin content in the glop, which is a fairly thick tissue, but have also tested thinner tissues by coating less volume.

Support materials

FactorNotes / tests
Support materialTried Yupo and various transparency materials, mostly polyester laserjet transparencies. Also tried ‘no support material’ by pouring tissue onto transparent tissue support (gelatin-subbed polyester sheet) and exposed through the back, developed on the substrate. Also tried gelatin- and acrylic-sized paper.
Pre-treatmentThe Yupo I have used straight out of the pack, but also with pre-treatments involving washing with sodium hydroxide and/or acid solutions, scrubbing with calcium carbonate + ethanol (like cleaning glass plates for wet plate collodion), sanding the yupo with fine-grit sandpaper, sizing with acrylic medium, and subbing with hardened albumen and with hardened gelatin.

The polyester sheets I have tried with hardened albumen and gelatin subbings, and also sanded with fine-grit sandpaper.

Tests with paper used paper sized with 25g/m2 gelatin hardened with chrome alum.

Glop & tissue manufacturing

FactorNotes / tests
TemperatureTried glop melt and off-gassing temperatures of 37.5C (sous-vide) up to nearly 50C (soldering hot plate set to 51C).
Bloom periodThe time allowed to let the gelatin bloom. Never controlled this precisely, but ranging from just a few minutes (<5) up to maybe half an hour. Very short bloom periods result in problems with unmelted gelatin particles.
Off-gas periodThe time allowed for bubbles to dissipate from the glop; I’ve gone as short as about an hour with still some bubbles remaining, up to many hours (~8?)
Tissue pour heightTest tissues generally poured to a wet height of 0.83 to 0.94 mm (15ml to 17 ml of glop per 12x15cm tissue). Tried thinner tissues down to around 0.5mm average wet height.
Tissue dry timeUntil thoroughly dry to the touch, but varied between just dry (tissue still flat and pliable even at low sugar content) to very dry (curly tissue).
LightingProduced tissue under regular white LED room light as well as red/orange only light. This factor extends also to lighting during other tissue handling activities (pre/post exposure, development).


FactorNotes / tests
Tissue ageNot controlled, but tried fresh tissues (<24 hours old) up to probably 2 months.
Tissue pre-treatmentTried wiping the tissue surface with ethanol or acetone. Rationale being the possibility of some kind of greasy film being present on the tissue, creating transfer problems. No effect; note also that exposure through the back of a transparent tissue support and development in situ circumvents any tissue-surface anomalies that may be present.
Light sourceTried Philips UV-BL tubes (peak wavelength ca. 360nm, spectrum ca. 350nm – 410nm emphasis on shorter wavelengths), 400nm LEDs, 365LEDs, combined 365 + 400nm LEDs, sunlight (direct and cloudy day). The 400nm LEDs and cloudy daylight exposures give problems with vastly insufficient hardening in highlights and upper midtones.

Direct sunlight, UV-BL tubes and 365nm and 365+400nm combined LEDs give similar results (also contrast!) when compensated for strength / exposure time. Direct sunlight had a small edge over LED, but this is probably due to better collimation than wavelength distribution etc. The better collimation makes tiny inter-grain image areas image more discretely, allowing them to stick a little better. The same highlight problems persist, however.
Materials in light pathI generally use a mask of rubylith to mask the edges, leaving the transparent part of the rubylith in place in the exposed center area. I tried removing this transparent material, increasing exposure speed by about 2/3 of a stop, but no improvement with the actual problem.
No material between negative and tissue. Regular 2mm picture frame glass in the contact frame, uncoated. Tried hardened 5mm glass on occasion as well (different contact frame).
Post-exposure treatmentTried letting exposed tissue sit unprocessed for a few minutes (maybe 10-15).
Tried heating the exposed tissue with a hairdryer on high heat setting; this brings out the latent DAS image and makes the tissue very supple as long as it’s hot, but no effect on actual image transfer.
Lighting pre- and post exposureTested with white LED light, daylight with pulled blinds or even broad daylight on overcast days, and red/orange LED light. No differences.

Note that tissue handling under (partial) daylight conditions should produce some visible fog at least, and certainly does so with dichromate tissues. But so far I have never ‘managed’ to print fog with DAS tissues, which underlines the problem.

Soak and mate

FactorNotes / tests
Tissue soak timeTested from a few seconds to a few minutes. Very long times (>3 minutes) result in increasing transfer problems, likely due to excess water absorption. Very short times result in many DAS-nitrogen bubbles and pronounced tear-marks on the prints, and difficult separation of tissue and support.
Soak bath pHIncreased pH with sodium carbonate to no effect. Decreased pH with acetic acid with apparently increase in transfer problems. Possibly because acetic acid is a solvent for gelatin?
Soak bath temperatureNot strictly controlled; used fresh tap water which varied between ca. 12C and 20C during the experiments, with no notable differences.
DAS nitrogen bubbles mitigationTried brushing, squeegeeing and pouring water onto the tissue to dissipate bubbles, with varying degrees of effectiveness, but no influence on the highlight adhesion problem.
Pre-mate dryingTried drying soaked tissue briefly or entirely – or (mostly) not at all. No influence on transfer properties. The pre-mate drying is to the best of my knowledge mostly/only intended for N2 bubble dissipation, and not essential for this purpose as other methods are just as effective (particularly squeegeeing).
Tissue and support matingTested squeegee and roller, underwater and with a ‘water bead’ on a work surface (with roller), with little and much force. Too little force makes the problem worse (increased adhesion problems); lots of force makes offers no improvement.
Mate wait timeVaried the time between mating tissue and support between <1 minute and >1 hour, to no effect.


FactorNotes / tests
TemperatureTried anything from temperatures where the gelatin barely melts at all up to the higher 40s (Celsius). Higher temperatures exacerbate problems like bubbles and speed up the development process; otherwise little/no effect on the problem.
Development bath pHAdjusted in the same way as the soak bath, to the same effect.
Time until separationWaited different lengths of time (usually not controlled) between putting the sandwich into the development bath and pulling off the tissue support. The only effect is that waiting very long tends to allow warm water to seep between the print and the tissue support, giving that area a head start and also making the highlight problems worse in that area. This seems to come down to the same as extending development time for the entire print, which also tends to exacerbate the problem.
AgitationAgitation always kept quite gentle, rocking the print around and tapping it with its edge to the wall of the tray to dislodge undissolved gelatin. Without the tapping motion, development progresses more slowly, but the end result (also the highlight problem) seems to be the same. Pouring warm development water onto the print from a pitcher, which seems to work quite well with dichromate-sensitized tissues, does massive and immediate damage to DAS prints.
TimeIn nearly all tests I cut the development at the moment problems were clearly visible, and at the stage where there is still some minor washing off of unhardened gelatin (maybe 99.5% of development complete). The defects are always present with technically complete development with no more runoff of pigmented gelatin.

What happens beyond development is not relevant, since at this stage, the problem is already there. Hence, procedures such as double transfer to a final support, clearing and drying do not contribute to this problem.

Conclusion and going forward (?)

At this stage, I feel I have exhausted the parameters I can think of in my testing. That’s rather annoying, as it leaves me empty-handed. During the more recent phase of testing, I have received many responses on a thread on this issue. All of these messages were very helpful, but they also underline the challenging nature of this problem. In part, it seems difficult to accurately convey the nature of the problem, with several suggestions going into directions that will not produce an actual solution (e.g. adjusting the tonal scale of the negative, or using halftone screen negatives).

Another part of the responses, in particular from highly experienced DAS carbon printers, mostly confirm materials and practices that I would also expect to work well – but that somehow in my hands fail miserably. What’s also unsettling, is that there appears to be nobody who actually recognizes this exact problem (i.e. has experienced or seen it before).

Well, I did find one fairly recent thread by Yanis Houssen who seemed to struggle with a somewhat similar problem. He successfully did DAS carbon in the past, and after a 2-year hiatus, ran into comparable problems I’m having. I reached out to him and received confirmation from him that changing to a different gelatin and ensuring a low glop melt/off-gas temperature has kept the problems at bay for him recently. I have ordered the exact same gelatin he uses, but am quite skeptical it make much of a difference, based on the absence of any clues in my testing with different gelatins that it makes much of a difference for this particular problem. But at least this gelatin I know works reliably for him and for Kees Brandenburg, whom I’ve queried on this in particular. I have also tested low-temperature glop several times now with no apparent effect, so that in itself can’t be the issue.

At several stages, the thought has occurred to me that what I’m seeing is just inherent to DAS carbon: that it has a far more pronounced (problematic) tonal threshold than dichromate carbon. To an extent, this is confirmed by some who have mentioned that the gelatin hardening action of dichromate is more robust or resilient than with DAS. On the other hand, when asked directly, top-level carbon printer Kees Brandenburg confirmed very firmly to me that the same kind of tonal scale and delicate highlights can be printed with DAS as can be done with carbon. This makes me confident that the problem I’m having is really an anomaly, and not just an unfortunate shortcoming of DAS carbon that needs to be worked around.

So that leaves me more or less empty-handed at this stage. I can only summarize what I think I know about the nature of the problem – or what I suspect:

  • It’s specific to DAS, as the problem is absent with dichromate (Yanis experienced this, too).
  • It’s not the DAS sensitizer itself, having A/B tested two sources and having confirmation of other printers who use ‘my’ DAS successfully. Moreover, Calvin Grier has dedicated a page on his website to DAS, and some of the stuff he has tried looks absolutely identical to mine. He notes that although the darker DAS doesn’t look as pretty to him as the lighter colored stuff (I agree, and I happen to have the darker material), he states that it “doesn’t seem to have much of an effect on the final image.” No mention of massive highlight problems. And despite that Calvin appears to mostly do carbon with halftone negatives, I imagine that during testing to set the density targets for his work, he should have noticed (and mentioned somewhere) any problems with printing low optical densities.
  • It’s probably none of the glop/tissue materials, or the support materials, since they work problem-free with dichromate, and/or problem-free for other DAS practitioners, and I’ve excluded influences of specific materials through simple exclusion or comparative A/B testing.
  • It’s probably none of the ‘usual suspects’ in terms of processing parameters, because I’ve tested all of them to the best of my knowledge, and very conservatively so.
  • The problem manifests itself mostly in the highlights.
  • As the problem is exacerbated by changes in materials or working methods, it tends to extend down on the tonal scale – first affecting midtones, ultimately shadows and the entire image.
  • I suspect the problem is ‘something’ that happens on the interface of the tissue surface and the transfer support. But I know it’s not something that’s only present on the surface of the tissue, since I’ve eliminated this through several tests.

Just a few more tests to do…one thing is that I want to dry some tissues in a different space, to exclude the possibility that something in my currently used drying spaces is offgassing some compound that interferes with the DAS. The other is to wait for the new gelatin to arrive. The gelatin I’ve ordered is a ‘type 2 photo gelatin’ retailed by GMW in Germany. According to Kees Brandenburg, porcine gelatins like these will harden over time and hence fog. Perhaps this is why bovine gelatins are generally recommended for carbon work. Kees recommends the type 1 gelatin sold by the same shop for DAS carbon tissues, but confirmed to me the type 2 also works, but apparently with fogging if tissues are stored for a long time.

I’ve not given up yet, but it’s disheartening for sure. If I wasn’t convinced that the process should work (after all, it does, for others), I would have thrown in the towel long ago. I’m convinced that the problem can be sorted, and that there’s one main factor responsible for the heart of this problem. The question remains, what it might be…

Leave a Reply

Your email address will not be published. Required fields are marked *