In my move to DAS instead of dichromate for sensitizing carbon tissues, I ran into a highlight retention problem. Thin gelatin layers (highlights and upper mid-tones) would slide off the temporary support material during warm water development. In testing, I identified exposure unit wavelength as a critical factor. However, the problem persisted especially in very high-contrast tissues. It seems I have now identified another part of the solution: a different way of subbing the intermediate support.
Let’s have a look at the problem first. Here’s an example that triggered my new bout of problem-solving:
This is a small area (3.5 x 2.5cm or so) of an 8×10″ carbon transfer, using a combination of Pbk7 carbon black and Pb16 cyan pigments. During warm water development, once most of the unhardened gelatin had already washed away, the mid-tone area depicted started to develop blisters and just lifted away from the support. Hence, in the white area, no part of the image-forming gelatin matrix is left; it’s all gone.
The problem seemed to vary a bit depending on especially pigment load, which makes sense. A higher pigment load means that the same optical densities require a thinner gelatin layer, and the thinner this layer, the more fragile it is. So initially I decided that there’s just a limitation to how high you can go with the pigment load in a DAS tissue. Consequently, I went in the direction of longer-scaled negatives, lower pigment loads and longer exposure times. Still, the problem also occurred in tissues that quite frankly should have just worked. The example above is from a fairly contrasty negative that really, I should be able to print well with DAS carbon. So something was amiss.
The first thing I did was to experiment with a couple of different glop recipes. I had a feeling there might be something in my glop formulation that resulted in degradation of the bonds in the hardened gelatin. I figured that the following culprits were possibly suspect:
- The actual pigment; most of my carbon presently uses Kremer Pbk7 dry pigment, which I like for its neutral image tone.
- Orotan 731K dispersant, which I use to make pigment dispersions from dry pigments (like the Pbk7 mentioned).
- Glycerin, which I have gotten used to add to keep the tissue more pliable/supple once dry.
- Gum Arabic; this one’s a bit of an oddball since I wasn’t in the habit of using it, but because it’s a common component of paints and inks (including the Talens India ink that I’ve used a lot for carbon transfer), I wanted to investigate its potential role.
I made a couple of ‘skeleton’ tissues, without any glycerin and also without the Orotan 731K dispersant, so their potential influence would be eliminated altogether. To test if the Pbk7 pigment itself played a role, I made a test batch using Talens India ink instead. I used Talens India ink specifically because Calvin Grier explicitly recommends it for carbon, and I know he has done a lot of DAS carbon with it. To verify that the gum Arabic present in that ink didn’t pose a problem, I made a test batch that used Pbk7 pigment, but was absolutely free of any gum Arabic.
Leaving out the Orotan dispersant required departing from the acquaeuous pigment dispersions I had been making using a simple rock tumbler. I went back to mulling, and actually found that it works quite well. In my early attempts at mulling, I used the Orotan dispersant and water, which gave a very liquid mixture and did not mull very finely. I’ve now mulled with more viscous media, and that seems to work a lot better. I’m using either a 30% gum in water solution (which is of a rather thick, syrupy consistency), or…honey.
Weigh out some pigment, add just enough gum or honey to make it mullable, and then mull away until perfectly smooth. It doesn’t take very long, and the result appears to be excellent, at least going by the initial tissues, prints and smear tests I’ve done so far. The only downside is that it’s impossible to retrieve all of the mulled paint from the muller, glass plate and other utensils. This means that there’s a small margin of uncertainty in how much of the dry pigment it actually makes it into my glop.
Anyway, I ended up making a couple of test tissues, with the following characteristics:
- 250mg Pbk7 pigment, mulled with 1ml of 30% gum. 10g gelatin, 3g sugar and 400mg DAS for a total volume of 100ml glop. This provided a very nice and smooth tissue surface without apparent pigment clumping.
- 1ml Talens India ink, 10g gelatin, 3g sugar and 400mg DAS in 100ml glop. For some reason, I seem to have pigment clumping issues these days with this ink, and it shows up pretty badly in the tissues (rough/uneven surface) and prints (very visible ‘pepper corns’).
- 250g Pbk7 pigment, mulled with 1ml honey. 10g gelatin, 2g sugar and 400mg DAS in 100ml glop. Note I used only 2g sugar because I already had 1ml of honey in there, and that’s basically sugar as well, I reasoned.
The gum-mulled tissue seemed to work quite OK, but still had some minor problems with highlight flaking. The honey and India ink tissues actually preferred pretty horribly, with very severe flaking. The India ink tissue also had the pigment clumping problem I mentioned earlier. So it seems I didn’t find any solutions this way – just a few more problems…darn!
Well, let’s put it optimistically: I learned that mulling pigment with nothing else than gum or honey actually works quite well. The stuff even stays nicely suspended in a gelatin glop left to outgas for a couple of hours, and that’s better performance than I had anticipated. Also, it suggests that constituents do make a difference in terms of highlight adhesion/stability in DAS tissues. Furthermore, leaving out the Orotan dispersant seems to cut down the foaming I got during the development stage, although this is only a minor inconvenience. Finally, gum doesn’t seem to hurt, and if anything, it may even help a little. At least the pigment-and-gum tissue seemed to be the best performer.
So I looked into a different direction instead – that of the temporary support. After all, the problem is with part of the gelatin matrix peeling off the support. This can be due to the stability of the matrix itself, but really, during initial warm water development, it seems to do just fine. It just seems to sort of blister, and then slide off. So perhaps it has something to do with the layer underneath the actual image layer.
For the temporary support sheets, I’ve been using albumen-subbed laserprinter transparencies. These have worked extremely reliably for dichromate double carbon transfer. However, is that also true for DAS-sensitized tissue? There’s only one way to find out…
So I tried replacing the albumen with gelatin. The subbing solution I tried was either 250mg or 500mg of gelatin to 100ml water (i.e. 0.25% or 0.50% w/v) with 1ml of 10% chrome alum. Bloom and then melt the gelatin in e.g. 20-30ml of water, make up to 100ml by adding tap water, then add chrome alum and use immediately. The workflow for making the subbed transparencies becomes as follows:
- On recycled/reused sheets, first soak in a weak sodium hydroxide solution to soften any remaining gelatin image or remnants. Scrape off softened gelatin muck, then rinse a couple more times. Skip this step for virgin-out-of-box sheets.
- Prepare subbing solution: weigh off 500mg gelatin, bloom in 20-30ml water, then melt in a hot water bath and add water to make 100ml. (The albumen version used 2ml liquid albumen per 100ml water and requires no melting). Add 10% chrome alum solution. The solution does not keep and must be used immediately.
- Put subbing solution into tray, then load in the sheets, one by one, swishing around the subbing solution so that both sides of each sheet are washed with it.
- Remove sheets from tray one by one, and hang to dry.
- Sheets are ready to use 36 hours or so after drying. (The albumen version is usable immediately upon drying.)
It’s a quick and easy process with essentially just one step, as opposed to preparations that first use a subbing bath, then dry the sheets, followed by a separate hardening step. The albumen version, which requires no wait time before use, works very reliable for me with dichromate carbon transfer.
More importantly, the gelatin version seems to work just fine for DAS carbon. Even the ‘tricky’ tissues (pigment mulled with honey, and pigment-clumped India ink) seem to transfer just fine to these gelatin-subbed sheets. I suspect this might be because the gelatin subbing swells and thereby makes good contact with the image matrix. It’s also possible that the gelatin subbing just chemically crosslinks better with the gelatin of the image itself, and/or that the gelatin subbing adheres a little better to the substrate material.
The only downside I can find so far compared to the albumen subbing I used before, is that the gelatin-subbed sheets need a day or so to fully harden. I’ve had mixed results with sheets that had just dried,
but success rate seems to rise to 100% if the sheets have been allowed to rest for 36 hours or so. [Update: Uhm, no – see note at the end of this post!]
Granted, I’ve only done a limited number of tests so far, but prints that were dodgy/hit & miss on the albumen-subbed sheets now seem to come out just fine on the gelatin-subbed ones.
What I take from this is that this set of experiments demonstrates how many variables interact with each other in carbon transfer. It’s often difficult to predict which factors play a role in any problems that occur. It really helps to systematically experiment, as I did here by attempting to isolate specific groups of factors (glop formulation, temporary support subbing) and specific factors (pigment, dispersants, subbing chemistry). I’m also quite happy with the ‘finding’ that mulling pigment actually works better the way I’m doing it now – it only takes a few minutes, and allows me to get from dry pigment to mixed glop directly, without storing a dispersion in the meantime with the associated risks of deterioration, clumping, sedimentation etc.
Update 11 May: well, I have to revoke my earlier optimistic claims of having solved the issue. In fact, the problem still seems to be there, quite as bad as it was before, so there are other parameters that play a role in this. Sadly, I’ve not yet figured out what those may be. I’ve tested a whole slew of other subbing approaches, changes to my transfer workflow etc. etc., but there’s just no pattern emerging. Hopefully I’ll revisit this in a while with a proper solution!