Earlier today I posted a blog with analysis of curves and especially hues of salted paper prints. I mentioned a comparison of curve shapes with some other processes as well, and since that’s fairly easy to do, I thought I could follow up on it right away. In this brief blog, let’s have a look at the H/D curves of salted paper, Van Dyke and DAS carbon.
Before I get slapped across the ears with all kinds of methodological concerns: this is not a hardcore quantitative comparison of these printing processes. Rather, it’s a couple of plots made with density measurements I had been doing for various purposes over the past few days anyway. I thought it might be entertaining to put them together in a chart. And one or two useful things can still be taken from this.
These are our test subjects. Top right is the salted paper print I did the other blog on today. Top left is a Van Dyke print I just made for the occasion. It’s a lousy print in many ways, but the purpose was to make a quick view of the curve shape and the Stouffer step wedge is plenty useful.
The shiny thing at the bottom is a carbon transfer print that’s not transferred to a final support – and likely never will be. It’s one of the 15 or so carbon step wedges I did over the past few days to play around with linearization of digital negatives. Which is something I may do another blog on, if I can muster the motivation. Boy, are those digital negatives a soul-sucking exercise. But it did give me a nice stack of step wedges to do measurements on, and fortunately, the i1Pro isn’t fooled too easily by quasi-reflective media!
Some notes about the prints and how they were made, to give you some context. The salted paper print is described in detail in the other blog. I use the readings of the gold- and heat-toned version in this blog.
The Van Dyke print was made as I always make them, and I’ll give you a quick run-through of the process:
- I mix 5 drops of 20% ammonium iron citrate, 4 drops of 11% silver nitrate and 2 drops of 8% tartaric acid. This yields a solution that is cloudy at first, but clears as the tartaric acid is added. This is brushed with a spalter brush onto a suitable paper. I used Fabriano Disegno for this test print. The paper is dried with a hairdryer.
- Exposure is typically around 3 minutes, and the exposure I did here with my 365+395nm dual wavelength light source was exactly that.
- Clearing is in two consecutive baths of tap water with citric acid added to it, for a few minutes in total. I keep clearing until the print has started to lose considerable density (which is regained during fixing).
- Fixing is in Fuji Hunt C41 fixer, 1+10 for about 2 minutes.
- Wash for 30-60 minutes under very slow running water. Air dry.
- The print used for this exercise was heat toned just like the salt print by running a clothes iron over it on high heat. The Van Dyke print was not gold toned.
The carbon transfer I made on a tissue with 4% w/w to gelatin DAS and 2% w/w to gelatin Kremer XSL Black. This is a fairly contrasty tissue formulation. Exposure was for 15 minutes (same as the salt print, coincidentally) under the same light source. I have exposed this tissue for 30 minutes and it still produces additional density at that point, so it has more ‘oomph’ in it than we get to see here in this blog. Transfer of the gelatin relief is to my standard laserprinter transparency sheets subbed with albumen and hardened with DAS. I took the density readings on the gelatin/emulsion-side. Density readings on a paper support will likely be subtly different, but the curve shape will be very similar.
Alright, let’s have a look at the H/D curves for these three prints. It’s actually an interesting picture:
In the curves above, ignore the horizontal spacing. How far apart the curves are horizontally is really random and not indicative of much. Well, there’s one thing: since the exposure time for the carbon and the salt print was identical, and the light source too, it seems that the carbon transfer works out as one stop faster than the salt print. But that’s within the specific process parameters of my working methods. Since the Van Dyke exposure was 80% shorter, its curve should actually be shifted to the left considerably. It would end up roughly overlapping the salt print curve, interestingly.
It’s immediately clear that Van Dyke poops out pretty early with a fairly low dmax. Keep in mind that this print wasn’t gold toned; gold toning would probably bring it close to the level of the salt print’s dmax. I do think this illustrates why these alternative silver-based processes benefit so much from gold toning. Without it, dmax is really lacking, at least to my taste.
Also interesting is how the carbon transfer seems to be shouldering off a bit, but it still has some zest in it. I could have exposed it for much longer and obtained additional density. But one can wonder how much good that does; beyond a certain point, it gets difficult to discern differentiations in those very murky shadows. Still, a deep, solid black is a strong statement, and in that sense, carbon does have a strong appeal.
The curve shapes are quite different too, looking at the toes and shoulders of the curves. Looking at the toe, the salted paper print starts off very gradual and eases its way into the straight part of the curve. The Van Dyke print starts also gradually, but sweeps up much more resolutely to the straight portion. The carbon transfer just lifts off of the paper base and snaps into a straight line pretty much instantly – there’s not much of a toe at all.
At the shoulder end, things are also pretty interesting. The salted paper print makes a classic shoulder that even bounces back down from dmax as exposure increases to the extreme of the curve. This is contrasted by the Van Dyke curve, which seems to bump into dmax pretty abruptly. There’s actually not much of a shoulder here – the curve just stops and levels off instantly. This is Van Dyke’s notorious habit of blocking shadows; while salted paper tends to ease off into the murky depths, with Van Dyke, you just hit a dark (brown) wall all of a sudden.
The shoulder of the carbon curve is hard to establish; it’s not in this curve, and in fact, it’s pretty hard to picture at all, because theoretically, carbon transfer can just keep adding more and more density as the gelatin matrix gets thicker. The hard limit is the point where the tissue is exposed all the way through to the base, at which point the actual transfer becomes impossible.
As a printer and viewer of photos, I experience the contrast of a print as relative to its own limits. Even though the Van Dyke has a lower overall dmax, and the curve shows roughly the same steepness as the other curves, the process ‘feels’ more contrasty to me. This is because I experience the contrast as relative to the minimum and maximum densities of the print.
So if I normalize the curves, I can more closely approximate the contrast of the processes as I experience it subjectively:
The punchy nature of Van Dyke prints now becomes apparent in its steep curve. The carbon and salted paper curves are much more similar, although the carbon curve is yet more gradual than the salt print’s. The carbon transfer didn’t really taper off into a shoulder – and arguably, it will only do so until it hits the theoretical reflective dmax or 2.5 or so. If I had exposed that longer, allowing it to build some more density, the curve would have been even more gentle in the normalized plot above.
The normalized plot also emphasizes the differences in toe and shoulder behavior. It illustrates my preference for salted paper over Van Dyke; the former is more gentle with a smoother transition in both shadows and highlights.
From the earlier plot, it’s also possible to determine the requirements on the density scale of a negative. These requirements are a tad subjective, since they rely on determining where the curves start and end. This is problematic, given how gently some (e.g. salted paper) lift off in the toe region, or how they may (or may never, in the case of carbon!) slope off into the shoulder. But approximating it a bit, I could make the following table:
| Salted Paper | Van Dyke | DAS Carbon | |
|---|---|---|---|
| First step # (highlight) | 20 | 12 | 18 |
| Density first step, logD | 3.00 | 1.80 | 2.70 |
| Last step # (shadow) | 2 | 4 | 0 |
| Density last step, logD | 0.30 | 0.60 | 0.00 |
| Delta log D | 2.70 | 1.20 | 2.70 |
The big caveat here is that the curve for the carbon transfer was not really complete, so the negative range of 2.70 is a bit of an arbitrary figure. A negative with a range of 3.00, and perhaps even higher, could be printed while still showing separation in all tonal values. For the salt print, 2.70 is really about the limit, and it covers the entire range of densities this salt print could image, from a nearly imperceptible highlight tone to virtually indistinguishable shadows. A range of 2.40 would also work quite well.
Coming back to digital negatives for a bit: this illustrates why making good digital negatives for DAS carbon and salted paper is really a notch (or two) beyond negatives for e.g. Van Dyke or even classic cyanotype. While virtually all inkjet printers will easily manage the latter processes, it’s different for salted paper and long-scale carbon, and limitations of optical density become a concern. Indeed, it’s at the limits of what I can do with the pigment inks in my Epson 3880 printer – although I understand that there are dedicated UV-blocking inks with very high UV blocking power that might be useful. I’ve never tried those. Maybe, one day?