Sloppy slopes – Green and blue LEDs for variable contrast black and white

Recently, I wrote about a new RGB LED head for color printing and reported on some step wedge tests I had been doing. The other day I was printing some black & white negatives with this new head, and I ran into some looooong print exposures. This led me to do some additional testing, which yields some interesting (I think) insights.

Let’s start with those long exposures. We’re talking about 20 seconds or so at f/4 for ca 13x18cm prints from 35mm negatives. That’s long – but I should mention that this was using green light only, because the negatives were essentially bulletproof. I was lazy, so I decided I would do semi-stand development with Fomapan 100 in Pyrocat HD 1+1+100. I agitated the film for the first 30 seconds or so, then for a few seconds at the 3-minute mark and a further few seconds at the 25-minute mark. Then let sit for another 30 minutes or so.

The result was some seriously, extremely overcooked negatives. My brand new Soviet-style densitometer suggests densities of around 2.10logD for the brightest image areas. Nice for salt or carbon prints; not so nice for enlarger work. That’s my punishment for haphazard laziness. As usual! To be clear – the only reason I did stand development was because I couldn’t be bothered to agitate every minute.

Not to worry; I still made some OK prints, and this was just target practice on a somewhat dreary day anyway, so what gives? Still, those very long print exposures did make me wonder – how well does this combination of LEDs actually work for variable contrast black & white printing?

So at the start of my next two B&W printing sessions, I included some step wedge tests, using a Stouffer T3110. I used two papers for this test: fresh Fomaspeed Variant 311 (glossy) and some old Adox MCP312 (matte) I had lying about. Manufacture of the Adox paper went on (continuing) hiatus several years ago, so it’s obviously not fresh paper and this likely will affect the results – we’ll see this later on.

My tests revolved around the two light sources I’ve been using recently: the one using my tried-and-tested combination of 650nm red, 525nm green and 450nm blue, and the experimental combination of 680nm red, 550nm green and 480nm blue. We can ignore the red wavelengths for variable contrast B&W; I did not use them obviously for the test exposures. Also, I only tested the pure colors: both flavors of green and blue by themselves, and no blue/green mixes.

The test consists of simple step wedges, contact printed onto the paper. The strips were developed for one minute in ID62 diluted at ca. 1+3. I use my ID62 replenished and replenishment is done fairly arbitrarily and in such a way that the developer is nice and vigorous. Development was to completion in any case.

In testing, I kept the exposure the same; i.e. I only varied the LEDs used, but printed everything at the same time and aperture. This allows something (but not much) to be inferred about relative speeds etc. Conclusions in this area are limited because there’s always the difference in LED efficiency to take into account, and this interacts with the spectral sensitivity of the paper. Unfortunately, I could not find spectral sensitivity plot for the MCP312 paper.

Measurement data were read from the step wedges using an i1Pro 2 photospectrometer and imported into Excel so I could make charts just like a real financial business risk analyst.

Let’s have a look at the Fomaspeed plots, first, with log-Exposure on the horizontal axis (simply the inverse of the nominal densities on the T3110):

I excluded parts of the plot that held no meaningful data; i.e. densities beyond the shoulder and before the toe. Especially the dmax readings fluctuate a little especially due to things like greasy fingerprints, but held no meaningful information other than an impression of approximate dmax values.

Some observations:

• Foma mentions a dmax of 2.1 in their datasheet for this paper. I got close in this test, but not quite; the paper seemed to max out at around 2.00-2.05 logD. It’s very well possible that a little optimization would be possible by adjusting the developer, but the practical utility of this would be limited. The blacks are pretty convincingly black as they are.
• There is a very odd plateau in the green curves, which is especially pronounced in the 550nm exposure. It seems to suggest an interaction between two emulsion elements, with one topping out at around in the 550nm exposure around 2.2logE and the other continuing to build density beyond 2.5logE. This plateau seems to erode away as the wavelength decreases, as it’s much less pronounced at 525nm.
• The 450nm exposure exhibits a nearly symmetrical toe and shoulder, but the 480nm exposure shows a longer toe than shoulder. I’d say it has a “green toe” but a “blue shoulder”, as the toe resembles the gradual ones of the green exposures, but the shoulder is the sharp bend we also see in the blue 450nm exposure. The steepness/contrast of the 480nm curve in the shadows is very similar to the 450nm exposure, but the highlight transitions are far more gradual.

I am tempted to say something about paper grades at this point as well, but this is at the same time problematic. From the curve, we can distill the exposure range it takes for the paper to go from white to black, and this exposure range is the basis for the paper grade. It’s very nicely explained (although without much detail) on this page: https://www.rogerandfrances.com/paper-grades/

The ISO-R number is used as the basis for the paper grade, with the ISO-R being the log exposure range needed to go from white to black, multiplied by 100. The paper grades are thus be defined as follows:

logE range white to black	ISO-R range (logE x 100)	Nominal paper grade
0.00 – 0.35	0 – 35	5+
0.35 – 0.50	35 – 50	5
0.50 – 0.70	50 – 70	4
0.70 – 0.90	70 – 90	3
0.90 – 1.10	90 – 110	2
1.10 – 1.30	110 – 130	1
1.30 – 1.60	130 – 160	0
> 1.60	> 160	00

Back to my bulletproof negatives I started out with: I measured a logD range of 2.10 or so, which means it would take a paper grade “deep into 00 territory” to print well. Indeed, that’s what I noticed – I used Adox MCP312 and exposed using only 550nm green light, and still wanted to burn some of the highlights to put down some more tone there.

Back to the paper grades: Foma does publish a neat collection of curves in the datasheet for their Fomaspeed paper. I could determine the paper grades based on the ISO-R definition given above. In fact, let’s do this for grades 0 and 5, which would correspond to pure green and pure blue exposure:

In blue, I drew in the range between the start of the toe and the end of the shoulder of the grade 5 curve, and the blue lines are the same for the grade 0 curve. These lines correspond with a logE range of approx. 1.00 for grade 5 and approx. 1.85 for grade 0. That’s a little odd, because the ISO-R definitions would make the ‘real’ ISO-R for Foma’s grade 5 an actual grade 2 (ISO-R 90-110) and Foma’s grade 0 an actual grade 00 (ISO-R >160). Apparently, we’re supposed to look at only the linear part of the curve, but especially with a long toe and/or shoulder, this makes it quite arbitrary which part of the curve is used to determine the paper grade.

If I put my own numbers to this based on the measured curves shown above, the picture is yet different. I arrive at around an ISO-R of 60 for the 450nm blue exposure (grade 4) and ISO-R 230-ish for the green 550nm exposure (deep into grade 00 – more like ‘0000’).

I already concluded a few years ago for myself that paper grades are a problematic construct and not really something to dogmatically hold onto. The brief exploration above demonstrates this, at least for me. What I take from it is that the determination of paper grades depends on many factors and a couple of (usually left implicit) assumptions, and as such, speaking of ‘grades’ is inherently fraught with uncertainty.

Simply put, one man’s grade 2 might be another man’s grade 4, and if that’s the case, the numbers lose their meaning. In the end, actual response curves are much more telling. If you’re a numbers person, I can only recommend getting a couple of step wedges. And if you don’t care much about numbers, just visually determine what blue and green (or yellow and magenta, if you’re using a dichroic head) exposures do for your paper and negatives, and proceed on that basis.

Anyway, let’s have a look at the MCP312 curves for the same light sources. Keep in mind that this paper has been sitting around for a few years. It was refrigerated for much of that time, but I suspect it will have been affected by age. Resin-coated paper tends to age much less gracefully than fiber-based paper, after all.

Again, a few things stand out:

• With 550nm green exposure, I didn’t actually get a convincing black in the first place. I chose to keep the exposure the same as in the other tests; I could have increased exposure by a couple of stops, but this was already 16 seconds at f/4, so I would have had to go well over a minute to try and hit the paper’s dmax.
• The green exposures show a fairly linear response. Especially the 525nm exposure, that wasn’t truncated, shows a nearly perfect straight line from its shallow toe all the way to its dmax.
• Just like the Foma paper, the green exposure by itself (looking only at the 525nm one here; the 550nm would have required more exposure) doesn’t quite hit the same dmax as the blue exposures. I truncated some data beyond the shoulder of the 525nm exposure, but the highest density I measured on that strip was around 2.02logD anyway.
• The dmax of blue exposures on the Adox paper is a little higher than on the Foma paper: it peaks at around 2.25logD.
• The 480nm exposure shows the same relation between toe and shoulder as the Foma paper: a shallow ‘green’ toe and a sharp ‘blue’ shoulder.

If we must do the grades thing…Adox doesn’t publish curves in its MCP datasheet, so I don’t really know how to interpret their grades to begin with. Looking at my own measurements, I arrive at an ISO-R of around 80 for the blue 450nm exposure, which makes it a grade 3, and an ISO-R of around 240 for the 525nm green exposure, so again a solid ‘grade 0000’ like the Foma paper.

These ‘grade 0000’ extremes are something I suspect the paper manufacturers don’t really consider as the normal bandwidth of their papers. I also doubt if you get these extremely low grades if you use Multigrade filters or a dichroic filter head, but I’d have to test that separately (but I don’t know if I actually will). There is the odd curve shape of the Fomaspeed paper when using only green light, which surely isn’t intentional. And there are the very, very long exposures required to hit dmax on these low grades. Yes, it’s normal that the extreme grades require more exposure than the middle ones, but pure green light (esp. 550nm on the Adox paper) takes this to a whole new level. Still, it works “in a pinch”, as demonstrated by my prints from the heavily overcooked negatives.

Now I’ve got these curves anyway, it’s just too tempting to pitch them side by side. This is what we get:

The Fomaspeed develops a lot more contrast under 450nm light. It makes me wonder if perhaps the Adox paper could be pushed to an even steeper curve with a lower wavelength. I might try this one of these days, since I am pondering about a lower wavelength test for Fuji DPII paper as well. But it’s also possible that the steep emulsion of the Adox paper just isn’t so steep anymore due to its age. I don’t know, really.

At 480nm, which is a cyan kind of blue that clearly affects both high- and lower-gamma emulsions n both papers, the curves also become a little more similar. The Adox has a rather long and sagging toe, the Foma has a rather pointed shoulder. The Adox paper is still a little less steep on average, but the difference is much less pronounced than at 450nm.

At 525nm, the Adox shows a near-perfect straight line response that’s a little reminiscent of something like salted paper or even carbon transfer. The Foma paper starts to do funny things.

At 550nm, I simply get the impression I’m doing something with these papers that they’re just not made for. The Adox paper just gets insufferably slow. This explains my initial experience I started this blog with – what’s going on with these super long exposures? I can only conclude that 550nm is probably at the edge of the paper’s spectral sensitivity. There’s really no point in going there, especially since 525nm green already gives a nice and long scale (in case this is required), but more sensible sensitivity. The Foma paper is just being wayward and odd. Again, there’s no point in subjecting it to this kind of torture.

Firm conclusions are hard to draw and they’re also not necessarily what I was/am looking for here. This is more an amusement kind of thing; “let’s see what this button does”. To the extent that conclusions are warranted, and please take them with a healthy dose of salt, I’d say something like this:

• 550nm green by itself isn’t very useful for VC B&W printing. 525nm green works well enough. That’s a bit of a pity, since 550nm is really perfect for color printing. The workaround is probably to mix in a little blue with the 550nm green when printing variable contrast B&W. I tried this yesterday and it seems to work quite well, also on the Foma paper, but I only printed images, no step wedges, this way.
• 480nm blue is a nice intermediate grade, but for a split-grade approach not very useful since it doesn’t produce the highest contrast the papers are capable of. I suspect that 450nm blue is pretty close to the shortest wavelength Foma needs for its highest grade. There’s a possibility that the Adox paper might do well with yet lower wavelengths, but I have yet to try this.
• There are obvious curve and density differences between the Foma and Adox papers, and in practice I find both print perfectly fine. But there does seem to be some reason to avoid the very lowest contrasts on the Foma paper, because that’s just dodgy stuff.

Well, to be continued, no doubt. These tests are quick enough to print- much quicker to print than to interpret, let alone report!