Review of Kremer’s 30th anniversary blue pigment assortment in oil paint

There’s an excellent review on the internet of this pigment set in watercolor, but I couldn’t find one about oil paints.

I bought this set to get a chance to try some of the historical pigments like smalt and azurite, and to explore the differences among ultramarine blue variants and cobalt variants without buying a lot of pigments. I got the set on one of Kremer’s fairly rare sales (around Memorial Day and Thanksgiving); get on their mailing list to be notified when that happens.  All in all, it was worth it, and it includes a few awesome expensive bonus pigments.  By the way, if you are starting out with pigments, I’d recommend the 25th anniversary set. It’s a great selection of modern and earth pigments across the color wheel.

My quick summary is:

  • Don’t sweat which of these ultramarine blues you buy — the differences are too subtle to matter much in practice. There are certainly slight differences, but I suspect they will be hard to see in most painting situations.
  • The cobalts and ceruleans on the other hand show distinct differences in color and value that might justify buying several, and also have interesting rheology that is almost never seen in tubes.
  • Hand-mixed cobalt violets are stunning — much better than tubed versions I have, which seem full of fillers.  Cobalt violets are pretty expensive, but not as expensive as buying high quality tubes, which I have seen at $90.
  • For color alone, I think the historical pigments like azurite, lapis lazuli, and smalt have been replaced for good reasons in oil painting. (Watercolor is a different story.)  Their muted masstone and dullness in tints and mixes hardly justify their expense when good alternatives are available. On the other hand, mineral  colors can be quite interesting for  unique optics in glazing and surfaces. My view of these pigments is changing since I have tried making paint all the way from rock and earth.  I reluctantly admit that we may be losing something by not grinding rocks ourselves for certain effects, but I don’t think very many artists want to go there. I think these Kremer pigments may only get you part of the way to the real interest of hand-ground pigments, though. I’m thinking if I buy more, I will look for coarsely ground pigment, and use a mortar and pestle to do the rest.

Column 1

  • 10010 – Smalt, very fine grind – I like the muted color of smalt, which I find useful in landscapes for violet tints in shadows.  It’s a dark blue with a hint of violet in masstone, but weak in tints (making an unusual silvery blue-gray, though, which I like) and transparent.  A close match for masstone can be made with ultramarine blue, a violet and a little black; it will be much stronger in tints than the original.  I don’t find the historical version useful or unique enough to pay the price, but I like it.
  • 10074 – HAN-Blue (Note that bottle is only 1/4 full; I checked with Kremer that this is on purpose due to expense) — This is a nice color, close in hue to Ultramarine Violet Medium.  It has the highest chroma of the historical pigments and holds its strength well in tints.  It is, however, really expensive and easily matched with mixes.
  • 10170 – Ploss Blue (3 ml) — Not advisable in oils, per Kremer, and they are right!  It began to turn greenish within a day.
  • 10180 – Blue Verditer (pb30) – Also not recommended in oils since it contains copper, though I have not yet seen any color shift.  This is called synthetic azurite, but I find it closer to a greenish cobalt blue than the real azurite included in the set.  I like the color; it’s not quite the same as any of the others, but not essential, and I’ll trust Kremer that it won’t work well in oils long-term.
  • 10206 – Azurite MP, pale (3 ml) – this is so expensive and so little is provided that I couldn’t bring myself to make a full swatch!  I like its dark moodiness in masstone, but would never pay the price for it. It’s muted, a weak tinter and fairly dull in tints. Mulling may improve this somewhat, but if I had mulled it, I would have lost most of the bottle on cleanup!
  • 104200 – Sodalite, 0 – 120 (3 ml) —  This is a transparent milky light gray with the barest hint of blue.  It’s gritty, one of the class of pigments I call ball-bearing pigments (nice term from Tad Spurgeon); they kind of slide.  The two dots are sodalite by itself and with added pv19. It dulled the pv19 a bit, but allowed a very translucent thick glaze.  Sodalite’s not extravagantly expensive ($19 for 100 gr); this is one historical pigment I might purchase to test further as an element in glazes, but my tests so far indicate it will gray everything a bit. Update – I compared its effect in glazes to calcite, which has a similar ball-bearing effect and is very translucent. It didn’t dull glazes like the sodalite, so I will pass on sodalite.
  • 10550 – Lapis Lazuli, bright pure blue — An attractive deep blue in masstone, similar to its modern match of ultramarine blue though less saturated, fairly transparent, muted in tints (though it holds in tints well for a historical pigment). It’s too expensive for me.
  • 10562 – Lapis Lazuli, sky-blue — even more expensive and much duller than above.  Note that I did not mull for these swatches, so I may be missing some intensity, but not that much!
  • 36000 – Indigo, genuine –This is a nice dark color, similar to Payne’s Gray, and a strong tinter.  Kremer claims it has excellent lightfastness, but no one else agrees that indigo can be that lightfast.  I might buy it except for that issue.  It can be matched pretty easily, so it’s not worth wondering if it will fade. Recently, I tried a Sinopia pigment called Tyrian Cobalt Purple. This turns out to be very similar to indigo in tints, so might be a good source for the color, though it might be marginally more purple.  Sinopia is not forthcoming about its contents, but it is not a single pigment color.

Column 2

  • 393701 – Colored Glass, Lapis Blue, opaque – pretty light in masstone, very transparent (maybe I didn’t load the pigment enough), weak tinter but interesting as a glaze, possibly.
  • 44700 – Ultramarine Green, genuine (pg24) — This is a beauty, but is unfortunately an extinct pigment so I won’t get poetic about it. The only way to buy it as a pigment is in this set.  It’s close in hue to cobalt blue turquoise dark but darker in masstone and more transparent. It probably can be matched; there are some hue versions on the market that will give an idea of possible mixes; I might try prussian blue plus viridian. It’s a nice bonus of this set that you get to try it, even if you’ll run out forever when you’re done.
  • 45000 – Ultramarine Blue Very Dark (pb29) –reddest of the  UB bunch, but the differences are subtle. In some light, there is no discernable difference among them. Along with UB Greenish light, the darkest in masstone (again, not by much).
  • 45030 – Ultramarine Blue Greenish Extra — marginally lighter and higher chroma in masstone of the bunch.
  • 45040 – Ultramarine Blue Greenish Light – Quite close to UB Very Dark.
  • 45080 – Ultramarine Blue Light — could have fooled me that it’s notably lighter; I think they are referring to the dry pigment. As I said above, don’t sweat the differences in the ultramarine blues unless you are a pigment connoisseur. Kremer has a few redder variants than these four, but I’d bet the differences aren’t strong. By the way, ultramarines tend to have interesting long rheology and can be pulled into sometimes spectacular strings.  Some people use additives to neutralize this, but I like it.
  • 45100 – Ultramarine Violet Medium (pv15) — a nice transparent color that makes a periwinkle in tints.  There are several variants of ultramarine violet; this ones leans blue rather than red and is pretty close to ultramarine blue.  (If you want to stretch the hue range more, you might try Kremer’s Ultramarine Red instead.)  It’s very dilatant, so has odd behaviour without additives.  Dilatant paints get stiffer when you push them hard, then relax when you stop (like cornstarch). Oddly, many can be smoothed into glass-like finishes; combined with its transparency that might make this one interesting for glazes.  I like that; it’s a favorite of mine for its unusual rheology but its hue can certainly be matched in mixes, and it’s a weak tinter.
  • 45202 – Milori Blue LUX (pb27) —  I think Prussian blue is quite lovely, and mixes with other paints very well (unlike indanthrone which seems to dull mixes).  It has a persistent reputation for darkening over time, so I’m a little wary of it, but it’s a great moody color.
  • 45350 – Manganese Violet (pv16) — Very nice color in a part of the color wheel with few choices. In the same hue sector as cobalt violet deep and dioxazine violet, but a bit redder than those, more natural than dioxazine and not as overwhelming in mixes.  Reddish ultramarine violets are much less expensive but are generally weak in tints. I don’t find it essential (you can mix similar violets from the ultramarine blues and quinacridone reds), but if you like single pigment variants of violet it’s a good one that won’t completely break the bank.

Column 3

  • 45364 – Copper Blue — Like Ploss blue, not recommended for oil paint.  It didn’t change color immediately, but over time got more green. Avoid in oil unless you have an interest in the patina effect. It’s a beautiful color once it changes, reminiscent of swimming pools or aqua oceans.
  • 45400 – Zirconium Cerulean Blue (pb71) — This is supposed to be a non-toxic cerulean.  Cobalts are not as hazardous as cadmiums, but are definitely in the “use caution and a good face mask” category for mixing.  (Always good practice to use a face mask!)  I don’t like it as much as pb35 and pb36, but that’s just personal.  It’s lighter in masstone than those, sort of a muted turquoise that could possibly be used without any white in a deep aqua sky.  For my money, I think I’d get pb35 and tint it to get close, but it may prove useful for its light  masstone.
  • 45700 – Cobalt Blue Dark (pb74) — This and the following two are typical of the most common cobalt blues in tubes.  You could pick any of them and be happy if that’s what you are looking for.  This one, like Sapporo, leans slightly to red. This is the reddest, pushing toward violet even, and quite close to smalt.  If you don’t want to pay for Sapporo (below), this is closest in hue if not chroma.  Of course, in this color range, you can get an ultramarine for much less than any cobalt if that suits you.
  • 45701 – Cobalt Blue Dark, greenish (pb28) — leans slightly to green compared to Cobalt Blue Dark, not quite as dark in masstone, more expensive. Compared to all 4 ultramarine blues, this is discernably greener (but not as green as the ceruleans below).
  • 45702 – Cobalt Blue, Sapporo, ( pb74) — the highest chroma of the three in masstone; I’d probably recommend it for an all-around cobalt blue that leans a bit toward the red, except that it’s pretty expensive. It makes the two above look slightly muted.
  • 457141 – Cobalt Blue Pale (pb28) — The lightest in masstone of the cobalts, it is a fairly neutral blue that is not quite matched in tints of any of the other blues.  Its light masstone makes it useful to me in skies.
  • 45720 – Cobalt Blue Light (pb35) — Usually called cerulean blue; its partisans consider it the genuine cerulean.  Like many cobalts, it has interesting long rheology.  It’s a bit lighter in masstone and, while slightly greenish, it’s a more neutral blue than…
  • 45740 – Cobalt Blue, Greenish (pb36) — Also usually called cerulean and probably the more common variant in tubed paint, since it’s less expensive than pb35.  I have always liked pb36 better than pb35 for its chroma in masstone and greenish cast in tints, but that’s personal. It will make sky tints that are more aqua than pb35. Many brands make hue versions of ceruleans, usually a phthalo green and blue mix, but to my eye they never quite match and they don’t have the interesting rheology of the originals.
  • 45750 – Cobalt Blue Turquoise Light (pb28) — This is a dazzler, if you have a use for it: a bright, clean all-out turquoise that mixes beautifully and has interesting, long rheology. The scan doesn’t show its saturation. I just bought Sinopia’s Cobalt Aquamarine pigment; it looks pretty much identical to this color, at about the same cost. Sinopia’s version is pb36.

Column 4

  • 45760 – Cobalt Blue Turquoise Dark (pb36)  — Another beautiful pb36 variant, slightly greener and darker in masstone than Cobalt blue greenish. It’s a very rich and evocative color that justifies its expense to me.
  • 45800 – Cobalt Violet Dark (pv14) — A lovely color, not too far from manganese violet in hue, slightly redder than dioxazine, but lighter in masstone than both and not as overwhelming in mixes. It has high chroma and is surprisingly strong in tints compared to tubes I’ve bought of it; they must be full of fillers. If you like the color it might really be worth making your own paint.  Less expensive brands use too many fillers and premium paint brands charge a fortune for it.
  • 45820 – Cobalt Violet Brilliant Light (pv49) — It’s expensive, but unique. It’s an unusual color that’s very vivid, in the range of what are called opera colors (the intensity doesn’t show well in the scan).  I don’t have anything else that matches it in hue.  A substitute might be quinacridone violet and quinacridone magenta and zinc white, but I bet it won’t be as bright.  If you’re a flower painter, this might prove indispensable; it has excellent lightfast ratings.  For my muted desert cityscapes, I probably can’t justify the expense, but it might be one of those unusual ones it just feels good to have around. It’s fairly transparent but retains chroma well in tints.

All in all, I’m glad I bought the set. It’s nice to get a better feel for the ultramarines and cobalts (and I can use them all eventually); there are a few unusual standouts like the cobalt turquoises and violets which will be great to have around; and I will treasure my ultramarine green until it’s gone. Someone, start making that color again!