Top Lightfast Watercolor Pigments! Build a professional versatile palette for unique color mixing.
A list of lightfast pigments that I highly recommend for watercolor painting. These colors are what I use most often in my palette. They are versatile for a wide variety of subject matter as well as capable of creating unique granulating and color separating mixtures for visual interest. It's important to me that these colors mix well together for overall harmony within this palette. I include several complimentary colors for quick and easy mixing of shadow colors, near-black and neutral grays. Easily replicate complex colors like Moonglow and dupe even the most "rare" pigments. You can create DIY mixtures that match Daniel Smith's Primatek Mineral paints.
While I choose paints with lightfast ratings of LFI to LFII (or Blue Wool Scale 7-8, with a low of about 6 diluted) there are exceptions. PR122 Magenta is sometimes under-rated due to an outdated sample test marking it as LFIII in 1999. That is an error, since many modern brands have been thoroughly UV tested by myself, Golden and Handprint resulting in LFI masstone (LFII when diluted/BW7-8). You can not always rely on ASTM or manufacturer ratings. Another poorly rated example being PR177 Anthra Red, marketed as a replacement for the fugitive PR83 Alizarin Crimson. It turns out that it fades just as much in tints/diluted (LFIV fading for a pigment often labeled as LFI). This is likely due to a test in masstone, oil/acrylic/colored plastic for appliances or misinformation from the dry pigment source. There are other reasons for unexpected fading, read more on the fugitive list.
I'll be listing each color in order of versatile mixing priority, with urgent colors listed sooner to help those looking to form the most compact limited palettes. I'll share my reasons for each color, but you may find that you require a different color priority due to the type of art you create. The needs of a botanical painter are different than a portrait, animal artist. Landscapes or abstract work may rely more on color separation and granulation.
LIST - Brand suggestions and reasons for each color are further down, or click a color name to go to that part of the page:
2) Magenta PR122,
8) Burnt Umber PBr7,
9) Viridian PG18,
11) Potter's Pink PR233,
14) White Gouache PW6,
--) Coming soon!
---) See lower page for notes on additional colors being considered. Also - pardon the clutter - this list is currently in progress. I'm continually editing my notes, adding updates as I create new videos, scanning new swatches and figuring out how to best replicate complex convenience mixtures.
Where to buy these watercolors? Click any swatch card below to visit where I purchased this paint. More info, general advice for beginners, as well as the benefits of buying tubes (instead of dry pans) towards the bottom of this page.
When brands create unique looking convenience mixtures, it's easy to be persuaded into buying them while assuming you couldn't re-create them. Daniel Smith makes a lot of beautiful color separating mixtures (Moonglow, Shadow Violet, Imperial Purple etc.) that are very tempting because of their dramatic color separating appearance. The deep purple (when used on dry paper) of Moonglow or Artemis separates out into blue-green and red when it hits water (or when you sprinkle table salt onto it as seen in the swatch cards above).
When I first started painting I had collected way too many colors made up of multiple pigments that I later discovered were easy to mix on my own. Having a decent collection of single pigment paints will give you the most versatility in the long run. If you spend some time mixing your own favorite colors, you may find yourself saving money over time by fighting off the urge to buy things you already have the core ingredients for. Here I've swapped out the fugitive Anthra Red PR177 they used (and misinformed customers that it was lightfast) with a more stable Quin Red PR209 instead:
I definitely recommend starting with primary mixing basics:
1) PB29 - Ultramarine Blue: My personal favorite is Schmincke's French Ultramarine (I usually buy them at Jackson's here because USA shops like Blick pass on costly import fees). This version provides the most extreme granulation resulting in interesting textures and color separating mixtures.
If you prefer a smoother appearance, Schmincke also makes an "Ultramarine Finest". There is value in having both French Ultramarine and Ultramarine Finest in your palette since they have such a different texture with variable end results in mixtures. As seen in the swatch cards, you can tell when a color has smooth finer particles based on the salt texture reaction. Salt will easily push these lighter fine particles away causing a star/snowflake pattern, while heavier granulating pigments stay put and do not have any reaction. You can't really go wrong with other brands if Schmincke isn't an option for you, they are all just moderately granulating middle grounds of these examples though. In general, if a brand has a regular "ultramarine blue" and also an ultramarine "deep" or "french" ultramarine, then typically the "deep or "french" is a darker, slightly redder shade that granulates more. When using brands other than Schmincke (that have less differences between their versions of ultramarine) it is not as worthwhile to have two varieties in your palette. If you're uncertain which texture level you prefer, having ONE moderately granulating cheaper Ultramarine Blue Deep from White Nights or ShinHan would be a reasonable option while you discover your favorite uses for this color.
2) PR122 - Magenta (alternatives are PV42 or PV19 sometimes called Permanent Rose): For PR122 my top picks are White Nights Quin Rose, Sennelier Helios Purple and Rosa Gallery Magenta who all make excellent versions available in both tubes or pans. Roman Szmal's Quin Magenta (not quin pink also made with PR122) is really nice, but that is available only in a dry pan (keep this convenience in mind if you plan to pre-mix your own frequently used multi-pigment colors in empty half pans). Qor Magenta is very strong, but unruly and can cause unusual flow and texture problems when mixed with other brands.
Some versions of PR122 are arguably a touch closer to blue on the color wheel than ideal. It's most important to avoid ones labeled as "opera, bright, opus, brilliant, neon or fluorescent" because those options mix in fugitive dyes (sometimes without warning that anything but PR122 is the ingredient). This beautiful pigment is just plain useful as a floral pink while also fitting the bill of a primary mixing magenta. There are actually several nearly identical pigments that can be used as a "cool red" to mix vibrant purples with ultramarine or phthalo blue. PV19 often called Permanent Rose as well as PV42 Quin Pink/Magenta are also solid choices. All 3 of these pigments can vary by brand/pigment source ranging from a near-purple to a warmer red. Often, the clean bright PR122 magenta color is difficult (impossible?) to recreate with mixtures of other colors. (Adding a yellow to PR122 can make it warmer to appear like many versions of PV19, but those versions of PV19 mixed with blue do not quite look like PR122.) I will have to determine which PV19 of the dozens available is the most UV stable, versatile and unique for color mixing before elaborating upon it as an alternative primary cool red.
3) PY150 - Nickel Azo Yellow: A few brands make perfectly acceptable PY150 options, but I really like DaVinci's Nickel Azo Yellow because it is a particularly strong version. Dilutes to yellow, but all brands will look nearly brown at full strength / when layered enough.
Additionally, if you don't mind Qor's aquazol binder (which has unusual flow properties) they also offer a very strong version. Mission Gold brand is the only one I really avoid for this specific color, due to some odd greenish discoloration both Denise Soden of In Liquid Color and I noticed several months after swatching. I can't be sure if this was just an old batch issue or possibly chemical interactions with our paper/nearby paint chemicals. Luckily there are plenty of of alternatives and this is such a strong pigment you're unlikely to come by a weak version. Be prepared to dilute it a lot to get a nice yellow, instead of the dark mustard-like brown when full strength. A little goes a long way, I struggle to think of a single more cost efficient paint.
PY150 is one of the very few totally transparent AND lightfast yellow pigments. Extremely useful for glazing/layering techniques. My first choice in yellow due to the realistic foliage greens it mixes when used with Phthalo or Ultramarine Blue. A slightly brownish honeycomb yellow that dilutes to a mustard-like color, this pigment is a really strong mixer (a tiny bit goes a long way) and its masstone to diluted value range is huge for a yellow. When diluted with water it is capable of passing for a mid yellow, but never quite achieves the brightness of a cool lemon yellow in case that is important for your chosen subject matter.
4) PB15 GREEN SHADE (sometimes also noted as PB15:3) Phthalo Blue: Available in many brands. Caution - some brands do not distinguish when it's PB15:3 (Blue-Green shade, not PB15:1 Red shade such as the Winsor and Newton swatch cards shown simply labeled as PB15). While you can't really go wrong with any type of Phthalo Blue for mixing, the green-shade version of PB15 is the closest thing we have to digital cyan. It is capable of mixing the most vibrant purple and deep emerald green secondary colors.
The red-shade variant is a touch too warm, being closer to Ultramarine Blue's purple-leaning side of the color wheel. As an alternate to mix things up sometimes, I also enjoy PB16 an even more green-leaning phthalo blue often called "Phthalo Turquoise". Another notable alternative for a quirkier palette would be Winsor and Newton's "Aqua Green" which has a touch of granulation and is made from a pigment with no code number called Palomar Turquoise manufactured for the automotive industry. Definitely check out the pigment database if you're interested in odd ball colors.
5) Bright Lemon Yellow - PY175 is more transparent than Cadmium Yellow PY35 and more reliably lightfast than Hansa Yellow PY3. Holbein Imidazolone Lemon PY175 is my go-to yellow for bright vivid yellow objects such as goldfinch or canary birds, lemons, florals, sunlight etc. While it can be used to mix spring greens with blue pigments, as well as acceptable oranges with red pigments, its best feature is pure bright yellow subjects that Nickel Azo PY150 just can not reach in bright vibrancy.
CAUTION - My previous go-to was PY74, commonly offered as a primary yellow, which failed a lightfast test when diluted. I then considered PY3 HANSA YELLOW LIGHT which is a stronger pigment with arguably better mixing performance. Unfortunately I've been reading about fluctuating lightfast issues with PY3 (stability varies both by brand and even batch to batch within a brand). I have yet to personally experience this, but am doing a thorough masstone and diluted range check for all brands at this time. Finding an alternate cool lemon yellow that is transparent (most versatile for mixing) AND totally lightfast is quite the challenge. PY3 Hansa Yellow Light is a bright yellow that feels quite perfect for mixing, but has less than perfect lightfast ratings. Often listed as LFII or 3 of 5 stars on Schmincke's scale. Golden, the maker of Qor watercolor, did a study about PY3's lightfastness which determined that it's SOMETIMES fugitive in one brand and lightfast in another. This was true even of batch to batch differences, which Golden discovered because one of their batches tested less lightfast than their previous batch. I'm going to gather all of my PY3 watercolors to do a side by side lightfast test to compare and see what I find.
The major problem with choosing a second ideal mixing yellow is that there are very few transparent yellow pigments in existence. Many are cloudy or totally opaque, such as PY184 or PY35. When you use an opaque color for mixing watercolor it limits the value depth of color layering/glazing. This can sometimes give you unexpectedly muddy or cloudy/chalky looking secondary colors. Holbein or Winsor & Newton's PY175 appears to be the best transparent alternative. Winsor and Newton's pan formula re-wets easier than their tubes left to dry in your DIY palette, which is good to keep in mind if you are in a particularly dry climate that causes you to have to scrub at your paints to reactivate them. I do NOT recommend Daniel Smith's "Lemon Yellow" because it is desaturated and dull looking (banana-like) compared to the brighter lemon of Holbein or W&N's version. Schmincke's is also a touch too dark valued. Lukas has an acceptable third place option.
6) PR209 - Quin. Coral, Red or Cherry: As far as the variety of "warm reds" go, this is about as cool-leaning (towards pink) as it gets. It is quite warm when compared to Magenta PR122, but with a pink undertone that makes it very useful for mixing. Your choice of red may vary by subject matter painted. I've found that PR209 is great for florals as well as portraits, mixing well with diluted browns for adding blush and lip color. Full strength it reminds me of the color of coral-pink make up products (blush and lipstick).
This is one of the best reds for mixing with granulating pigments to create dramatic color separation. Most brands are very transparent with lightweight finely ground pigment particles, which allow it to spread out across the surface of wet washes. It is just cool (pinkish) enough that it mixes purples instead of the dull wine/browns that result from warmer reds mixed with cobalt teal, ultramarine blue or viridian. While warm in comparison to PR122, PV42 and many varieties of PV19, this red is quite cool compared to other standard "red" colors. It looks red in masstone and dilutes to a punchy coral pink-leaning color. It's a color that is so unique that I can spot it within a sea of other pink or red swatches, which only further establishes it's unique importance. The diluted pink range of PR209 makes it a perfect substitute for PV19 in replicating convenience mixtures like Misty Morning by Roman Szmal (see Cobalt Teal further down the page).
There is versatility to transform PR209 into a warmer red by mixing it with a tiny touch of yellow. It can be further mixed to make ripe orange colors when paired with lemon yellows such as PY3 or PY175. I prefer Sennelier's "Quin. Red" (tube) and Roman Szmal Aquarius "Cherry Quin Red" (pan) both are superbly smooth versions with fine particles. Sennelier provides amazing salt reactivity (snowflake-like, see swatch) and very flat washes with no texture issues like I noticed in M Graham and Mission Gold's versions. This texture appears like tiny gritty specks, minor, and more subtle than granulation. Daniel Smith's Quin Coral is a suitable option, being a bit of a balance between the oddly textured versions and the smoothness found in Sennelier/Roman Szmal.
7) PR254 Da Vinci or Winsor Red - You may not require a second red depending on your art style/subject matter. I do not like Daniel Smith's Pyrrol Red because it has an unusual wet to dry shift, darkening as it dries. Some artists may prefer an even warmer red or red-orange pigment in this palette slot.
I chose PR209 first, as it is one of the only transparent and lightfast pigments fitting into the generally "warm" category of red pigments. PR254 (Winsor Red in W&N's pro watercolors) has a slightly deeper masstone and is a stronger mixer, but it is semi-transparent. This subtle opacity varies by brand. It is an acceptable alternative to PR209 with just a touch more warmth that takes away that coral/punch look that some may not appreciate (especially if you're not using it to create color separating mixtures). It's also a red that botanical artists may need for painting red flowers without the pink undertone of PR209. I can see value in having both PR209 and PR254. If you prefer to skip PR209 = By mixing PR254 and PR122, OR by adding a PV19 Permanent Rose color later on, you may be able to replicate the role of PR209 in color separating mixtures.
Artist preference notes: If you think you'll need stop-sign or fire engine reds, other Pyrrol or Cadmium Red PR108 are also options, but I do not reach for them colors often. Despite its opacity, Cadmium can be notably unique. PR108 is a very powerful, gently granulating, solid matte red that dilutes to a warm peach color (can be helpful in portraits). Because Cadmium is opaque, mixtures with it are limited in value depth (reducing it's usefulness in glazing/layering). Some artists love deep rose reds, like Alizarin Crimson PR83 and Anthra Red PR177 - both of which are fugitive with severe fading issues. I am currently lightfast testing hundreds of reds in search for a good replacement. As an alternative to Alizarin, PR179 Perylene Maroon may be a valuable mixing color as a near-brown deep red (that lends itself to the shadow values of red rose flowers). Winsor and Newton has a lightfast replica called Permanent Alizarin using PV19 and PR206. Sennelier offers a few rose-red colors, but I am weary of them due to including PY83 which fades in tints.
8) PBr7 Burnt Umber: Da Vinci brand's Burnt Umber is my favorite version of this extremely useful brown. It's slightly warmer and smoother flowing than White Nights or Roman Szmal, but those brands also have a very strong example of PBr7.
Avoid the brands that have weaker/pale earth browns (Sennelier, Winsor and Newton) or offer odd multi-pigment mixtures for their "Burnt Umber" (like Mission Gold or Schmincke, who mix red iron oxides with black pigments). This is a priority right after the primary colors due to the fact that it's so time saving to not have to mix all your colors together to get such a frequently used color. It can be used as a convenient brown for trees and branches, animal fur as well as landscapes. PBr7 can be used as a skin tone (dark skin in masstone or light when diluted). It's also a common way to mix neutral shadow colors and even a dark black when paired with Ultramarine Blue PB29. This instantly adds a quick way to paint "black" objects that look more natural / realistic, while being much more versatile than adding a black pigment at this time. Be aware that your choice in PB29 Ultramarine Blue changes the granulation and color separation in wet washes:
By mixing more blue than brown you will have a cooler gray, like Jane's Grey. From there your mix can be more neutral (near-black with very little color temperature bias) with the addition of more brown. It gets warmer with more brown, which can be helpful as a neutral tint or natural looking shadow color. You may hear the advice "never use black" in watercolor, which is primarily due to true black pigments having little color bias, looking flat, dead, lifeless or unrealistic. By using a mixture you can find the most accurate representation for warm and cool shadows that contain hints of color to represent reflected light of nearby objects. That being said, for those who are not painting realism, particularly graphic designers, illustrators and abstract artists - I will elaborate on true black pigments later in my list. An ink-like deep true black can be very helpful for those doing certain types of art.
9) PG18 - Viridian: This seems to vary substantially in strength and granulation levels between brands. M.Graham makes the most saturated, granulating and easily re-wet version. This appears to be due to the fact that Viridian pigment dries hard, requiring more scrubbing to reactivate than most pigments. The honey content in M. Graham and Roman Szmal keep it hydrated, making this pigment easier to re-wet than other non-honey brands.
If you own a hard to re-wet brand, you could try helping it by stirring a drop of vegetable glycerin or honey into your pan (when filling with a tube paint).
Daniel Smith and Schmincke also make good versions with a decent amount of granulation, but Daniel Smith is slightly preferable due to less binder gloss compared to Schmincke. I did not like the weaker, less granulating, harder to re-wet versions from Rembrandt or W&N. DaVinci falls in the middle with good pigment load but no interesting granulation. Lukas had a very close match to Phthalo Green (blue shade) PG7, a similar color that is much smoother - non granulating and staining. Viridian by Lukas lacks the granulation necessary for color separating mixtures.
While Viridian is not a very intense pigment, and is a weak mixer in comparison to the similar hue of Phthalo Green (more paint is required, easily overpowered in mixtures) Viridian is still a vital part of my palette. Because of Viridian's unique pigment particle size and weight, it behaves much differently than PG7 when used to make your own color separating mixtures. For instance, in Da Vinci's Artemis, or Daniel Smith's Moonglow (made of PG18, PR177 and PB29) the Viridian clumps closer to the also heavier/thicker particle Ultramarine Blue to cause a blue-ish green separation away from the Anthra Red. When using PG7 instead the mix becomes prone to creating a brown and blue separation (thin particle Phthalo sticks to the Anthra Red where they BOTH separate from the Ultramarine Blue).
Viridian mixed with PR209 creates a variable range of gray, which can be extremely interesting in wet washes where granulation is encouraged. While PG18 is able to mix nearly the same masstone on dry as PG7, the differences are apparent when used on very wet paper. If you are not a fan of granulation, you may consider selecting the more intense and staining PG7 in this palette spot instead.
In all honesty, I prefer to keep two versions in my palette, because different brands offer dramatically different granulation levels and slightly varied hues (blue leaning or green leaning). Please note that the most common version of PB28 is Cobalt Blue. I do not recommend the darker blue version of PB28 because it's such a close match to the more useful ultramarine blue PB29. The ideal, brighter teal versions of PB28 are typically marked as "Turquoise". For example Rembrandt's Cobalt Turquoise Blue is made from PB28, but is a very similar hue to the PG50 versions of this color.
When mixing replicas / dupes of other brand's convenience colors you can get variable results using the same pigment codes from other brands, and remarkably similar results using different codes altogether. It's all about finding a color that is most similar, not just the same ingredient code. PR209 was a good replacement for PV19 and PB28 from White Nights is remarkably beautiful in the second mixture below.
You can decide, just like as in PB29 Ultramarine, if you want one certain type of teal or one granulation level or if you need one of each like me. The most granulating is Daniel Smith. A more subtle version that behaves uniquely in mixtures is White Nights Cobalt Turquoise, which also happens to be a very affordable version of this typically expensive pigment. The color varies between near-neon bright sky-blue to more green leaning teal. There are a LOT of GOOD options for these pigments. You are unlikely to go wrong with any reputable brand. Sennelier's bright blue-leaning "Turquoise Green" is also amazingly bright and I also love Winsor & Newton's near perfect balance of not too blue not too green but plainly a teal "Cobalt Turquoise Light". This color is ideal for painting the edges of tropical oceans and highlights on Robin's eggs.
11) PR233 - Potter's Pink. I'm currently using MaimeriBlu, but Schmincke is a close second place. This is a dull, nearly brown, pink with extreme granulation. PR233 creates an amazing mixture when combined with Cobalt Teal/Turq PG50 (as shown in the background of the bird painting below) where the darker vintage pink granulates and separates out from the bright blue.
This pigment can be hard to re-wet from dry, it may be helped by adding a drop of glycerin or honey to your pan (when pouring from a tube). If buying PR233 from Winsor & Newton, definitely get the pan version if planning to use it from dry which re-wets easier than their tubes. Daniel Smith, MaimeriBlu and Schmincke all make similar versions of PR233 which are very granulating but slightly darker in hue than W&Newton. I do not like the Roman Szmal version, as it is weaker with a more pale dusty residue that can look chalky.
The background on my Wren bird painting above uses a pre-mixed Cobalt Teal + Potters Pink combination (Schmincke's limited edition Glacier Green convenience mixture). It eagerly color separates in wet washes, ideally used on cold press or rough papers where the pigment particles can really settle into the dips and valleys of the paper surface. Potters Pink can also take on a peach hue when you mix it with a touch of yellow, or more vibrant pink when mixed with magenta PR122 or rose PV19. If you're looking for a granulating red, those are very rare. Cadmium Red PR108 is a possibility, but the texture amount varies dramatically by brand. PR108 is always opaque and its overpowering nature can make it tricky to work with.
12) PBr11 OR PY119 - Lunar Earth, Aquarius Brown or Magnesium Brown. I prefer Daniel Smith's Lunar Earth because it is available in tubes compared to Roman Szmal's pan, but both produce a lot of granulation texture. Tubes allow me to set up pre-made custom mixtures, which I find particularly useful with this color (squeeze into empty pans, stir with a toothpick, so I don't have to mix my favorite "Serpentine Genuine" look-alike each time I paint).
Winsor & Newton's Magnesium Brown PY119 is an extremely similar color, but has slightly finer particles resulting in a more subtle granulation texture.
This lighter orange-brown is very opaque in masstone, making it ideal for use in single-layer wet washes for texture effects. It does not lend itself to layering/glazing unless used very sparingly for texture with transparent colors like PY150 Nickel Azo Yellow and PB15:3 Phthalo Blue. PBr11 + PB15 is capable of making remarkably beautiful deep ocean teal blue-greens, as well as brown/blue combos with striking color separation similar to popular colors like Roman Szmal's Ocean Blue or Daniel Smith's Cascade Green.
13) PBk11 - Black Iron Oxide / Lunar Black / Mars Black / Magnetite. This granulating magnetic black is available from most brands but varies greatly in particle size. Some brands grind this pigment down very finely resulting in tiny dirt looking particles vs chunkier thick flakes. Lunar Black from Daniel Smith is a very chunky black with a tendency to be distracting in mixtures. It can be quite lovely when extreme texture is the effect you were after. Mars black in Winsor & Newton, Van Gogh or Rembrandt are all finer, which lends itself well to speckled fur patterns, stones/rocks/beaches as well as DIY replicating of the Van Gogh's "dusk" colors.
Commonly available from natural sources as well as being synthetically reproduced, all varieties can be moved with a strong magnet while in a wet wash. This allows for an odd ability to form the way the granulation is moving, such as into streaks and lines for fur. You may find the best control when using this texture trick by moving the magnet underneath the paper rather than from above.
PBk11 mixed with other colors may be a good way to replicate Hematite Genuine, Violet and Burnt Scarlet shades of the natural mineral paint by Daniel Smith.
14) PW6 Titanium White Gouache - I use Schmincke or M. Graham, but many major companies like Holbein, Winsor and Newton or DaVinci make opaque white gouache with extremely fine particles. These white paints are more opaque than white watercolor, but can be diluted to use in the same way. It's a quick and easy way to mix any watercolor into a pastel color, especially if you want a delicate opaque color for painting on black paper or to cover up other colors.
PW6 is often found in "mint" or "teal" paints that use PG7 (sometimes with PB15) to mimic the more costly Cobalt Teal. I personally use PW6 for highlights, as the final layer in my paintings.
M. Graham's 50ml (2oz) tubes of white gouache are huge, affordable and can be intermixed with any watercolor. It often costs far less than most other brands smaller tubes, so I recommend checking around for the best deal:
It can be time consuming to preserve white space in your watercolor art. It can also be consistently distracting to think about during the painting process. Some artists choose to use masking fluid (a liquid rubber latex that works similarly to rubber cement). It's expensive and prone to drying up in the bottle (limited shelf life is bad for those who use it infrequently). Some use white gel pens, which have a fixed line width and are prone to skipping/dry tips and sometimes unexpected blobs. I prefer using a tiny detail brush with white gouache for suggesting areas where the light is shining in my art. This particularly brings life to subjects meant to be glossy or wet, like water, the edges of shiny feathers or the reflection in the eyes of people and animals.
15) PV15 Ultramarine Violet OR PV62 Cobalt Violet HUE - Winsor and Newton has a great version of PV15 which is granulating and clearly defined as a purple (unlike some brand's blue-leaning versions that are too close to Ultramarine Blue, or even ones that are actually mixtures with PB29 like in Schmincke Horadam). Rembrandt's version seems to have some unusual color separation, slightly tinged with pink around the edges of a wash and slightly blue under salt reaction. Schmincke's Cobalt Violet Hue is a rare pigment very similar to PV15. I believe these may be interchangeable in mixtures and will be doing further testing on this speculation.
I found that using Winsor & Newton's version in larger areas with a lot of water really helped form pronounced granulation texture. I used it here in the sky with Cobalt Violet and Nickel Azo Yellow:
16) Manganese Blue Hue or Cerulean. Depending on your subject matter, you may find a pale sky blue color convenient to have on hand. This is a much more subtle, less staining option compared to PB15:3, ideal for those who use this color frequently in sky/landscapes/urban sketching.
It is not easy to mix unless you add a white pigment (which could lead to unwanted opacity). It has a softer, lighter valued appearance that is not quite accurately replicated by diluting blue with water. Holbein offers Horizon Blue, but you could just mix Phthalo Blue PB15:3 with Titanium White PW6 for a sky blue if a color being opaque does not bother you. MBH is a much lighter value alternative to Ultramarine Blue and more transparent than Cobalt Teal. Unfortunately, mixing PB29 and PG50 results in a dull desaturated color that does not match the vibrancy of Manganese Blue. Turner brand has the same vibrancy of DS, but unfortunately does not flow as well and lacks granulation.
I enjoy Manganese Blue Hue from Daniel Smith. It is made with a unique granulating and easily lifting (coarse particle) PB15 with a fluorescent additive. I realize that this is a controversial pick, but it is a unique and irreplaceable one, most likely appreciated by those who have had the opportunity to try the original Manganese Blue Genuine PB33. That pigment was quite toxic and stopped being mass manufactured around 2016, resulting in paint brands like D.Smith, Old Holland and Lukas discontinuing it. This caused a lot of sadness in the art community, as PB33 was considered a near perfect primary cyan blue with intense granulation. It was much less opaque than Cobalt Teal / Turquoise or Cerulean. This hue from Daniel Smith is the closest match to PB33 I've ever seen.
The unique colorless fluorescent chemical (invisible in normal lighting) causes it to double as a UV black light special effect paint. This glow effect spreads in water where the blue color does not reach, and fades independently of the blue. Fluorescent effects under black light are always fugitive (this should not be easily detected in normal daylight/regular lamplight). It will stop "glowing" blue under UV after too much direct sun exposure (as seen in my 1 year window lightfast test). The underlying PB15 blue, visible in normal daylight, is lightfast (won't be a problem for most artists). PB15 is LFII, so very subtle dulling of any phthalo blues always occur, regardless of any fluorescent additives. This appears similar to Cerulean, where fading/lightening does not occur, but the vibrancy may dull over very long periods of time. For MBH, the fluorescent effect for black light use requires protection from direct window light. In most indoor lighting situations (no consistent sunrise/sunset beams) Cerulean, PB15 and optical brighteners should remain unchanged for decades.
Daniel Smith Manganese Blue Hue very likely contains a special resin or Optical Brightener (also called a fluorescent whitening or brightening agent). This is a common colorless additive found in laundry detergent and printer paper. It results in bright whites, the illusion that makes yellow leaning items appear more white by refracting blue light. This has a similar glow effect as most brands of invisible ink and specialty black light products. It is not quite like the dyes I've seen in neon blue acrylic and gouache paints which utilize fluorescent blue colorants (often cyanine or naphthalamide dye in polymer and sometimes an optical brightener suspended in thermoset sulphonamide-melamine paraformaldeyde resin, which can be dried and pulverized into powder for use as a lake pigment) that usually fades visibly even in normal lighting. Most cyanine dyes and fluorecent dyes used with phthalo "neon blue" paints also have a fugitive blue undertone not present in this colorless clear/white optical brightener. Because most companies I emailed said that the fluorescent ingredient is proprietary, I can not 100% rule out additional trade-secret chemicals, or rarely used minerals like Fluorite (unlikely due to expense). Daniel Smith's paint is not advertised or promoted to be "fluorescent", which is unfortunate because it could be of interest to the UV/Black Light kids room or party market. The normally visible blue in daylight is not prone to fading, unlike many "fluorescent" or "neon" colors you may have heard always fade (which is more commonly a problem with Rhodamine dyes in pinks/violet colors like "opera"). Manganese Blue Hue has a very strong blue "glow" under black light (fluoresce blue light? or a resin that refracts UV-A?).
For science enthusiasts: Refract vs fluoresce? I probably misspoke in the video below (3:40) regarding what visually appears like light being reflected (a glow effect) from the optical brightener. The molecules may get excited under UV (black light), fluoresce (emit blue light) and be considered "fluorescing" instead of "refracting". Refract is a term for light changing direction, usually as it passes through something else. Perhaps the paint and I both got excited (lol), but the thought behind "refraction" here has to do with the resin particles involved (used in the making of fluorescent pigment powders). I asked one of the manufacturers of this type of UV effect paint about what was causing the glow. I was told that "the resin itself is refracting UV-A light". Kremer says the resin (either an optical brightener itself, or a binder for it) is hardened/dried, then pulverized into a powder like tiny clear crystals to be used in paint.
In addition to being valuable as a replacement for the old Manganese Blue PB33, this unique fluorescent phthalo mixture is what Daniel Smith uses to create Lunar Blue (mixed with PBk11). I believe a small amount of MBH is also included in Cascade Green (easily replicated with a touch of normal, finer particle PB15 and coarse PBr6/7 pale earth brown like Raw Umber). These colors can only be less accurately replicated with other versions of Phthalo Blue (no fluorescence, less texture and other Phthalo Blues have different flow characteristics). Manganese Blue Hue is also the closest way to replicate Blue Apatite Genuine, a naturally fluorescent mineral (sometimes glowing blue-violet under UV light when the fluorite content is high). A cerulean PB35 or 36 could be used with any black iron oxide PBk11 (Roman Szmal has particularly good options for both), but this combination will diminish the UV refracting glow and appear slightly more cloudy/opaque.
While there is debate on just how much "genuine" unaltered mineral is present in Daniel Smith's Primatek paints, that controversy does not overall change my recommendations. I tend to recommend mixing Primatek yourself with a more limited palette anyway. A handful of useful single pigment colors will be more affordable and versatile. I choose colors based on what helps me the most as an artist (unique versatility in mixtures, lightfastness and expense being important to me). I like the creative freedom of using single pigment paints and the joy of coming up with unusual mixtures myself. While beautiful, nearly all Primatek colors perform like multi-pigment mixtures (which is essentially what they are on a chemical level, since rocks are formed in sedimentary bands of multiple types of unfiltered minerals/different pigments). I feel capable of replicating them as needed, but if you paint the same type of subject frequently you may find that you repeatedly go through a specific mixture so often it's worth owning a "convenience color". If you were considering building a broad collection without knowing which color you'll really use the most, you may find that using the suggestions on this page instead may save you some money (while also allowing you to mix a larger variety of colors).
If budget is a concern, but you're really tempted by Primatek mineral paints, keep this in mind - Manganese Blue Hue, especially when accompanied by PBk11 Lunar Black, is very useful for replicating at least two or three other Daniel Smith paints. PBr11 mixed with affordable primary mixing colors (ones you likely to already own if you've bought any basic pan set) can result in the ability to replicate at least five additional Primatek mineral colors. If you don't already own a pan set of basic mixing colors, most of the top 10 colors in this list can be purchased individually for as low as $3 to $12 per 15ml (such as the super affordable White Nights, ShinHan, Holbein, Paul Rubens or even Mission Gold when sold in sets). Many of the Primatek mineral paints cost $15 to $30 per fluid 15ml from Daniel Smith (more if importing outside of the USA). I won't get into the complexity of comparing brands that only offer pans, but I will say Daniel Smith can suffer from excess water evaporation shrinkage, so despite the lower stated milliliters Roman Szmal pans are also a great deal. That being said, there are colors I feel are worth checking out from Daniel Smith, due to being truly unique, rare or particularly helpful. Manganese Blue Hue being one of them.
Other options: If the invisible fluorescent additive is undesirable, or Daniel Smith brand is just not a good fit for you, this palette pick could be replaced with the non-granulating Manganese Blue Hue by Winsor and Newton. That would be ideal if you just want a convenience sky color. For textural granulating/color separating mixtures, a Cerulean Blue PB35 (or or the PB36 chromium version if you like a darker value) could work as well. Roman Szmal and Winsor & Newton make very good options. Cobalt Teal (particularly the highly granulating DS version) mixed with a touch of Phthalo Blue PB15:3 will make a similar hue to Manganese Blue. Sadly it will not have the same type of granulation as DS MBH and is more prone to color separation in wet washes. Only Daniel Smith's brand glows in black light, but an alternative fluorescent medium is clear invisible ink, a good option for including secret black light effects or hidden messages that won't change the lightfastness of other colors.
This list is not yet complete. In no particular order here's some more pigments I've bookmarked as "must haves":
x) PV19 - Permanent Rose. This middle of the road cool red is often the top pick for artists sole primary mixing magenta/red. When picking multiple reds, as in a split primary system, it loses some of it's value in favor falling between the cooler PR122 and warmer PR209, PR254 or PR108. It's complicated though, because PV19 can actually vary dramatically by source. It can even be slightly less lightfast between brands. Each source varies slightly chemically, causing the hue to shift from a cooler purple leaning color to a warmer rose red. Figuring out which brand version will best fill the slot between the other 2 reds I've already chosen will be tough. I am in the process of doing a lightfast test on every brand to determine which has the least fading issues in the diluted range / in tints / mixtures. For now I am favoring Magenta PR122 instead.
PV19 mixed with PR206 (Brown Madder/Permanent Alizarin) is a great replacement for fugitive Alizarin PR83/Anthra Red PR177. I like Winsor and Newton's "Permanent Alizarin Crimson" (hue) as a good botanical/rose red.
x) Pbk26 - Spinel Grey by Rembrandt, a rare pigment. Smooth ink-like matte black that dilutes to gray with very little temperature bias. Ideal for when illustrators, graphic designers or abstract artists are looking for a true black. Ideal for monochromatic grayscale studies and representing pitch black areas where you might be inclined to use black gouache instead (a deep hole/chasm in the ground, space, a tear or rip to nothingness etc.). This fills an illustrator's need for a neutral and non-granulating black, performing much differently than PBk11 or PB29/PBr7 combos. PBk26 is also made by Maimeri as their Neutral Tint, but I felt like this brand was slightly weaker than the Rembrandt version.
x) Goethite. Available from Daniel Smith (PY43 tube) and Roman Szmal (labeled as "Pbr?" on their pan, and because they were uncertain how to classify I'm uncertain if it's truly just a form of PY43 chemically). Varies by batch from more yellow to more brown, but always makes a lovely beach sand color when diluted. Other versions of PY43 and PY42 Yellow Ochre are not as granulating as this "Goethite" color (which is labeled as "PBr? Natural Earth" on the Roman Szmal Aquarius label, but has been established as PY43). When mixed with blue it creates a lovely gently granulating succulent type pale green.
x) PY42 or PY43 Yellow Ochre variants, or other pale yellow-browns such as PBr7 Raw Sienna and Raw Umbers - Earthy yellows and browns are currently being tested for versatility. I need to test if Goethite replaces the need for some of these yellow-browns. Winsor and Newton's Raw Umber has a thick, heavy particle size that works well in color separating mixtures. Such as mixing with PB15 for replicating Daniel Smith's Cascade Green.
x) PO48 Quin Burnt Orange vs PBr7 Burnt Sienna - warm orange-browns. P048 is commonly mixed with PY150 for "Quin Gold" hues, a replacement for the long discontinued PO49 Quin Gold (genuine) pigment. I personally like the color separating effects in Daniel Smith's Undersea Green, which is a mixture using PY150, PO48 and PB29. I am currently testing if PO48 can be swapped out with other orange-browns for replicating Undersea Green.
x) PBr33 - Mahogany Brown only available from Schmincke, this intensely granulating brown has a unique color separation offering multiple hues in a single pigment paint. Ideal for complex neutral mixtures with other granulating pigments such as ultramarine blue. Daniel Smith's Environmentally Friendly Brown Iron Oxide also has dramatic granulation and may be a suitable alternative, with less color separation.
x) Up for debate = many Daniel Smith PrimaTek genuine mineral paints... single pigment ground minerals / gemstones with intense granulation and sometimes color separation. Purpurite, Blue or Green Apatite, Serpentine, Zoisite are lovely. While Purpurite is a beautiful stone, it is just the impure version of PV16 Manganese Violet with speckles of other trace minerals like Black Iron Oxide PBk11. It's pretty easily replicated, as are most of the PrimaTek colors. Mixing tests are currently being performed to determine which of these are the most versatile. I love Hematite for animal fur, but believe we may be able to replicate its effect with PBk11 mixtures. I have recently discovered color change/stability problems with tigers eye genuine and sodalite in storage (non UV issues). I have also determined it's easy to mix near-exact matches for many of Daniel Smith's colors using the other pigments we've already covered in this palette. You can see more about the issues I've found on the Daniel Smith lightfast testing and review page here.
x) -In testing- This is a helpful but optional pigment, as this particular type of PR101/102 can usually be replicated by mixing red and black iron oxides (PR101/PBk11). If one of these proves to be hard to replicate it will be the winner. For now here's my notes on PR101 Mars Brown or Caput Mortuum (PR102 in some brands):
Specifically the deep granulating red-brown version of red iron oxide. Often opaque to semi-opaque and has color separation in wet washes. Makes magic when mixed with other granulating pigments such as Ultra Blue or Ultra Violet. It's also a great way to tone down Phthalo Blue PB15:3 into stormy blue-grays. MaimeriBlu has a Mars Brown that is slightly more orange (vs a dull brick red Caput Mortum) with a lot of color separation in wet washes or with salt. Winsor and Newton and Sennlier both make excellent Caput Mortum. Not to be confused with english red, burnt sienna hues and other non-granulating versions of PR101. This pigment has a huge color variety between deep red-brown to terra cotta orange, it can also be manufactured as transparent to opaque. It's an inexpensive pigment that is often mixed with black iron oxide PBK11 to create "Burnt Umber" (a hue for the traditional PBr7 version).
PG50 (Green Version, different from Teal/Turq due to minor changes in manufacturing, heating and trace minerals) - Cobalt Green Light or Deep: I'm currently evaluating if this pigment can be swapped out for Viridian to leave granulating green texture in mixtures. Schmincke's new tundra and forest green mixtures with brown pigments seem particularly suited to animal art. I have used Lukas, Maimeri Blu, Winsor and Newton and Roman Szmal's Cobalt Greens - all are lovely granulating grass greens that work well for color separation in wet washes. I have had stability issues (browning/yellowing of binder) with PG19, Cobalt Green Pure from Daniel Smith, but Schmincke's version seems much better.
OPTIONAL ADDITIONS: The following colors are not necessary, but I have found myself drawn to them. These are beautiful, strong, lightfast colors that you may find useful. Each of these can be closely replicated by mixing a color already on my palette, but I could see adding them if you know you'll reach for them often enough.
Optional - mixable - PY128 or PY129 Green Gold is a beautiful color, but it can be easily replicated with PY150 and blue mixtures. If you're on a budget and already own PY150, try mixing a touch of any blue or green you have first.
Optional - mixable - PG7 Phthalo Green Blue Shade - a strong green that is non-granulating. Similar in hue to PG18 Viridian. It makes a lovely "mint" green when diluted with water or mixed with PW6 Titanium White. Some artists find this to be a great color for mixing neutrals, particularly with a cool red. The emerald-like green is often too unnatural looking for botanicals directly, but it mixes well with PY150. I more often choose to mix PB15:3 with PY150, which are also very strong/saturated/high chroma colors that can result in a very similar green.
Optional - mixable - PR206, PR179 - Deep rose reds, red-browns, perylene maroon and "Alizarin Crimson Permanent/Hues" from multiple companies attempting to replace the fugitive PR83 or PR177.
POPULAR OPTIONAL COLORS I PERSONALLY AVOID: I have seen these on other artists palettes, here's my notes about why I didn't include them.
Avoid - chemically unstable - PV14 Cobalt Violet (sometimes also called "light") - this pale pink-violet, highly granulating color is very unique. It's hard to replicate in any mixture and it pains me to remove it from my palette - but it is chemically unstable. While it often passes lightfast tests in northern and dry climates, art sent to those in humid sunny climates can fade. It appears to be either a heat or humidity/acidic/PH level change issue which only takes place in select environments.
Avoid - fugitive - PB27 Prussian Blue is rated as LFI only because when it fades it has a chance to recover to the original hue when left in shade for some time. It fails lightfast tests with fading up to LFIV and many brands never recover even after long term shade. I avoid this fugitive pigment (read more about how PB27 has a chemical reaction that occurs in any brand) but you might be interested in PB60 Indanthrene Blue as a deeper valued alternative to Phthalo Blues like PB15/16. Intense
Discretionary - opaque - Cadmium pigments such as red PR108, orange PO20 and yellow PY35 and PY37 have minor toxicity and are very opaque (limiting their usefulness in mixtures and layering). I do not reach for Cadmium colors often because they do not do anything that excites me personally, such as significant granulation texture or sinking in color separating mixtures (like opaque Cobalt colors do). Artists love them because they are strong, lightfast and have a lot of covering power for use as highlights and blocking in large areas of masstone. Gouache (in essence, opaque watercolors) may also be a non-toxic, more environmentally friendly alternative.
Discretionary - fugitive, but useful for realism in botanical art - Neon or Opera pink (PR122 combined with fluorescent dye). This is the only way to achieve some bright floral colors otherwise impossible to replicate with lightfast pigments. Fluorescent dyes don't scan well (cameras do not read the light reflection correctly) so you'd have to adjust it to bright pink in photoshop for prints. If prints is the goal, it may be better to do a more subtle traditional art piece that is lightfast, then up the contrast digitally after scanning before printing. If selling original art lighting advice should be passed on to the buyer, as only certain types of artificial light in rooms without windows (dark museum style lighting) would be suitable for displaying fugitive dyes. In rooms with large windows (with walls that receive beams of light at dawn/dusk for about an hour per day) neon/opera/fluorescent colors can start to visibly fade in as little as 3 months. I can see enjoying fluorescent colors for UV black light glow art, sketchbooks and personal projects. Though, Daniel Smith's Manganese Blue Hue is lightfast and also glows bright blue under UV light, so I'd be more inclined to use that for special black light glow projects that do not specifically have to be pink.
--- IN PROGRESS NOTE ---
As time goes on I'm adding more and more hand painted swatch cards to the pigment database which is helping me narrow down which brand makes my favorite version of a specific color. I'm also running lightfast tests on over 1,000 colors by dozens of companies at this time. As results become available, verified UV stable choices will continue to be added to this page. Fugitive color results are being added to the fugitive pigments list here.
Happy painting :)
---------- Tubes vs pans and other tips for beginners ---------
Tips for beginners starting their first professional palette on a budget: If you want to start off with the smallest number of useful colors, buying a primary mixing trio made up of Magenta, Yellow and Cyan will allow for countless combinations from just 3 colors. Ideally expand this to a group of 6 "split primary" colors which includes a warmer and cooler version of each primary color (a "cool" lemon yellow being closer to green on a color wheel compared to an orange-leaning warmer deep yellow, a cool phthalo/turquoise leaning blue, a deep ultramarine purple-leaning blue, a cool purple-leaning magenta that dilutes to pink as well as a warmer red that is capable of vibrant oranges in mixtures.) If you're new to all this, it may be easier to buy a pre-made set such as this split primary tube set by Daniel Smith and spend a little time playing with mixing your own colors.
I recommend researching color theory to understand mixing relationships. To avoid accidentally mixing "mud", it's good to know that complimentary colors (across from each other on a color wheel, such as red and green) will mix brown to near-black gray. That can actually be useful when creating "neutral" and shadow colors.
Buying tubes vs pans: Some brands only offer pans, some only tubes. When possible I'll list multiple brands that offer a good version of a specific pigment, so there may be a choice to be made on your part. For the type of palette set up I'm using, it is ideal to buy paints in tubes instead of pans - especially for colors you see listed as a base ingredient in multiple of your favorite mixtures. In the case of PY150 Nickel Azo Yellow, this pigment is present as an ingredient in a huge variety of pre-made "green" paints on the market. Having a tube allows me to pre-mix my most-used colors into pan trays (use a toothpick or needle to stir). Tubes containing 10-15ml cost more upfront, but are a better deal over time. Not only do you get more paint for the money, but it can save you from buying extra pre-made pans of those go-to convenience mixtures.
You will find that you use up less paint at a time and have better control while painting if you allow your tube paints to dry in a pan before re-wetting them with a wet brush later. Pre mixing your frequently used colors can save you time if you find, for instance, that you are constantly mixing a purple. You can mix a cool red (like PR122) and blue (PB15 or PB29) together for your perfect custom hue of pre-made purple with your chosen amount of pink or blue hue.
I've seen people recommend pans so that you can afford to collect more colors. This is OK advice when applied to single pigment paint, but it may cause you to buy extra pans of pre-made mixtures and/or cause more work to create complex mixtures like Moonglow. When using dry pans, you have to re-do your favorite mixtures in a palette with a brush each time you paint. You won't need as many pre-made brand mixtures like "Moonglow" (often called "convenience colors") in the long run if you can mix your liquid paint into your own pans. If you aren't sure how dedicated you'll be to watercolor painting, it would definitely be cheaper in the short term to buy a pre-made pan set (such as a 24 to 48 assortment by Paul Rubens on Amazon).
If you do not enjoy granulation (texture and color separation effects) you may still find some good options within this list. Some pigments have smooth or rough texture choices. This is the case in pigments such as PBK11 Daniel Smith's Lunar Black (thick particle) vs most other brand's Mars/Black Iron oxide (fine), PR101 Red Iron Oxide/English Red are often smooth while Caput Mortum can be granulating as well as color separating, and most commonly - the widely variable PB29 Ultramarine Blue. Each pigment code (ingredient number) can vary from extremely granulating to only subtle minor texture depending on how fine the pigment powder was ground up before being turned into paint. This is particularly important to know for choosing your primary Ultramarine Blue - such as Schmincke's "French Ultramarine" heavily granulating vs "Ultramarine Finest" with very minor texture.
If you love granulation texture you can encourage it the most by using very wet washes, allowing the pigment to flow over the surface of water. This works best on textured paper like cold press or rough press (not smooth hot press) where the pigment particles can settle into the tiny dips and valleys of the paper surface.
Where do I shop for art supplies?
Rosa Gallery brand from the Ukraine is available from Spain at Artmiranda, for USA check Etsy (OmniaPro shop) and on rare occasions they list sets on Amazon. Most other companies can be found at Blick (USA) or Jacksons (UK and worldwide) below.
Jackson's has the best price on Schmincke and MaimeriBlu. Blick has US brands like Da Vinci and Daniel Smith. Both have great prices on Sennelier, M. Graham and watercolor paper (Arches or Winsor and Newton Cold Press is great!) and brushes (Princeton Heritage for point/spring, Neptune for holding a lot of water). Winsor and Newton has trade agreements preventing international shipping, so get them at Blick if you're in the USA.
My favorite American art supply chain store is Blick Art Materials. They have a massive catalog and competitive prices, with quick shipping options here in the USA. Best major retailer for Da Vinci.
One of my favorite places to shop for a world-wide selection watercolor paint and brushes is Jackson's. They have affordable shipping to the USA and a lovely selection of items not easily found in American stores. Only major retailer for Roman Szmal.
Amazon USA continues to offer more and more art and craft supplies that can be found no where else. They often have import sets, such as Chinese brands like Paul Rubens, that are not available in the more common art stores. This page contains affiliate links. As an Amazon associate, I earn from qualifying purchases.
I use affiliate links. As an Amazon associate I earn from qualifying purchases. Thank you!