The lightfastness tests for my watercolours were finished some time ago, but though I removed the charts from the windows, I was unable to scan them because Windows 8 decided it didn’t want to recognise my scanner anymore, and I never got around to buying a new one. I scanned them the other day on a scanner at uni and today my old man bought us a scanner for home, so I can now wrap up my watercolour lightfastness tests.
Note: In addition to some annoying lines through the swatches left by the uni scanner, the colours have come out a lot more saturated than they did on my older scans and in some cases diluted areas look much more faded than they actually are, whereas some that I scanned on our new scanner have also slightly changed the hue and saturation in some swatches. However the fading (where it has occurred) is still easy enough to see from the horizontal lines through the middle of the swatches, so the scans can still serve their purpose. You can also see Part 1 (test start) here, Part 2 (4 months) here and Part 3 (8 months) here. (if you want an idea of what the paint colours look like, I’d recommend looking at these older posts, as those charts reflect what the paints look like in real life much more accurately than the scans I’ll be including in this post). Once I play with our new scanner’s settings I’ll update this post with better quality scans (at this stage the yellows, reds and violets are from the uni scanner and the blues, greens, earths and shades are from our home scanner).
Just to recap technical information for these tests:
AS = Art Spectrum
DR = Daler Rowney
DS = Daniel Smith
HWC = Holbein
LUK = Lukas
MB = Maimeri Blu
MG = M Graham
OH = Old Holland
REM = Rembrandt
SCH = Schmincke
SEN = Sennelier
WN = Winsor & Newton
Paper used: Daler Rowney Aquafine
Winsor & Newton’s Indian Yellow has faded in the washed areas (more noticeable on the version with PO62 in it), but not where the colour was used at full strength. Daler Rowney New Gamboge and Cadmium Yellow Deep Hue also suffered very slight fading. Lukas Permanent Yellow Light, which had already faded slightly at the 8 month mark, has faded even further, even in the areas where the colour was more concentrated. Sennelier and Schmincke’s Lemon Yellows have both faded noticeably. Lukas Naples Yellow also shows very slight fading across both diluted and concentrated areas, though it is barely visible to the naked eye. All other yellows remain unchanged.
Very few colours on the red and orange chart remain unaffected. Most noticeable (not surprisingly) are the genuine Alizarin Crimsons from all three brands (Lukas, Sennelier and Winsor & Newton) have faded drastically since the 8 month scans, especially the SEN and WN, which have almost completely disappeared even in the concentrated areas. WN’s Permanent Alizarin Crimson has also faded, though not to the extent of the genuine pigments. Art Spectrum’s Coral has faded a little more, as have Lukas Permanent Red and Permanent Orange, Old Holland’s Golden Barok Red and WN’s Winsor Red Deep, Quinacridone Red, Scarlet Lake and Bright Red. The cadmium reds and most of the oranges escaped unharmed.
The only really noticeable fading to occur in this lot is with Rembrandt’s Quinacridone Rose. WN’s Perylene Violet has lost a little of its intensity , though saturation remains the same (this is visible to the naked eye but does not show up as well in the scan). There has been slight fading in some other colours, including WN Permanent Rose and Opera Rose, Schmincke’s Mauve, Lukas Purple and Dioxazine Purple, and Art Spectrum’s Flinders Red Violet. The other colours have not faded at all.
Most of the blues have suffered no fading throughout the test. WN’s Smalt Blue has faded very slightly, but this is only visible to the naked eye; it does not show up in the scan. AS Tasman Blue and Lukas Prussian Blue have also faded a little, while Lukas Indigo has faded significantly, losing some of its ‘reddishness’ (it seems that the red PR176 has faded from the paint, leaving only the blue and black pigments). Otherwise the blues have pretty much stayed the same.
Both versions of WN’s Olive Green have faded a little, as has Sennelier’s Hooker’s Green. Lukas Cinnabar Green Light, Permanent Green and Olive Green have also faded very slightly, though it doesn’t show up well in the scan. No change in the other colours.
Though AS Australian Red Gold had faded a little at the 8 month mark, it doesn’t seem to have got any worse. However, several of the Lukas browns – Burnt Green Earth, Burnt Umber and Raw Umber – have faded significantly, and like the Lukas Indigo, all of these colours contain pigment PR176 (which seems to be a variant of Alizarin Crimson).
No change to any colours.
So, that concludes my watercolour lightfastness testing. At some point if I get more colours, I might do tests for them and add them to these posts, but I don’t think I need to buy anymore paints (I have far too many already). The main thing to take away is that it pays to check what pigments are in your paints, especially if you’re going to use them in art that will be sold or displayed. Even some colours that have the same name across different brands may have different pigment combinations (depending on when your paint was made, even the same colour by the same maker may have a different makeup). For example, why they included such a fugitive pigment in these browns – or why they are mixes at all, given there are perfectly good single-pigment alternatives for these colours – is beyond me. Not only that, but sometimes one version of the same pigment may be lightfast while another isn’t (eg. some of my PV19 paints faded while others didn’t change at all), so if you can, do your own lightfastness tests on any colours you’re not sure about. If it turns out some of your paints are fugitive, not to worry; they can still be used for art journals and sketch book work or projects that will be scanned and and distributed digitally.