Lightfastness Tests 2023

Since we’re deep in the darkest part of the year, I figured I’d take down my lightfastness tests. (To be honest, I forgot about them; I meant to only do them for 3 months but some of them have been up for over 6.) Here is what I learned.

Method

First, I painted a swatch of each paint on a small card, marking both the left and right sides with the name, brand, and pigment number.

Lightfastness test swatches, before cutting in half.

I then cut the swatch in half, and stashed one half in an envelope while posting the other half in a western-facing window with direct afternoon light. I attached all the window swatches to a backboard for ease of use (and to avoid putting tape on the actual glass of my window.)

Lightfastness swatches, attached to backboard.

I live in the northeast United States, the Boston area, so everything was done from about 42 north latitude. I got these started in the summer so there would be some days of intense light.

Sometimes I posted the left half of the swatch and sometimes the right. I did this intentionally so there would be a randomized mix (in case I always paint the left half a of swatch darker or something), but tbh I wish I had just systematized it. Comparing the swatches side-by-side is easier if a reliable side is always the “sun” side and they go together like puzzle pieces.

I tested a fairly random selection of paints. There are still some I wish I had done, but there’s always next year!

Highest Performers: No Noticeable Change

Strong Performers: Slight/Subtle/Nearly Imperceptible Change

Iffy Performers: Noticeable Change

Conclusions

Biggest Surprise

Deep Scarlet faded a lot for a supposedly LFI color. Lies, Daniel Smith! All lies!

Biggest Bummer

Which Quin Magenta and Red Rose Deep didn’t fade as much as fugitive Opus, both noticeably changed. There is no such thing as a perfect magenta.

Also, I thought PR101 would perform perfectly, so I was surprised by the fairly consistent subtle muting. Indian Red did better than Transparent Red Oxide.

Pleasant Surprises

PY74 and PR264 performed better than I expected.

Overall

I’m still not convinced that lightfastness matters to me, so maybe this was a waste of time. Still, I feel I’ve learned a few things:

  • Fewer colors than I expected can withstand 4-6 months of light with no change. It’ll be easier to find colors if you decide not to care about the subtle changes.
  • It’s relatively easy to find yellows and blues that don’t fade much, and relatively difficult to find reds.
  • Don’t necessarily believe printed lightfastness ratings. (Especially from Daniel Smith.)
  • To be fair, lightfastness may also vary from batch to batch… but if that’s so, I definitely need to stop caring, because it seems like A Lot to go through a three-to-six-month process before cracking into a new tube.

Did my results surprise you? Do you care about lightfastness?

2 thoughts on “Lightfastness Tests 2023”

  1. I do care, but not about changes as minor as your subtle changes. I can’t even see most of them, and sometimes I think the exposed strip looks better! (Although that Holbein PR264 is pretty blatant, and annoying…)

Comments are closed.