The Great Fabric Medium Experiment, Part IV

I finally got some time to finish up my patch tests. Instead of buying another new T-shirt just to paint up, I pulled an old, bleach-stained blue T-shirt from the rag pile. This means that as colors become less opaque, the perceived color will shift towards the blue, rather than towards the red. I shouldn't see an issue with more opaque colors, and the increasing thickness of the paints in my test suggested that my colors would be more opaque.

The last three paints

As you can see, that was not quite the case. While there was little color shift with the dark Martha Stewart Multi-Surface Satin Acrylic Craft Paint in Chipotle, both the Folk Art Acrylic paint in Wicker White and the Liquitex Acrylic Paint in Light Portrait Pink showed significant opacity issues with fabric-medium dilution. Also, the Folk Art paint — which I had expected to be a better quality craft acrylic paint and more opaque than the Martha Stewart paint — showed the same opacity issues as the other acrylic paints. (This might be an issue of the particular pigment: Folk Art also offers a Titanium White, which is considered to be a more-opaque white pigment.)

Also noteworthy: the Folk Art paint was also subject to the same sort of bleeding exhibited by the most fluid paints.

Opacity Issues, Random Controls

The blue shirt also has a couple of control swatches: on the top left, I have Americana Shading Flesh by itself, and Golden High Flow Acrylic Cerulean Blue Hue both by itself and diluted 1:1 with GAC-900. The difference between these color swatches and those on the red shirt suggest that even apparently opaque paints are subject to color shift on colored fabrics. There's also an open square of GAC-900 by itself, It has discolored the fabric, but I'm looking to see how it will work as either a surface prep (like gesso) or a resist. Similarly, there's a closed square of Delta Ceramcoat Textile Medium on the right side, along with a rectangle of Fabric Creations Soft Fabric Ink in Yellow (which is not nearly as opaque as it had appeared to be when creating my "Star Struck Person With  Headscarf" portmanteauji T-shirt).

Preliminary Observations

Before heat treating, my impression is that for any non-fabric-specific acrylic paint thicker than Golden High Flow Acrylic, the paint swatches are very stiff until we get to a 1:1 paint-to-medium ratio. Unfortunately, this is the point at which pigment dilution issues start to become obvious. Also, since all of the fabric mediums were thinner than all but the Liquitex  Professional Acrylic Ink and the Golden High Flow Acrylic, the lower the paint-to-medium ratio, the higher the likelihood of bleeding.

Because of pigment dilution, a more concentrated pigment is needed for better results on non-white surfaces. This suggests that our usual method of painting T-shirts with craft paints is probably the least effective method for getting a good result.

The samples were done with a single coat of paint, with a second layer applied to the upper half of the Apple Barrel and Americana control samples; a second coat may be necessary (but may further stiffen the textile).

Next step, heat treatment.

The Great Fabric Medium Experiment, Part III

I decided to use the back of the red t-shirt to run the next few sets of tests. These used different brands of craft acrylic paint, which is about the same consistency of Golden Fluid or Liquitex Soft Body acrylic paints, but less expensive, and is the general paint of choice for fabric painting.

Closeup for bleeding check

This second set of tests covered Craft Smart Multi Surface Premium Satin Acrylic Paint in Rich Lilac, Apple Barrel All-Purpose Craft Paint in Victorian Green (a discontinued color), and Americana Acrylic Paint in Shading Flesh (also a discontinued color). Going into this set of paints, I found less bleeding with the GAC-900 than I did with the lighter-weight artist colors. On the other hand, the opacity of these craft paints started out pretty low, and only got worse as the paint dried.

The camera enhanced the opacity of the paints.
The true appearance of the controls is closer to that of the 1:2 paint:medium ratio.

You can see from the photo (but not barely on the garment itself) that I ran a second control paint layer over half of each of the Apple Barrel and Americana tests.

Liquitex mentions on its web site that its Basics line of acrylic artist paints is not as pigment-dense as its Professional line of paints. It's not a far stretch to assume that, at about 10% of the price of the professional soft/fluid acrylics, craft paints are even less pigment-dense.

I'm going to need to move to a second T-shirt to finish testing my craft paints (Folk Art and Martha Stewart) and move to my very old Liquitex Acrylic Paint (which, in the absence of a product-name modifier, is probably closer to their Basics paint than their Professional Heavy Body paint). I'm also curious to know how brands and formulations from the same manufacturer vary in response, so I will probably run a control of similar colors of Apple Barrel, Folk Art, and Martha Stewart paints (all from Plaid), and I may need to purchase a small tube of Liquitex Heavy Body and a tube of Golden in one of my existing colors of those brands to find out.

The Great Fabric Medium Experiment, Part II

Now that I'd decided to do a "scientific" experiment on the choice of paints and fabric mediums for t-shirt painting, I had to gather my supplies. Searching "fabric medium" on Michaels' website came up with three relevant products: Liquitex Fabric Medium, Americana Fabric Painting Medium, and Delta Ceramcoat Textile Medium. While we only sell the Liquitex product online, we do sell Golden GAC-900 and Martha Stewart Crafts Tintable Fabric Medium — both of which did not show up in the search — in-store. This gave me four fabric medium products to test.

Fabric Mediums

My craft paint stash includes Apple Barrel, Americana, Folk Art, and Martha Stewart Crafts acrylic craft paints as well as a few old tubes of "Liquitex Acrylic Artist Color" and a couple of colors of Fabric Creations Soft Fabric Ink (which does not require the addition of fabric medium).

Craft Paints

Old Liquitex Acrylic Paints

I added to this a bottle of Liquitex Professional Acrylic Ink in Cadmium Yellow Light Hue, a bottle of Liquitex Soft Body Acrylic Paint in Cadmium Orange Hue, a bottle of Golden High Flow paint in Cerulean Blue Hue, and a bottle of Golden Fluid Acrylic Paint in Teal. Based on Golden's website and the samples on the bottles, I expect both Golden paints to be relatively opaque.

Thinner-Bodied Artist Paints and Inks

Because I'm looking at doing this in our classroom during off time, I'll also have access to the CraftSmart acrylic paints and Tulip Soft Matte fabric paints leftover from previous events. While I should probably add our house-brand Artist Loft acrylic paints to the mix, I think I have enough variety — four mediums at five paint-to-medium ratios times nine or ten paints, plus a control (no fabric medium) for each... we're talking over a hundred samples, and possibly over 200. The samples need to be at least an inch square, if not two inches square, to have enough material to bend to check flexibility and tendency to crack. Because of the number of samples I'm going to need, I'm not going to heat treat until all my samples are done, and I'm going to wait the full week after that until laundering them.

I brought my materials into work a couple of hours early on Monday, so I could start work on the project. After about an hour and a half of solid painting work, I'm four paints into the project, and the entire front of one T-shirt.

First four - a bit overexposed to show the Liquitex Ink (top set)

I'm debating whether to use the back of that shirt for my next set of samples, or if I should start a new shirt so I can hang the patches side by side.

Meanwhile, it's become apparent that (at least among the thinner paints I've tested so far) paint-to-medium ratios of less than 1:1, produce is some bleeding. In general, the thinner the paint, the greater the bleeding — but what surprised me is how much more bleeding there was with the GAC-900 — which is supposed to be the highest quality of the fabric mediums I used — compared to the other fabric mediums.

GAC-900 is the top row of each paint set. Control at far left; ratios go from 3:1 on the left to 1:3 on the right

I want to see if the bleeding reduces as I test the heavier paints. It may be that the paint-to-medium ratios, and the preferred fabric mediums to use, may change with the viscosity and pigmentation of the paints used. It's also apparent that the dilution effect of fabric medium decreases the opacity of the paints  —  and that some fabric mediums are worse than others. Again, this may change with both the quality and thickness of paint as well as the individual fabric mediums. Since I've only gotten through about 40% of my test slate, the answers are still waiting to be found.

Pigments, opacities, bleeding, and resist

Spoiler alert: I can't resist playing with "new toys", even if I'm not ready to run an "official" test with them.

---

When I was a little girl, my father used to match inks for a living.

Back in the days before process-printing (where all colors come from the overprinting of between four and eight mostly-transparent inks) was king, a lot of color printing was done color by color, using inks formulated from the same pigments used by artists. It was Dad's job to look at the colors of original artwork and determine what combinations of pigments would be needed to match each color exactly. There could be some variation in the mixture depending on both ambient and processing conditions (heat, humidity, pressure), but the result had to be the same. Each color for each project had notes, in case a second print run was needed, so that the colors of the new printing would exactly match those of the old printing.

That sort of work required knowledge of pigment materials and colors, some knowledge of chemistry, and a very keen eye for color — which he did his best to pass on to me. For most of my life, though, my internalization of color has been relative, not pigment-based. It wasn't until I started working at Michaels that my childhood memories of pigments came back to me.

Today, a number of paint manufacturers sell products with the old pigment names followed by the word "hue" — for example, Alazarin Crimson Hue. As opposed to the HSB digital colorspace (where "hue" refers to redness, blueness, etc.), "hue" in a pigment name means that the color was manufactured to come as close as possible to that of the original pigment, but either avoiding toxic materials, or at a cheaper price point, or both. (While there are differences in the behaviors of real and "hue" pigments, I also remember Dad telling me how many of his colleagues died in their thirties because of exposure to those chemicals.)

---

My first new "paint" acquisition for the Great Fabric Medium Experiment was Liquitex Professional Acrylic Ink in Cadmium Yellow Light Hue. I've been wanting to play with artist acrylic ink since finding out that Speedball non-black calligraphy inks are acrylic-based, and I can apparently use these inks in dip pens as well as with brushes. I chose the yellow because it was bright and light; I wanted something that should show up on both a dark shirt and a white shirt. What I didn't know is that Liquitex Cadmium Yellow Light Hue Ink is a transparent (see-through) color, meaning that some of the color of the medium (or underlying paint layers) will show through. A dropper full of well-shaken ink all but disappeared into the red shirt scrap I'd been using for small-scale testing for my Michaels app T-shirt. [I warned you I can't resist playing with new toys!] I learned from Golden's web site that transparent pigments are better for mixing and blending than opaque pigments. Both Liquitex and Golden publish charts that rate their color products for transparency, lightfastness, and permanence.

Another issue I'd run into in my recent T-shirt projects was bleeding. Basically, the thinner a fluid, the further a drop of it will spread — and the more absorbent the surface it is dropped upon, the more it will seep into the surface (rather than rest on top of it). This is one of the reasons that artists cover their canvases with gesso (and let it dry) before putting paint to canvas. (Note: prepared canvases, like the stretched canvases we sell over at Michaels, have already been primed with gesso and are paint-ready.) Acrylic paints in general, and craft paints in particular, are usually thinned with water to create the correct consistency for their intended use. If a paint has settled and has not been sufficiently shaken up, or if a paint has been thinned too much for the surface, it will either run (drip down an upright or angled surface) or bleed (absorb into the surface and spread beyond the borders of the brush stroke). One way to stop bleeding is to set a boundary beyond which the paint (or dye) is not allowed to spread. This boundary is called a resist.

The dark halos around the lighter paint patches indicate bleeding.
The light halo around the paint patch on the right is where the paint has encountered a test resist.

If you've ever done tie-dyeing, you understand the basics of resist. If you've ever used stencils with paint, you understand the basics of resist. Batik, another fabric-coloring technique, uses wax as a resist. Since dried, bled-in acrylic paint seems to serve as a resist for an overlaid color of paint, I'm wondering if fabric medium can do the same. Also, we sell a lot of Elmer's Washable School Glue. I wonder if that can be used as a resist medium...

While I don't have a formal experiment planned for testing those two fluids as resist mediums, I might just add a couple of test patches along the way.