When making some of the works from the γ-series, I noticed, or seemed to notice, that the watercolour was retained in the brush to a greater extent when using real sable hairs as opposed to synthetic materials. As this can influence the amount of control when has on the introduction of the watercolour to the droplet, I decided to test a number of different brushes.
For this experiment, a single batch was made of an unspecified amount of Royal Talens Rembrandt branded Quinacridone Violet (593) dissolved in 0,5 mL of distilled water.
Using a micropipette, a 5 µL droplet of distilled water was placed on Hahnemühle 290gms Agave watercolour paper.
Each brush was then dipped into the watercolour solution, rinsed two times by dipping it into distilled water and then placed into the droplet, as vertical as possible.
The action of adding the watercolour to the droplet was recorded by video. As the droplet had the tendency to move towards the brush, the first frame is recorded when the first movement of the droplet is observed, together with the frames 0,1 and 1 second after after this initial movement.
The following brushes were tested: Da Vinci Colineo, size 2/0; Da Vinci Fit Synthetics, size 0; Da Vinci Forte Synthetics, size 3/0; Escoda Perla Sintético, size 3/0; Gerstaecker, size 1; Winsor & Newton Synthetic Sable Round, size 00; Winsor & Newton Cotman 111 Round, size 00; Raphaël Martora Red Sable, size 3/0.
The experiment was repeated with Royal Talens Rembrandt branded Chromium Oxide Green (668). This colour consists of a simple inorganic pigment in the form of chromium(III) oxide, as opposed to the aromatic quinacridone pigment found in Quinacridone Violet.
This gave the following results:
 |
| Da Vinci Colineo, size 2/0 |
 |
| Da Vinci Fit Synthetics, size 0 |
 |
| Da Vinci Forte Synthetics, size 3/0 |
 |
| Escoda Perla Sintético, size 3/0 |
 |
| Gerstaecker, size 1 |
 |
| Winsor & Newton Synthetic Sable Round, size 00 |
 |
| Winsor & Newton Cotman 111 Round, size 00 |
 |
| Raphaël Martora Red Sable, size 3/0 |
An overview of all these tests gave the impression that the chromium oxide had more of a tendency to leave the brush than the quinacridone pigmented watercolour. The amounts also seemed to vary from brush to brush, with the additional remark that brushes of the same brand seemed to show similar diffusion.
Of these brushes the Escoda Perla Sintético brush had the least diffusion into the droplet. As this was also the most previously used brush, I wondered if this was perhaps related to the amount of use this particular brush has had. In order to find out, I bought an identical brush and redid the above test with both the new and the old brush:
 |
| Escoda Perla Sintético, size 3/0, used (top) and new (bottom) |
In addition to the Quinacridone Violet and Chromium(III) oxide, I also used Royal Talens Permanent Lemon Yellow (254), which consist of bismuth vanadate.
There was very little difference between the new and the old brush. Perhaps the older brush had slightly more diffusion than the newer brush, but this can't be said with any certainty.
Another noteworthy observation is that the Quinacridone Violet didn't seem to dissolve into the droplet at all from the Gerstaecker branded brush.
Additionally, the only brush that contains the much-coveted real sable hair, the Raphaël Martora Red Sable, has a very different diffusion pattern from all the other brushes. The hairs of the brush spread far more easily than those in any of other, synthetic, brushers.
Related to this is the observation that the water in the droplet is very much attracted to the water present in the brush. This attraction is so strong that with a minimal amount of water, the droplet is almost 'sucking' the water out of the brush:
