MITRA Forum Question Details

It was a common practice in the 17th-18th century to have just such a double ground. In the 1600s it was often done out of economy. The rough texture of the canvas would be filled with a very affordable red or brown earth. The gray or white layer containing the much more expensive lead white pigment would only need to be thick enough to obscure the colored ground below.

It was also common in the 18th century to have a ground just as you describe where the indirect effect would yield a cool smokey appearance to the surface due to the turbid medium effect. It is also true that paints containing lead and zinc whites become more transparent over time due to increased transparency resulting from metal soap formation.

As to your current situation, I think that it is fine to add additional layers to achieve your intended effect. It is true that in general, the more layers, the more likelihood of potential problems. Having written that, I think your plan to add some titanium white to an additional layer would be sound. You may have to adjust the mixture as titanium white is both far more opaque but it is also a much colder white.

There may be some darkening overtime, but this may not be a major problem, especially if you paint the paint proper with some body. Much of the problem encountered in past is that the brightest highlights remained white but thinly executed mid tones were lost due to the increased transparency.

It's helpful in working over an underpainting like this to add a medium to whites and flesh passages that renders them resistant to staining and quick to dry.  Doing so helps amplify light colors and increases contrast. This practice can be seen in the work of Artemisia Gentileschi, where some pentimenti from the initial work can be barely seen, next to figures and drapery which are brilliant and prominent. Modern painters can use one of the alkyd-based mediums to achieve this resistance, which earlier artists achieved with semi-fossil resin varnishes in cooked oil.​

17 Rembrandt Canvas Gray Ground.jpg

I have always found that when oil priming if you use a heavy cream-like consistency gently applied with a spatula you have a far more covering surface than when applying a paste-like consistency of oil ground. Back in art school I would often prime canvas and linen with heavy-paste lead white unthinned using a long priming knife. I saw the same issue with the heads of the threads remaining relatively uncovered even after multiple coats. This is far less the case when doing the above.

There is no doubt that filling the interstices of the fabric with a fabric like color and then priming with a white or off-white priming like I describe creates less of the issue you describe. I do believe, though, that doing the same with cream-like primings using only white lead would do the same, it just would have been more expensive.

I have included an image of a double ground that I reconstructed for a project about dutch strainers and Rembrandt's painting technique. You see that there is no strike though and the top ground is quite opaque. This is an emulation of the earlier, 17th century ground practice and not the smokey indirect effect that the OP is talking about.