Old Silver - First Sketches
Continuing the theme of planning I dealt with in the last post, this post covers the early stages of the planning of a new painting.
The subject of this painting is a collection of old silver objects. I know, silver again. But I like to paint silver, especially old silver. Apart from it being very pretty, it represents quite a difficult challenge in terms of the values. For some time now, I've been concentrating on values, and I expect to do so for some time to come. I think it's important.
It's a fundamental truth of painting that the value range that we have available to us with our materials is limited when compared to the range of values that we see in nature. The lightest light we can get is white, say a titanium white. But that's still not a true white in value terms, it's slightly darker. Even if it was a true white, it's still not as light as the reflections of direct light off a shiny surface like silver. That's basic physics. Paint doesn't reflect as much light as silver. That's because it's not as reflective. (I worked that out for myself, you know.)
If you're faced with the problem of making a convincing painting of a shiny object, and you know that you can't match the lightest light in the subject, you're necessarily forced to work with a compressed value range. But how to deal with that problem? Make everything a couple of steps darker? Then what happens when you get to the bottom of your value range, the dark end? You'll run out of range at this end instead of at the top, exchanging a glass ceiling on your lights for a glass floor, so to speak, on your darks. I've tried it, it doesn't work.
The way I've been trying to deal with this when I paint silver is to try to ensure that I don't let any lights get too close to the highlights on the silver. There has to be a noticeable step in tone between the highlights and the next lightest tone in order for them to be sufficiently light to read as reflected light. It's all in the relationships.
Although I understand this from an intellectual point of view, making it work in practice is another thing entirely. So for this painting, I thought it might be an idea to explore this problem a little before I started getting into planning the composition. The original idea for this painting was that the silver objects - a sugar bowl, butter knife and a collection of spoons - would be resting on a white cloth, seen from above.
The white cloth compounds the problem somewhat, because the lightest light on the cloth must be some way below the value of the highlights on the silver. It could easily end up looking like a grey cloth, if the relationships between the rest of the values aren't carefully handled.
This is the butter knife that will be in the final painting. The white cloth kind of reads alright as white, I think, but the highlights on the knife aren't exactly jumping out.
This sketch was a struggle. Although I could get the value of the shadow area of the cloth working ok, I fought with the colour of the shadow for the best part of a day. Was it warmer or cooler?. Brownish or blueish? In the end I decided it was mostly grey, but slightly pinkish in places. Either way, I wasn't happy with this sketch and started to revise the idea of the white cloth. Perhaps a darker cloth would work better in the painting.
A cop out? Well, maybe. But this painting is going to give lots of value challenges in any case, and I do want it to be an attractive painting. I wasn't convinced that it would be with this white cloth. I don't want it to look drab and boring.
These two spoons are also going to be part of the supporting cast in this painting. I particularly like the old soup spoon.
From a point of view of the tones, this darker cloth is going to work better. The silver stands out more clearly against it, and it presents me with less value problems.
I wasn't all that happy with this sketch either though. The values haven't been handled all that well, and the highlights are plain white, when in reality they are slightly yellow. Even looking at it now, I find it hard to put my finger on exactly what the problem is with the values, but I know there is one. It's been my experience that when you get the values close to right, the objects seem to come to life and jump off the surface of the painting with a certain physicality that no amount of detail can replace. The light looks convincing. That's not happening here. It could just be that I haven't painted this very well, but I have a feeling that the value of the light area of the green cloth is wrong. Too dark in relation to the spoons, perhaps.
Things started to improve a bit on the next batch. Each of these sketches took a day or so, due to more care being taken with the judging of the values.
A big help here was the new grey scale I've just made. The grey scale I've been using up to this point is (I think) a printer's grey scale. It's been bothering me for some time that the lighter end of the scale seems to be compressed, with smaller steps between the tones in comparison to the dark end.
Recently I got hold of the New Munsell Student Colour Set, (I'll go into more detail on it in another post), which includes a set of chips representing a more evenly spaced scale of neutral values. I've based my new grey scale on these values, and it seems to be more practical in use.
The first teaspoon here is on the same green cloth as the two in the last sketch, but I've got the value of teh cloth much closer I think. This teaspoon seems to live more on the canvas than the last two did. The light is more convincing to my eye. But still, the highlights aren't really working convincingly. There's not enough distance in value terms between the highlights and the next nearest light values, I think. Also, these old silver objects have a slightly green tinge, and the spoon is getting a bit lost on the green cloth. Thus the red cloth used for the next two sketches.
This red is by far the brightest colour I've had on my palette for months. Since I've been dealing with value as much as possible for some time now, I've kept my colours deliberately reserved. I half expected this red to give me trouble, but it didn't. I believe that's because I spent a lot of time making sure I had it's value right, using my Munsell scale, before I thought about the chroma or the hue. If the value is right, everything else falls fairly naturally into place, I think.
I like this red against the greenish greys of the silver. It sets them off nicely. I think the values are coming better here too, particularly on the sugar bowl. The highlights are working better.
An interesting point here is the difference in hue between the two reds. This is the same cloth, painted on different days. On the sugar bowl, the red is warmer, on the soup spoon, it's distinctly more blue. At first I thought I must have judged it incorrectly on the sugar bowl sketch, although I knew I'd spent a lot of time trying to get the colour as close as I could to what I saw, and at the time I thought I'd got it pretty close.
What I think changed between the sugar bowl sketch and the soup spoon was the light. It is, after all, all about the light. The sugar bowl was painted on a fairly bright but overcast day. No blue sky, just high grey cloud, and a neutral light. The soup spoon was painted early the following morning, and the day was clear and bright, with lots of blue sky. I think that the difference in the hue of the cloth on the second day was due to the light coming in the window being reflected off blue sky, and this cooler light meant a cooler red being reflected off the cloth. In reality, the difference between the two is much more marked. This camera doesn't deal well with reds.
Regardless of the hue, I think that the value is much more important, so I'm not terribly concerned about the difference. After the value, I think the next most important thing is the chroma, the strength of the colour. These two things, if correctly handled in relation to other elements in the paintings, will define the feeling of light much more than the hue. Who's to say what particular hue of red that cloth is? Especially when it appears to be a different hue in different light.
I came across an interesting quote by Frank Reilly, illustrator and teacher, the other day which is pertinent here:
"80 percent of the picture depends on getting the values correct, 20 percent on correct chroma, and hue doesn't matter much at all."
You might argue over the percentages, but I certainly wouldn't argue that it's the values on which the strength of a painting primarily depends. At least, this kind of painting, where we want the stuff we're painting to look like the stuff it is.
On the subject of Reilly, I recently came across a very comprehensive set of lecture notes taken by a student of Frank Reilly's. This page might not last forever, so I'd go and visit now if I were you. Go to www.dhfa.net and click on the 'Reilly School' link on the left. And be amazed.
The next stage.
Back to the painting. The next stage for this painting is already underway - yet more sketches. I'm not entirely happy that the balance of the values is as good as I can get it yet, so today I did another sketch, this time with a few white flowers in the sugar bowl. This brings the challenge of the white cloth back, since I need to make sure that I keep the lightest parts of the white flowers far enough below the highlights of the sugar bowl to keep the silver looking like silver. But the flower still needs to work as white flower. Todays sketch went fairly well, but the verdict is still out on the flowers, so I'll be repeating it tomorrow. Hopefully I'll be posting these sketches soon.
Once this value problem has been cracked, or at least I've got to a point where I think the values are working sufficiently well, I can begin to work out the composition. The good Mr. Loomis will be my guide, and it will all be recounted here as it happens. If, after all this preparation, this doesn't turn out to be one of my better paintings, I'll eat my entire tube of flake white. Or maybe cut off my ear. So you'd better hope it turns out well, unless you want to see blood all over my web site. Still, if it's the right value and chroma, it might look quite good...