05-22-2019, 11:00 PM
You seem to be working a lot on values right now and I can kind of tell things aren't really clicking into place. On a lot of the color and value studies on prior pages seem to be focused on gradients, modelling and copying over some general values rather than trying to organise your values in a manner where you can more easily understand them. At times this isn't too costly when you're working from reference (although things would be better if things were better organised) but you end up really paying for it in your imaginative work where you have to fall back on your understanding of light and values. The things you seem to be struggling with are very common and that makes it much easier to address.
One of the primary issues you're running into when it comes to values is that you overestimate small differences and underestimate large differences. So when you're trying to model the form with halftones, you overestimate the how different each halftone is from the one next to it. This can often make the skin feel dirty or metallic and it really weakens the overall impression. On the other hand, the difference between light and shadow is often underestimated, your lights are too dark and your shadows are too light. This is further weakens the light impression combined with how much you overestimate the halftones. So instead of having a clear light shape and shadow shape, the image devolves into a mist of halftones.
It's a good idea to develop a way to orient yourself within a value range. The traditional way to do this is by grouping values together and focusing on getting the big impression established between these groups before subdividing them into smaller sections. In the real world, values group themselves much more than you maybe think. We all have a natural tendency to look at details and small variations in the halftones on a surface but our brain is poorly wired to put that into perspective with the big differences of the impression. So artists for as long as art has existed have looked at ways of overcoming this tendency.
This kind of big value grouping is part of what you want to study because it's that big impression that is the core of your painting, it's what makes stuff look like stuff. So try to squint down and break every local color/value into two values forming the light shape and shadow shape with a well studied edge between them varying from sharp to soft depending on the type of light and form. But really try to keep things flat, try to communicate more with the transition edge between these two values rather than breaking it up with further half-tones. When squinting you'll probably notice that most shadows tend to merge into one and the same value and it's a great opportunity to group them together to strengthen the core impression.
If you keep your values well grouped and flat, changing them becomes very easy because you just have to change the color of one shape rather than trying to figure out how to change a complicated gradient. The even flatness makes the information more solid and tangible, like how straight lines are more tangible than complicated curves, it lets you work with them in a much more analytical manner.
I've moistly addressed direct light but you may wonder what things are like under ambient light. It's basically the same thing but instead of two values, you're mostly just comparing local values, so every object basically just gets one value that contrasts with whatever is behind/next to it. Try and keep away from gradients and communicate the form with the contour information, with its drawing and edge.
Once technique I remember Krøyer talking about in some of his letters is basically value triangulation. This is when you've establish a two values on either end of the spectrum, often your lightest light and darkest dark (but this can also be subdivided into smaller areas of value), and you compare it with a third value and compare the contrast between both sides of the spectrum to see if one needs to move that value closer to the light or further to the dark. So if your lightest light is value A, and your darkest dark is value B, your new tone is value C. You look at your subject, find value C and first compare the contrast between A and C, then compare the contrast between B and C. You then do the same thing to the the values on your painting and if the contrast difference is the same between A and C, as well as B and C, then the new value is probably correct. If however it's contrasting too much with one and too little with the other, you then adjust the value to fit correctly between the two.
This system allows you to work with your values in a fairly easy manner while remaining completely relative to your subject, meaning that you don't have to work in 1:1 values, but rather you're expressing the contrast range instead of the literal each object is. This is very useful when working from life because you can't always paint the value of the lightest light or darkest dark, so you have to compress things and work in higher or lower key. I use this system a lot and it was a little awkward at first but you quickly get into it and it really helps.
If you post a study, consider including the reference and explain how you tried to study the values, it makes it much easier to talk about it. Value studies don't need to be perfect drawings so don't put so much pressure on that quality when making them, and critiques addressing the drawing of a study that is about the big impression of values can probably be ignored. Having more information on what you're doing, what you're trying to learn and what you're struggling with will really help anyone interested in giving you feedback. I know feedback on the internet generally sucks but maybe we can change that.
But regardless of any advice on particular methods and techniques, remember that broadly realist art has a history spanning hundreds of years and there are a lot of ways very different artists have solved the same problems they run into. So feel free to think for yourself if something doesn't resonate with you.
One of the primary issues you're running into when it comes to values is that you overestimate small differences and underestimate large differences. So when you're trying to model the form with halftones, you overestimate the how different each halftone is from the one next to it. This can often make the skin feel dirty or metallic and it really weakens the overall impression. On the other hand, the difference between light and shadow is often underestimated, your lights are too dark and your shadows are too light. This is further weakens the light impression combined with how much you overestimate the halftones. So instead of having a clear light shape and shadow shape, the image devolves into a mist of halftones.
It's a good idea to develop a way to orient yourself within a value range. The traditional way to do this is by grouping values together and focusing on getting the big impression established between these groups before subdividing them into smaller sections. In the real world, values group themselves much more than you maybe think. We all have a natural tendency to look at details and small variations in the halftones on a surface but our brain is poorly wired to put that into perspective with the big differences of the impression. So artists for as long as art has existed have looked at ways of overcoming this tendency.
This kind of big value grouping is part of what you want to study because it's that big impression that is the core of your painting, it's what makes stuff look like stuff. So try to squint down and break every local color/value into two values forming the light shape and shadow shape with a well studied edge between them varying from sharp to soft depending on the type of light and form. But really try to keep things flat, try to communicate more with the transition edge between these two values rather than breaking it up with further half-tones. When squinting you'll probably notice that most shadows tend to merge into one and the same value and it's a great opportunity to group them together to strengthen the core impression.
If you keep your values well grouped and flat, changing them becomes very easy because you just have to change the color of one shape rather than trying to figure out how to change a complicated gradient. The even flatness makes the information more solid and tangible, like how straight lines are more tangible than complicated curves, it lets you work with them in a much more analytical manner.
I've moistly addressed direct light but you may wonder what things are like under ambient light. It's basically the same thing but instead of two values, you're mostly just comparing local values, so every object basically just gets one value that contrasts with whatever is behind/next to it. Try and keep away from gradients and communicate the form with the contour information, with its drawing and edge.
Once technique I remember Krøyer talking about in some of his letters is basically value triangulation. This is when you've establish a two values on either end of the spectrum, often your lightest light and darkest dark (but this can also be subdivided into smaller areas of value), and you compare it with a third value and compare the contrast between both sides of the spectrum to see if one needs to move that value closer to the light or further to the dark. So if your lightest light is value A, and your darkest dark is value B, your new tone is value C. You look at your subject, find value C and first compare the contrast between A and C, then compare the contrast between B and C. You then do the same thing to the the values on your painting and if the contrast difference is the same between A and C, as well as B and C, then the new value is probably correct. If however it's contrasting too much with one and too little with the other, you then adjust the value to fit correctly between the two.
This system allows you to work with your values in a fairly easy manner while remaining completely relative to your subject, meaning that you don't have to work in 1:1 values, but rather you're expressing the contrast range instead of the literal each object is. This is very useful when working from life because you can't always paint the value of the lightest light or darkest dark, so you have to compress things and work in higher or lower key. I use this system a lot and it was a little awkward at first but you quickly get into it and it really helps.
If you post a study, consider including the reference and explain how you tried to study the values, it makes it much easier to talk about it. Value studies don't need to be perfect drawings so don't put so much pressure on that quality when making them, and critiques addressing the drawing of a study that is about the big impression of values can probably be ignored. Having more information on what you're doing, what you're trying to learn and what you're struggling with will really help anyone interested in giving you feedback. I know feedback on the internet generally sucks but maybe we can change that.
But regardless of any advice on particular methods and techniques, remember that broadly realist art has a history spanning hundreds of years and there are a lot of ways very different artists have solved the same problems they run into. So feel free to think for yourself if something doesn't resonate with you.
Discord - JetJaguar#8954