Made some further changes to the castle, to the forest, to the grass and the tree. Still haven't refined the mountains, the clouds or the overall palette.
I added some yellow grass, I don't know if it stands out too much. Someone once criticized me for thinking too naively about colour, colouring the sky blue and grass green without any deeper understanding of colour. I thought that was interesting, so I'm looking to challenge myself a little bit here. Maybe I should challenge myself more in that regard in the future.
@Ai: In regards to the tree, I think of the foliage as a semi-spherical object, being hit by a beam of light. And the light source is almost directly to the side of the viewer. If I only highlighted the edges of the foliage, this would suggest the light source is behind the tree. As the foliage is semi-spherical, it makes sense to have plenty of light on the front left side, as seen from the viewer, I think.
You ask if crepuscular rays are shadows. My understanding is that they're basically beams of light (light reflected off moisture and dust in the air) contrasted by space with less light. But I don't quite know how to show that correctly. For example, I was very unsure about the dark space below the tree's foliage, or the diagonal shadow I originally had from the castle. Because a shadow doesn't work the same was as light. A shadow isn't visible in the air. If there's a bright snow-covered mountain in the horizon, and a big building in front of it, the building shouldn't make the mountain look any darker. The shadow would only be visible on the ground. On the other hand, if there's a lot of dust and moisture, then all the air will be reflecting light, except where the building is blocking the sunlight. So in that sense, crepuscular rays are shadows. And I suppose, in that sense, it makes sense for the space beneath the tree to be darker. Maybe I should bring back the shadow from the castle, now that I think about it.
Thanks for making me reflect on this.
In regards to workflow, I've seen plenty of good artists work with my method, which is why I moved away from using lines in the first place. But I also notice that different artists have different methods, regardless of their talent. E.g. some people draw people using the blobs which are later refined, other people draw skeleton-like lines that are later covered by muscles and tissue and then clothes.
In regards to reference images... well, it's really tricky to find images that show exactly what I'm looking for. And it's very easy to be misled by Google Images, because sometimes you simply can't find good examples of what you're looking for. If you look for a red moon, for example, you will generally find the most red moons out there, as people tend to upload pictures of nights when the moon was really, really red. They may also upload images that have been filtered or photoshopped to make the moon look extra red. The same goes for red trees, fog, rainbows, etc. You'll find a lot of extremes and a lot of photoshopped / filtered images. I may be wrong, but I think both of your refs are photoshopped.
Thanks for doing the searches for me. I try to use Google Images the same way when I'm looking for refs
@Gil: That's actually a very smart and constructive way of doing edits, now that I think about it. I've previously tried to shy away from extreme edits, as people may instantly reject them because they're too far away from their original vision. But as I've come back to your edit again and again, it's grown on me. And I think I'm moving toward a compromise I'm comfortable with.
Now the question is just if I've managed to find a good compromise or if I'm falling between two chairs.
Also not sure if I have enough contrast and the right balance. This is where it gets tricky for me.