Starting with your question about shadows that end up filling in cracks, I do have some extra notes about those kinds of situations that I'll share in a moment, but I wanted to specifically address the meat of your question first, as those notes aren't directly tackling that issue. You are right - given a crack in an otherwise level surface, that crack can only ever get completely filled with shadow.

The way we address this in the gradient comes down to elements we can choose to control - one that falls entirely within the scope of the exercise, and another that isn't entirely ideal but is still a choice we can make within this exercise in order to get the most out of it.

First the solution within the scope of the exercise: we set the point at which the average sized crack in our texture is "filled" with shadow to be at the farthest point to the left. In essence we're given a range that exists between "shadow is barely cast" on the left side just before the solid white bar, and "shadow is cast to its fullest extent" on the far right, with the latter usually having an "infinite" cap due to the textural forms being projected out from the surface, rather than into it. We are of course in control of where these points fall - our reference doesn't dictate those decisions, it merely gives us information to work with and leverage towards our goals (or more accurately the goals of the exercise). So, while the shift from one end to the other might be very gradual and subtle, given the narrowness of the cracks, this still allows us to cover a range and create a gradient that transitions to the sparsest shadows to the largest.

The second is to actively choose to either work with a texture that has a lot of cracks (a greater increase in density will make up for the narrowness of your biggest shadows) or artificially increase the density to be greater than what your reference image may display. While we generally stick with what the reference shows us in terms of the nature of the texture (and so the density is part of the texture's properties, what makes it unique), the purpose of the exercise is still just to practice implicit markmaking and conveying the presence of those textural forms with cast shadows. So we still want to prioritize the main focus, even if it means undermining some of the more superficial aspects - although in truth increasing the density as a whole is the same thing as simply "zooming out" from a surface, so it's not that far astray from even those less important aspects of the exercise.

One thing we don't want to do in this regard is actually increase the density of the cracks on the far left side, but not the far right - we want that physical density (in terms of the 3D forms present) to be consistent across the whole gradient. What changes is the density of the cast shadow shapes being used to convey it. The whole thing about implicit markmaking is that the texture itself remains the same, but we get to decide how sparsely or densely we convey that detail.

As to the other info about cracks and other similar textures, here's those notes I mentioned:

When it comes to those tires with shallow grooves, or really any texture consisting of holes, cracks, etc. it's very common for us to view these named things (the grooves, the cracks, etc.) as being the textural forms in question - but of course they're not forms at all. They're empty, negative space, and it's the structures that surround these empty spaces that are the actual forms for us to consider when designing the shadows they'll cast. This is demonstrated in this diagram. This doesn't always actually result in a different result at the end of the day, but as these are all exercises, how we think about them and how we come to that result is just as important - if not moreso.

Finally, looking at your work I'm actually very happy that you asked that question, and gave me something to go on about at length, as your actual homework submission, being very well done and demonstrating a really solid understanding of how to imply the presence of forms by their cast shadows alone, doesn't really give me much to latch onto! The only issues I really noticed were related to your question - for example, how the bottom-most texture here ended up struggling to mask the hard edge of the black bar to the left - but the reason for this was simply that the nature of the texture made it difficult to obfuscate that edge.

I do have a couple suggestions on how I'd try and soften the issue in this particular example though-

• You've got the texture such that all of the intertwined strips are running at 45 degrees to the direction of the gradient - if you rotate the texture itself so that one set of strips runs straight up and down, and the other runs across left and right, that would help as you would no longer have to deal with shadows that are set at an angle.

• While the main textural forms are the strips themselves, the strips do have their own little texture as well, which is relevant here. You opted to capture them with stray marks, but in truth ideally we'd be capturing those smaller shadows as shadow shapes too, and leveraging them in the gradient too. At least, that's if we choose to include them. We ultimately decide whether we want to leave those smaller elements out or not, but in this case you did include them, so it's always good to leverage them to their fullest extent, if they're included at all.

The last point I wanted to make was that in the rocks texture at the bottom of this page (and to a lesser extent the one above it), you momentarily strayed into using a sort of "pointilism" effect to achieve midtones. This is essentially the same as using hatching, and should not be leveraged in this exercise. Every mark should be an individually designed cast shadow shape that requires you to think about the relationship in 3D space between the form casting it, and the surface receiving it.

Anyway, all in all, your work here is very well done. I'll go ahead and mark this challenge as complete.