Hacker News Reader

Comment by npgatech | Hacker News Reader

This article doesn't teach the most fundamental aspect of drawing, i.e. to internalize a 3-dimensional object and being able to express it on a 2d plane. In order to efficiently and quickly learn this, you need to be able to accurately draw basic 3D objects (such as a cube, cone, cylinder, etc.) from any viewpoint. You don't need to focus on perspective and foreshortening just yet, an isometric view is fine.

Now, glue these basic objects together and make something slightly more complex. If you can draw it from any viewpoint (not just orthographic views), you have successfully learned the most fundamental and crucial part of the art of drawing.

Ideas of chiaroscuro are simply the tools to express depth on a 2D plane. They are not to be learned until you can master the whole internalization part.

This bothers me so much. No one knows how to teach this well except for a few YouTubers and college professors. Check out Sycra's drawing tutorials on YT.

View on HN

I agree with you that this is a good approach for learning the genre of drawing you are describing, but the art of drawing in general has just about as much to do with representing 3D forms on 2D planes as the art of programming does. It's a niche, especially these days.

I have only studied the subject of drawing very superficially, but it is easy to find people who say "the trick is drawing what you see, not what you know is there". Then there are other people who seem to agree with you, and break everything down into basic shapes, which seems to contradict the previous statement. Different schools? Different approaches to suit different persons? A fundamental divide in the drawing community? :-)

I think the two stances are consistent, and the drawing of shapes actually enforces the "draw what you see".

Breaking an object down into its constituent shapes, is based on what you see, rather than the image in your mind. It turns out the mind is really bad at retaining the shape of something, even when we're looking at it.

We "know" what a bicycle looks like, and our mind often skirts over the detail. When asked to draw a bicycle from memory, people are terrible at it[1]. Drawing from memory is like a slightly different task, but when people draw "what they know is there", they look at a scene and remember the objects in it, then look down at the paper and draw them from memory. Not a memory of what the scene looked like, but what it was. Each object in the scene becomes the butchered, bike form memory version of itself.

If you're building the scene out of basic shapes, then you first have to look at what the scene looks like, rather than what it is.

There are some shortcuts and techniques which might seem to go against the purist version of "draw what you see". Knowing the ratios between the eyes, mouth, nose etc. But actually these are based on previous practice in "drawing what you see" (or received practice) and are still taking you away from drawing a "face" and instead drawing a set of shapes - ie "what you see".

[1] https://www.wired.com/2016/04/can-draw-bikes-memory-definite...

Reply on news.ycombinator.com