Tips On Drawing Realistic Faces

On December 1, 2016, I asked myself the question: With only one month of practice, can I learn how to draw realistic portraits with only pencil and paper?

On December 24, 2016, after 26 hours of practice, I found out that the answer was yes.

During the month of December, I documented my entire learning process in a series of 31 daily blog posts, which are compiled here into a single narrative. In this article, you can relive my month of insights, frustrations, learning hacks, and triumphs, as I strive towards monthly mastery.

New month, new challenge.

For the month of December, my goal is to draw a realistic self-portrait with only pencil and paper. Along the way, in order to learn the fundamentals of drawing and portraiture, I will also draw many other faces, which will hopefully keep this month's posts more varied and interesting.

This new challenge starts today, December 1, 2016, and, by December 31, I hope to be a master of portrait drawing.

My starting point

I've had strong artistic tendencies since I was a kid, but I've never invested much in my fine art skills. Instead, I've channeled my artistic impulses mainly through music, film, and computer-aided design.

Thus, to set a baseline for this month's challenge, I've drawn a before self-portrait with my current drawing skills. Although it's not the absolute worst thing ever drawn, it sadly doesn't look very much like me.

Measuring success

Measuring success for this challenge is certainly more subjective than last month (where I successfully memorized a deck of cards in less than 2 minutes).

In this case, the best I can do is show a photo that demonstrates the level of drawing I'm aiming to reach…

This portrait is the example drawn in the Vitruvian Studio Portrait Drawing Course, which is the course I'll be following this month.

Clearly, there are major differences in realism between my starting drawing and this example portrait. So, if I can match the level of this example (which will be, of course, a subjective, but hopefully honest judgement), I will consider this challenge a success.

With my goal set, it's time to start drawing…

In my life, I've created a fair bit of (what I'll call) art. However, I've done so, not by relying on well-developed fine art skills, but instead, by cheating my way through the artistic process.

Basically, I've used everything at my disposal (except for fine arts skills) to create artistically.

You can decide if this is cheating or not, but either way, this month is going to be different. This month, I am actually going to invest in my fine art skills. This month, I'm going to take a pencil and paper, and nothing else, and make it happen.

However, before I make it happen, I thought it would be fun to share some of my previous works.

1. Lego Portraits (with the help of Photoshop)

During high school, whenever I was tasked with making someone a gift, I usually opted to construct a custom Warhol-inspired portrait out of Legos.

Here are two portraits that I made for my cousins Adam and Marissa.

And another one I made for my grandparents.

While these pieces may look like they required some amount of artistic genius to pull off (do they?), that's really not the case. Instead, these pieces just required some clever computational analysis, planning in Photoshop, and executional patience (while glueing and placing each Lego piece).

The computer was the real artistic champion here.

2. Counterfeit paintings (using optical tricks)

I've also experimented using optical tools (like mirrors and lens) to mechanically create. Although, I haven't invested enough time to produce anything worth sharing.

Tim Jenison, on the other hand, does have something worth sharing. Without any artistic training, he painted a nearly-exact replica of a Vermeer painting solely using optical techniques.

Tim's journey is documented in the Penn and Teller-produced film "Tim's Vermeer", which I highly recommend you check out.

Here's Tim's final painting.

This month I'm only using pencil and paper

While technology-aided art still should probably count as art (in some capacity), this month, I'm committed to creating using only the tools shown below: 9 black pencils, 1 white pencil, a few different erasers, and a gray piece of paper (which I'll explain another time).

It's going to be hard, but that's the point.

This month, to learn how to draw portraits, I'll be following the Portrait Drawing video course from Vitruvian Studio.

Today, I spent 2.5 hours starting the course and beginning my first portrait.

Selecting who to draw

For my first piece, rather than drawing the model from the course, I've chosen to draw Derren Brown, who originally inspired me to pursuit portrait drawing.

Derren is a British illusionist, who I've been following for a while now, and who, I recently learned, casually paints portraits on the side.

Here are a few things he's casually painted.

After seeing these, I decided I too would like to be the kind of person that casually paints impressively good portraits on the side.

For now, before I get to the painting, I'll start off by mastering the drawing part of program.

This is the picture of Derren I'm drawing.

And here's my setup.

Starting the drawing

The first module of the course focuses on mapping out the portrait, which includes determining the shape of the head and locating the features.

Finding the top and bottom of the head

I started by arbitrarily drawing two lines on the page to indicate the level of the top of the head and the level of the bottom of the head.

Then, I arbitrarily marked, on the top level, the highest point of the head, and then used the angle between this point and the bottom of the chin, to locate the bottom of the chin on the page.

I also drew in the level of the notch of the neck. The first time, I drew it too low, so I moved it up. I gauged this distances as a proposition of the head length.

Find the leftmost and rightmost parts of the head

With the topmost and bottommost points identified, I then needed to identify the leftmost and rightmost points.

To do this, I used a new technique I learned called triangulation. To triangulate a new point, I first sight (try to visualize) the angles to this new point from two existing points. Then, I draw lines from the existing points in the direction of the new point based on the sighted angles. Finally, I mark the new point where the lines intersect.

After checking the angles again, I updated these two new points.

To check, I then sighted the angle between the two new points, ensuring this angle matches what I see on Derren's head.

Drawing the shape of the head

With these four outer points drawn, the next step is to draw in the shape of the head. To do this, I continued to triangulate more points, and draw in the necessary curves to connect them.

I continued in this way, until I outlined the entire shape of the head.

It didn't look quite right, so I checked a bunch of angles.

Once it seemed closer, I added in the neck and shoulders.

With the neck and shoulders in place, it again didn't look right. So, I checked more angles and made adjustments as necessary (mostly to broaden the jaw)

The head was now looking pretty good, but the neck and shoulders needed a few adjustments. I retriangulated, and adjusted the collar upwards.

That's it for today

Getting to this point took me 2.5 hours, which was split between watching the video course and drawing my Derren portrait.

So far, the portrait doesn't look like much, but I still learned a bunch today. I particularly like the triangulation technique, which makes drawing much more procedural and mathematical (a.k.a. easier for me).

Tomorrow, I'll continue following the course, and start drawing in the facial features.

Yesterday, I started following along with the Vitruvian Studio portrait course, and began drawing a portrait of Derren Brown.

Here's what I accomplished yesterday.

And here's my end goal (more or less).

Today, I spent another 2.5 hours watching the course and working on the portrait.

Today's progress

Drawing in guides

The first thing I did today was add construction lines to my drawing. These construction lines are designed to act as landmarks and help me eventually place the facial features.

First, I drew in the vertical center line, which will help me laterally place the features.

Then, I marked eye level, to start gauging the features' vertical placement.

I followed up with the levels of the brows, nose, and lips.

I made a bit of a mistake here. I drew the horizontal construction lines perpendicular to the center line (which seemed reasonable), but did not mimic the angle of the features in the actual drawing.

So, I sighted the correct angles, and adjusted the construction lines accordingly.

Blocking in the features

With the construction lines as references, I was then ready to start blocking in the facial features.

I started by adjusting the center line slightly for the nose, and marking the nose's outer boundary.

Then, I drew in shapes for the brows.

Next, I included the eye sockets and some more detail around the nose.

Finally, I added in shapes for the eyelids and eyes, and finished up for the day.

Reaching this point took another 2.5 hours.

Progress still seems fairly slow on the drawing, but I'm making a conscious effort to work carefully through the blocking in phase (so I can practice what I'm learning, and so I can ensure the portrait is built on a strong foundation).

I'll start detailing the features tomorrow.

Today, for the third day in a row, I spent 2.5 hours on my Derren Brown drawing. However, unlike the other days, today, I feel like I made a lot of progress.

Finish blocking in the features

Picking up where I left off, I continued to block in shapes for the features.

I added in the center line of the lips and the shadow on the nose.

I then finished the lips and added a line for the chin.

Lastly, I blocked in the main structures of the ear and added an outline for the beard.

Drawing in shadow/highlight shapes

With the features in place, I next blocked in shapes for the shadows and highlights.

With these tonal contours in place, I darkened the shadow areas slightly, giving the portrait some roundness and three-dimensionality.

Detailing features

With the features and shadows blocked in, I detailed the features, starting with the eyes.

Left eye done.

Right eye done.

Nose done.

Lips done.

Finally, I finished up for the day with the ear.

After 7.5 hours of work (2.5 hours over the past three days), I'm finally hopefully that this portrait will resemble Derren Brown.

Tomorrow, I'll starting adding tonal values (i.e. shading) to the drawing.

Observation about today's session: Based on the output from today, it may seem like today's drawing was the most technically challenging. But, in fact, I found just the opposite.

Because I spent the past two days meticulously locating and blocking in the features, it was very easy to add the incremental detail. (Trying to draw big shapes is much harder than trying to draw little shapes. Little shapes are a lot easier to visually understand and replicate)

In fact, I suspect that today was least consequential to the outcome of the portrait. If I mess up the shape of the head and the location of the features, I have very little chance of capturing a likeness. If the features are not quite accurately detailed, but in the right place, I still might have something…

Yesterday, after 7.5 hours of work, I finally finished sketching / laying out my first portrait. Today, I started adding tonal values (a.k.a. "shading the drawing").

Before I show today's progress, I want to share two techniques I learned that make it significantly easier to accurately add tonal values to portraits.

1. Start with the most extreme values and then meet in the middle

The human eye is really bad at assessing tonal values in isolation — which is why your brain thinks squares A and B below are very different colors, when, in fact, they are the same.

Thus, instead of relying on visual inferences, tonal values can be better approximated through a simple, not-so-interpretative procedure.

Here's how it works:

Start by identifying the absolute darkest and absolute lightest areas of the drawing. For the darkest areas, shade them as dark as you can/want. For the lightest areas, highlight them as light as you can/want.

This establishes the entire tonal range of the drawing, which is called the key of the drawing.

Establishing the key is straightforward, and doesn't require much visual interpretation (i.e. it's easy to find the lightest lights and the darkest darks).

Once the key is established, and the lightest and darkest values are in place, the intermediate values need to be introduced. Again, this can be done procedurally, by identifying and shading/highlighting the areas which are slightly lighter than the darkest darks and slightly darker than the lightest lights. Continuing recursively in this way, the tonal values eventually meet in the middle, and the drawing (or the relevant part of the drawing) is complete.

2. Squint to better see tonal shapes

When keying the drawing (and developing tonal values in general) it's important that the shapes of the tonal areas are captured accurately.

In other words, if the highlight on the forehead is angular, drawing it with rounded edges wouldn't properly capture the form.

This sounds obvious, but again, your brain and visual system can play tricks on you. Your brain is attempting to see a face (via your psychologically skewed, emotions-based mental model of a face), and not just tonal blobs.

In fact, this psychological problem of misinterpreting faces is so common, there are entire drawing systems (like drawing upside down, drawing the negative space around the face, etc.) designed to combat these problems.

Side note: Here's a video of Derren Brown, the subject of my portrait, when he used to have hair, experimenting with some of these alternative methods of painting. It's a pretty cool trick.(If you're going to watch, stick it out until the end).

In order to accurately see tonal shapes, and avoid psychological errors, I've found one method to be surprisingly successful: squinting.

Basically, you look at the area you want to draw, squint your eyes (so the image becomes blurred and your brain no longer sees a face), and identify the tonal shapes you see through your eyelashes. This works super well. (I didn't invent this method, I've just validated that it works for me).

Today's progress

With these techniques newly-learned, I began to add tonal values to my Derren Brown portrait.

First, I started with the eye.

In the course, the teacher mentioned that it's good to start with a small area that exhibits the full range of tones.

However, the eye was too small to help effectively establish the key. So, I keyed the drawing more aggressively, starting with the shadow on the nose and the highlights on the forehead and cheek.

I continued shading the darkest areas along the right side of the face.

Additionally, while doing this, to check the accuracy of my key, I started developing the eye.

I finished up my key, by adding shadows to the lower face and the back of the head, and was ready to begin modeling the form (finding the intermediate values between the darks and lights).

I started with the forehead.

Added a bit more detail.

And then smoothed everything out.

This is where I stopped for the day, after another 2.5 hours of working.

Derren looks a bit too shiny right now — a bit like a mannequin or the Tin Man — but I'm optimistic that this effect will vanish once I model the rest of the form.

I'm guessing I have another 5 hours of work left on this.

Today, like yesterday, I continued adding tonal values to the portrait. I spent a little less than two hours, and am getting really excited about the results.

Here's where I stopped yesterday.

I proceeded today by first addressing the nose.

Then, I addressed the right half of the face — further developing the shadow.

Next, I moved on to Derren's hair and beard.

Since the demo portrait in course is based on a long-haired female model, I had to do a bit more freestyling at this point. I think it works.

I continued with the upper part of the beard, and finished up for the day.

Tomorrow, I need to finish the mouth, the ear, the neck, the lower part of the beard, and perhaps the clothing.

Getting close…

Today, after another 2.5 hours of work, I finally completed my Derren Brown portrait.

In the coming days, I will write a few detailed posts about what I've learned, how I plan to move forward, etc., but for now, I'll just share the final photos of my progress.

Today's progress

I started off by detailing the lips.

Then, I added the mustache.

With this facial hair momentum, I finished off the beard.

Then, the ear.

Finally, I completed the neck, decided not to address the clothes, signed it, and I was done.

For my first portrait of the month, I'm quite happy with how it turned out.

Nine days ago, I began my 30-day quest to learn how to draw photorealistic portraits. Since then, I've watched the entire 10 hours of the Vitruvian Studio drawing course, as well as spent 14.5 hours working on my first portrait.

Here's the result…

And here's a video documenting the progression.

Considering where I started only nine days ago (see the before portrait), it's hard for me to believe that I actually drew this. It's not perfect, but I'm definitely excited about the outcome.

Part of me lacks the motivation to continue drawing, as I feel like I've already accomplished my goal. The other (more overpowering) part of me realizes that I have another 21 days to improve even further, so that's what I plan to do.

In particular, I'm going try to reduce the amount of time necessary to complete a portrait like this. With some practice, I think I can reduce my time down from 14.5 hours to 4–5 hours.

Tomorrow, I'm going to go through my previous posts (1, 2, 3, 4, 5, 6) and write up a "Portrait Drawing Cheat Sheet". Then, I'm going to break down the cheat sheet into isolated, practicable skills and drills, work on those individual skills for 1–2 weeks, and then start working on my self-portrait to finish off the month.

Here is my "Portrait Drawing Cheat Sheet", which features step-by-step instructions on how to draw a portrait.

These steps are based on the excellent portrait drawing course by Vitruvian Studio, which I highly recommend you purchase if you are serious about learning how to draw.

The Instructions

Mark the top of the head. Arbitrarily draw a line towards the top of the page. This represents the top of the head.
Mark the bottom of the chin. Arbitrarily draw a line near the lower third of the page. This represents the bottom of the chin.
Mark the notch of the neck. On the subject, using your pencil as a guide, measure the distance from the lowest point of the head to the notch of the neck. Determine how many of these distances can fit inside the vertical distance of the head. Use this is as guide to draw a horizontal line towards the bottom of the page to represent the notch of the neck.
Find the highest point of the head. Arbitrarily determine a point on the top line. This represents the highest point of the head. Often, on the subject, this point sits far back on the head.
Find the lowest point of the chin. Using your pencil as a guide, determine the angle from the highest point of the head to the lowest point of the chin. Draw a line at this angle from the highest point of the head (as marked on the page) down towards the bottom of the chin line. Draw a dash where these lines intersect. This intersection represents the lowest point of the chin.
Find the leftmost boundary. Identify the leftmost boundary on your subject. Determine the angle to this leftmost point from the highest point, and draw a line at that angle from the highest point towards the leftmost boundary on the page. Do the same from the lowest point. Draw a marking where these two lines intersect. This intersection represents the leftmost boundary. The technique used to find this boundary is called triangulation.
Find the rightmost boundary. Again, triangulate from the highest and lowest points to find the rightmost boundary of the head.
Check the angle. On the subject, use your pencil to find the angle between the leftmost and rightmost boundaries. Check if this angle matches the angle represented on the page. If not, retriangulate and check again.
Draw the outer-boundary of the head and hair. Triangulate points around the head and connect them with straight lines. Once the general shape seems right, smooth out the kinks. Check the angles between various points on the subject and on the page to make sure everything looks right. If there seems to be inconsistencies, retriangulate and adjust. Do the same for the hair line.
Draw the vertical center line. Pick some central point that looks like its on the vertical center line. Triangulate from outer-points inwards to find this central point. Check the angle from the bottom/center of the chin to this point. Use this as a guide to draw in the entire vertical center line. As the center line approaches the top of the head, it typically flattens, as it rounds back behind the head.
Draw the level of the eyes. The level of the eyes typically falls about halfway between the top and bottom of the head. Use this as a starting point. Draw in this level, and then check angles to confirm. Move up or down until everything checks out.
Draw in the level of the brows and bottom of the nose. If you divide the face length into thirds, typically the level of the brows fall on the upper third line and the level of the nose falls on the bottom third line. Use this as a starting point. Draw in these level, and the check angles to confirm. Move the level up or down until everything checks out.
Draw in the level of the start of the nose. The nose begins somewhere between the level of the brows and the level of the eyes. Gauge where this is and draw it in.
Draw in the bottom and middle of the lips. If you divide the distance between the bottom of the nose and the bottom of the chin into halves, the level of the bottom of the lips typically falls at the halfway point. Use this as a starting point to draw in this level. Then, gauge where the middle of the lips falls relative to the distance between the bottom of the lips and the bottom of the nose. Draw that in.
Adjust the center line for the nose. Starting from the level of the start of the nose, adjust the center line so its angle matches the center line of the nose. Typically this will be in two parts. The angle outwards from the level of the start of the nose to the peak of the nose, and the angle inwards from the peak of the nose to the bottom of the nose.
Adjust the center line for the mouth. The mouth typically has some volume, which pushes the center line forward. Adjust the center line forward below the nose to account for the volume in the mouth.
Draw in the shape of the eyes and eye sockets. Triangulate the corners of the eyes, and then draw in the complete shapes. Do the same for the lids and the eye sockets.
Draw in the shape of the brows. Triangulate the corners of the brows, and then draw in the complete shapes.
Draw in the shape of the nose. Triangulate the peak of the nose and the wing of the nose. Then, draw in the complete shape.
Draw in the shape of the mouth. Triangulate the corners of the mouth. Then, draw in the complete shape.
Draw in the level of the chin. Triangulate the level of the chin, and draw a line to distinguish the shape.
Draw in the shape of the ear. Triangulate points of angle-change around the ear. Connect these points with appropriately angled lines, and then smooth out the kinks.
Draw in shadow shapes. Identify shapes of main shadow areas. Triangulate their boundaries and draw them in.
Darken the shadow shapes. Lightly shade in the shadow areas of the portrait. Use a soft, clean paint brush to smooth out the material on the page. This will introduce some 3-dimensionality to your portrait, which should help you better visualize if anything doesn't seem quite right. If there is something that seems incorrect, fix it.
Detail the eyes. Draw in the iris, pupils, and other details.
Detail the nose. Draw in the nostrils and other details.
Detail the lips. Smooth out the shape of the lips.
Detail the ear. Draw in some of the main inner land marks.
Key the drawing. Identify the lightest and darkest tones on the subject, and add these tones to the page.
Modeling an area. Pick an area of the head (like the forehead), and detail some of the main places of tone-change. Identify and add in the main light and dark areas. Using a shading stump and the necessary pencils, fill in the transition tones. To better see the shapes of highlights and shadow, squint your eyes until the face isn't recognizable as a face, but rather a collection of tonal blobs.
Model the remaining areas. Continue as above until all areas are modeled.
Sign it. And you're done.

A few days ago, I finished drawing my first portrait. Since then, I've reread my notes, reviewed some parts of the course, and wrote up my "Portrait Drawing Cheat Sheet".

With all the steps documented, it's now time to deliberately practice the most important skills.

In particular, as I said on Day 35, I believe that it's most important to accurately capture the proportions of the head, the head shape, and the level of the features. If these things are done correctly, the rest of the process is very forgiving. If not, the portrait will end up beautifully shaded, but won't look like the subject.

Today, I'm going to practice finding the correct proportions of the subject's head using a few celebrities: Matt Damon, Natalie Portman, and Morgan Freeman.

Matt Damon

Here's the photo I'm using.

Here's my attempt to locate the peak of his head, the lowest point of his chin (which is located on the chin's left side), the leftmost point of his cheek, and the rightmost point of his ear.

In Photoshop, I overlaid my sketch on the photo to check. I was pretty accurate.

Natalie Portman

Here's Natalie.

And here's my attempt to locate the peak of her hair, the lowest point of her chin (again on the chin's left side), the rightmost point of her cheek, the leftmost point of her hair, and the notch of her neck.

Checking in Photoshop, everything seems pretty accurate. Although, the low point of the chin may be slightly too far left.

Morgan Freeman

Here I try to locate the peak of his head, the lowest point of his chin, the rightmost point of his ear, the leftmost point of his ear, and the notch of his neck.

This one looks right on the money.

With each of the sketches, unlike with my Derren Brown portrait, I felt that I was able to see the angle on the subject and accurately replicate it on the page with limited effort.

This is a good sign…

Yesterday, I practiced triangulating the proportions of a few celebrity heads.

For example, here's one I did of Natalie Portman.

Today, I practiced triangulating the complete head shape and gauging the level of features.

It took about 45 minutes.

To assess my work, I overlaid the sketch on Natalie.

My Critique

The face shape is accurate
The level of the features is accurate
The angle of the features is accurate
The center line curves a little too quickly as it moves up between the eyes
The neck shape is inaccurate — I especially misestimated the starting point of the neck on the right side.
Above the right eye, the angle of the head/hair is too steep
The peak of the head is too steep
The angle of the hair above the ear isn't steep enough

Overall, I'd give the sketch a B-.

Since I was accurate with the face shape and the level of features, if I continued working, I suspect I would develop the face fairly accurately. As a result, I would likely have enough accurate information to gradually correct the major mistakes with the head and hair shape.

Tomorrow, I'll practice again on a different celebrity.

Today, I didn't have too much time to draw. So, I quickly progressed the Matt Damon sketch I started two days ago.

Here's what I shared on Sunday.

Today, I spent 30 minutes sketching the head shape and feature guides.

Just looking at the sketch, the head shapes seems a little narrow for Matt Damon. But, overlaid on the photo, it seems to match up.

With the exception of the oddly tiny ear, everything else seems to line up well. The head shape, face shape, and hair shape seem accurate. The level of the features and the center line seem accurate. The wing of the nose is a bit too far to the right, but I really just threw that in for fun.

Overall, I'm pretty happy with the outcome — especially since I sketched this fairly quickly. I guess that means I'm improving…

Last month, when I was learning to memorize a deck of cards at grandmaster speeds, I started unintentionally seeing playing cards in the real-world. In particular, real-world things (like wheelchairs and airplanes), which have association in my mnemonic system, were triggering images of playing cards, without any conscious thought on my part.

Simply, I was rewiring my brain.

This month, as I learn to draw faces, I'm experiencing a new phenomenon… For the past few days, I've found myself scrutinizing and deconstructing other people's faces on the train, at work, on the street, at Whole Foods, etc. Wherever there is a face, I can't help but try to analyze it, and imagine how I'd draw it.

Now (and I hope this eventually wears off), when I see a new face, my first instinct is to estimate the ratio between the height and width of the head. Other times, I just look to see what shapes the eye sockets are. Or how prominent the brow ridge is. Or if the nose and brows equally break the face in thirds.

Basically, I can't stop staring at people.

So, thank you people of San Francisco for not getting totally creeped out. I promise I'll stop soon.

For the past couple days, I've been itching to start my self-portrait. So, today, I did just that.

After working for about an hour, I was able to finish sketching the outline of the head, hair, and neck.

And here's a video of today's progression.

So far, so good. Tomorrow, I'll start blocking in the features.

Today, I continued working on my self-portrait. Although it's coming together nicely, I made a mistake upfront that's definitely costing me now.

Before, I get to that, though, let me first share today's progress.

My mistake

Although I'm loving the composition of my self-portrait, I've sadly draw everything 10–20% too small.

Take a look at the self-portrait side-by-side with the Derren Brown portrait. My head is noticeably smaller.

Again, I think this is okay compositionally, but it's still a bit of a problem — particularly, for two reasons.

A smaller drawing offers smaller margins for error. If I slightly misplace the corner of the mouth or the height of the brow, the distance between the correct and incorrect placements represents a proportionally larger difference on a smaller drawing. In other words, smaller drawings are less forgiving and errors are more pronounced.
A smaller drawing means finer details. My pencil sharpener doesn't seem to work very well with the pencils I have, which means I'm drawing the tiny eyelids on my self-portrait with a tree trunk. Basically, the smaller drawing requires that I work in finer areas, which is challenging with the tools I have.

Nevertheless, I will persist, since, even with the sizing mistake (and the associated challenges), I'm quite happy with the portrait so far.

In fact, challenges are probably a good thing (I hope). Ideally, they push me to become a better artist.

Anyway, I think the takeaway is that I need to invest in a better pencil sharpener…

Today, my self-portrait progress is broken into two parts:

Finishing the sketch
Defacing the sketch (a.k.a. adding tonal values)

Finishing the sketch

Yesterday, I was able to sketch about 80% of the portrait. Today, I just need to add the final details.

I start by blocking in shadow areas near the mouth, on the forehead, and on the neck.

Then, I darken the hair and eyebrows.

I add detail to the eyes, and the portrait jumps to life.

Finally, I detail the ear, which is one of my favorite parts of the whole process. (Ears are just weird looking and fun to draw)

With the ear done, my sketch is complete.

Interestingly, this completeness is a bit problematic: Because the sketch feels whole (and, from my perspective, represents an interesting, standalone piece of art), I struggle to continue working on it.

The portrait just feels balanced at this point. As soon as I start adding tonal values, that balance will be disrupted, and won't return until I'm nearly done with the whole portrait.

It almost feels unnatural to add tonal values to the sketch, as if I'm defacing something I worked hard to create.

Nevertheless, I must continue. So, here I go… Time to temporarily deface my work.

Defacing my portrait

I start by blackening one of the eyebrows. This is easy, and hopefully will help me build momentum.

I continue with my black pencil, darkening the other eyebrow and the hair.

I can't seem to easily get the hair to be one smooth black mass. Instead, the grain of the paper is very noticeable, giving me a nice salted look. Even after aggressive blending with a blending stump and a dry brush, I still can't get the material distributed nicely on the paper.

I may need to invest in some powder graphite (but I'll return to this later).

Next, I start on the prominent eye. This is where the real defacing starts, as it's going to be a while until it doesn't look like I'm wearing makeup.

After many more minutes of work on the eye, I stop for the night. I'll continue more tomorrow.

Today, I spent a couple hours working on the eyes and nose area of my self-portrait.

My tonal approach is noticeably different than that used on the Derren Brown portrait.

With Derren, I wanted to ensure the portrait emanated three-dimensionality, so I pushed aggressively on the contrast of the portrait. I also didn't care much for the micro-gradations of shadow/light, as I was more concerned with the correctness of the bigger shapes.

As a result, the portrait definitely has a stunning roundness, but I wouldn't call it photorealistic.

Thus, this time around, with my self-portrait, I'm aiming to more closely match tones, while also paying attention to the smaller areas of light fall-off. With this attention, my hope is to create a more realistic rendering of my face.

It's still hard to tell whether I'll be successful, but we'll find out soon…

In most of my posts, I tend to be pretty positive (i.e. "Whoa, today went better than expected…", "I'm really pleased with today's progress…", "I can't believe how good this is…", etc.).

This is mostly because I'm very bullish on this entire project.

However, in my past three posts (I made a mistake, Intentionally defacing my self-portrait, and Fighting for photorealism), I've tried to interrupt this trend, and share some of the day-to-day challenges I face.

While I am still very positive about this project, and happily take on the micro-challenges, I thought sharing some of these things would be more interesting than writing about how every day is always better than the last.

Anyway, continuing with this theme, today, I want to share an interesting struggle.

The Light Situation in San Francisco

For some (perhaps, legal) reason, most apartments in San Francisco don't have overhead lights in their main living areas. Usually, apartments only have overhead lights in the bathroom and (sometimes) the kitchen, which is the case for my apartment.

As a result, the rest of my apartment is lit via Ikea floor lamps, which, although they do a 90% good job, it turns out, at night, there's just not enough light for detail-oriented drawing.

During the sketching phase of my self-portrait, I didn't need to see precise tone, so sketching at night was no problem.

However, now that I'm trying to carefully model the lights/shadows of my face, I need more light.

I considered drawing in the bathroom, but this isn't entirely comfortable. Especially because I was worried that the portrait would get wet/damaged on the sink, whose counter is the most viable drawing area.

Since, without deconstruction, the kitchen table doesn't fit through the bathroom door (I tried…), I needed to find somewhere else to work tonight.

I ended up across the street from my apartment at a well-lit coworking space, which was great for drawing, but not-so-great for picture-taking. The abundance of overhead lights meant that, however I positioned my body, I was always casting a shadow on the portrait.

Thus, once I finished drawing, I came back to my dark apartment to snap a photo.

After my light-seeking adventure, here's what I was able to accomplish.

Today, I only had ten minutes to draw, so I spent all ten darkening the hair and eyebrows on my self-portrait, until they were as black as I could get them.

This greatly improved the portrait in two ways:

The relative tones of the face to the hair are much more accurate now, which helps with realism.
The shape of the hair on the left side of the portrait wasn't quite right, so this gave me the chance to fix it.

Here's the before…

And the after

At first, the blackness of the hair is a bit jarring, but it accurately represents the "exposure" I'm going for (where the hair is emitting no light, and thus, shows up as pure black).

Although today's darkening session improved things, the portrait still seems a bit odd and unbalanced because of the nakedness of the mouth and cheek. I'll start tackling those areas tomorrow.

Yesterday, I declared that today I would start working on the mouth and cheek areas of my self-portrait. And yet, somehow, the day is over, and the mouth and cheek areas are still naked.

Instead, I got caught up making micro-changes to the parts of the portrait I've already worked on (the eyes, nose, forehead, etc.). It seems I can make small improvements forever.

This is clearly not the right approach. Especially because… As I begin shading the mouth, I will need to make adjustments to the nose area, so everything fits together. As I begin shading the cheek, I will need to make adjustments to the eye area, so everything fits together. And so on.

Perhaps, I'm just stalling out of fear: Once the mouth and cheek are developed, I'll have a much better idea if the portrait is any good.

If I am fearful, I definitely need to get over it.

To do so, tomorrow, I'll focus, not on perfectly detailing the mouth and cheek, but instead, broadly blocking in the right tonal values.

With the general tones in place, I'll have enough momentum to push the portrait towards completion.

Today, I spent an hour developing out the rest of my self-portrait.

It went from looking like this…

To looking like this.

It's starting to look like me, but it still looks like a drawing — mostly because I haven't blended the newly developed areas like the neck, cheek, mouth, ear, forehead, etc. Pretty much the whole thing.

I've been holding off on the blending because my blending stump is unusably dirty.

Tomorrow, I'll go swing by the art store and pick up a few fresh ones.

I picked up some new blending stumps today, and went to work smoothing the value changes over my face and neck. Here's the result…

When compared with the before, the difference is pretty striking. In the before portrait, I look like a sickly, pencil-sketched version of myself, while the after version has a much nicer roundness and weight to it.

Tomorrow, I'll make some minor tweaks, sign it, and hang it on the wall.

24 days ago, to kick off December's challenge, I tried to draw a self-portrait.

Then, over the next 3.5 weeks, I completed a 10-hour drawing course, drew a few other people, and then spent 8 hours on a new self-portrait.

Here are the before and after.

And here's a time-lapse of the 8 hours of drawing.

I'm happy with the result, and actually think the self-portrait looks a lot like me.

Tomorrow, I'll write up a more thorough critique. But until then, I'm declaring this month's challenge a success.

Yesterday, I declared this month's challenge a success, noting the differences between my before and after self-portraits.

And while my most recent self-portrait is a major improvement, and does look very much like me, I still do have some quick critical thoughts on it, which I've broken down into two parts: 1. Likeness and 2. Artistry.

1. Likeness

Overall, the likeness is strong. The portrait unequivocally looks like me. Although, it isn't perfect.
My expression/emotion in the portrait is plausibly mine, particularly in the eyes.
The shape of hair near the ear and back of the head is very accurate. However, the hair line doesn't seem completely right, and it's probably the second biggest reason why the portrait doesn't look perfectly like me. The hair line should probably come down on the forehead and should be less rounded. When I snapped a photo of myself (on which I based this portrait), I had just gotten a shorter-than-normal haircut, which is probably why I'm not used to the haircut I drew.
On paper, I feel I captured the nose perfectly, but, as a result of the shadow, it may seem slightly too small/short. To address this, I could have accentuated the tonal difference between the cheek and the shadowed part of the nose, but I wanted to remain as tonally accurate as possible and chose not to.
I'm very happy with how the neck turned out. Its weight and main features (the Adam's apple and the notch at my collar line) seem accurate.
There is something odd about the ear. It seems a bit out of place.
The eyebrows may be the slightest bit thin, but they are very close to reality.
The biggest potential miss is my cheek. While I do have prominent cheeks when I smile (which I'm not doing here), I also have a fairly slender face and a reasonably defined jaw. Depending on how I look at the cheek, it sometimes appears too round and too full. Other times, when I look at the portrait, my eye renders this area properly. If anything, I probably could have made the bottom of the face (in the rolling shadow) a bit more angular.

Nevertheless, even with these critiques in isolation, the portrait as a whole comes together nicely and captures a strong likeness. Thus, I've left it as is, since I care more about an overall likeness (versus a non-cohesive collection of individually accurate features).

2. Artistry

Before I drew my self-portrait, I drew a portrait of Derren Brown.

This portrait has two big advantages over my self-portrait: 1. The tonal range over the face is much greater, and 2. The midtone of the face matches the tone of the paper.

With my self-portrait, I strayed from both of these advantages. For one, on purpose. For the other, less so.

1. Narrow tonal range

Purposefully, I chose to base my self-portrait on a photo with a tighter tonal range, since I wanted to challenge and push my abilities (Drawing a portrait with heavy contrast requires less subtly and is, in my opinion, easier).

Arguably, the contrast of the Derren Brown portrait makes it a more visually compelling portrait, but this is another topic completely (first, I wanted to master accurate portraiture before tackling well-composed portraiture).

Even with the narrow tonal range, my self-portrait still maintains a believable roundness and depth.

2. Dark midtones

Less purposefully, I chose a photo where the midtone of my face was darker than the paper.

This was a bit of a mistake, but a good learning opportunity. As a result of this decision, unlike with my Derren portrait, I had to pencil-shade the mid-tones on my face, leading to a slightly dirtier portrait. (In the case with Derren, where there were midtones, I left the blank paper untouched and clean).

Especially before I smoothed out my face, it looked as if I had just been cleaning chimneys.

While the Derren Brown portrait (with its ultra-contrasty tonal range) may be a more dynamic portrait, my self portrait seems closer to photorealism, which is the main improvement I was aiming for.

Overall, I'm very happy with the result.

After spending nearly a month learning to draw portraits, I'm more convinced than ever that anyone can draw. Even if you don't have any artistic talent.

To me, drawing is a bit like doing your laundry. Before you do it for the first time, you feel it's much more complicated than it actually is, and thus, you feel incapable of trying. Then, you're shown that doing your laundry is only a matter of putting your clothes in the machine, pouring in some soap, and clicking a button. Much easier than you thought.

It turns out drawing is very similar. From the outside, it seems much more complex than it actually is. However, once you learn the two or three basic principles, drawing (at least, at my level) becomes nearly as straight forward as doing your laundry.

In fact, in order to draw a reasonable portrait, you only need to know the two following skills:

1. Triangulation

2. "Outside-in" Shading

Once you're equipped with these two techniques, you'll be ready to follow the "Portrait Drawing Cheat Sheet" and draw your first portrait.

You'll be surprised at how well it goes. I know I was…

As I mentioned at the beginning of this month, British illusionist Derren Brown originally inspired me to start drawing portraits. In fact, to acknowledge this inspiration, Derren was the subject of my first portrait.

However, Derren didn't inspire me with his drawings, but rather, his paintings, like these…

Of course, these paintings are built on a prerequisite foundation of drawing, but they also introduce a whole new skill set that I would love to cultivate.

Watching Derren paint, it seems like there are clear parallels between shading a drawing and painting a portrait: He sets a mid-tone color, adds the lights and darks, works his way towards the middle, and then adds detail.

There are also clearly major differences, like evaluating and mixing colors, general painting hygiene (letting paint dry, etc.), and best practices I'm probably not yet aware of.

And while this seems like a major leap from my drawing studies, I now have the artistic confidence to attempt a painting like this, without any (or very little) additional instruction.

In the coming months, I plan to start sketching a portrait on canvas, and then experimenting with paint.

Last month, I memorized a shuffled deck of cards in under two minutes, which required obsessive, consistent practice. If I were to stop practicing, over time I would lose this skill.

However, I don't think the same is true for my newly-found drawing skills. Mostly because… I didn't learn anything new this month.

Well, that's not exactly right. While I didn't cultivate any new drawing-enabled motor skills or artistic skills, I did learned to structure my already-existing skills inside of a better drawing process.

In other words, if I can remember the process, which, in my opinion, only depends on two very straightforward insights, I will always be able to draw at the level I can now.

In 20 years, even if I don't practice from now until then, as long as I can remember triangulation and outside-in shading, I will be able to fully replicate my results from this month.

I think that's a pretty cool thing, so look out for my Medium post in 20 years.

On December 1st, I drew this.

26 hours of practice later, I drew this.

In other words, after practicing for about an hour per day for 26 days, I majorly improved my portrait drawing skills.

Last month, it only took me 22 hours to become a grandmaster of memory.

I think this is going to be a theme for the entire Month to Master project: If my practice is deliberate and consistent, it's going to take a lot less time than expected to master these seemingly expert-level skills.

The trick, then, is to create a mechanism to force deliberate and consistent practice month after month. This is the hard part about learning these new skills, not the time required.

Something to think about as you start planning your 2017 resolutions…

Today, I flew from San Francisco to Florida to meet up with my family for a few days. I'll be here until January 4th.

I left all my drawing supplies behind, so I'm definitely not drawing any more this month.

I did, however, bring a Rubik's Cube with me in preparation for January's challenge (which starts in two days).

I'm definitely eager to start a new challenge, since I like the idea of always being in pursuit of something (which maybe suggests that I need to learn how to relax). Nevertheless, instead, these past two months, I've finished both challenges on Day 24 (of the month), and thus, needed to wait, without a challenge, for a week, until the next one began/begins.

Should I just start the next challenge once I finish the previous one? I'm not sure. On one hand, this seems reasonable and time-efficient. On the other hand, there is something very tidy about starting on the first of each month.

Clearly, I have some amount of obsessive compulsiveness going on, but I'm curious to know what you think…

Should I wait for the first of each month to start a new challenge, and enjoy my few days of relaxing (if available), or should I just use my extra time towards future challenges and start immediately?

Let me know.

Today, to celebrate the New Year, I decided to compile my personal highlights from 2016, which includes Month to Master, but also everything else from my life.

Rather than writing another M2M post today, I'll encourage you to check out that post if you're interested.

This post is part of Max's year-long accelerated learning project, Month to Master.

Max Deutsch is the co-founder of Monthly — an online education platform that partners with some of the world's biggest YouTubers to create one-month, highly-immersive online classes.

If you want to follow along with Max's year-long accelerated learning project, make sure to follow this Medium account.

For exclusive content on accelerated learning, discipline, and lifestyle design, subscribe to my once-in-a-while newsletter.