Aunt Herbert的论坛贴

What I am currently experimenting with, and I don't know if it makes sense or could be recommendated, is measuring from memory.
The set-up: I have this site here open, and some youtube-video, usually some music video with boring visuals, like a concert or something.
I start a figure drawing, look at the figure for a few seconds, but before I start drawing, I switch to the youtube-clip, so I can't see the reference any longer, then I try to draw what I remember.
Initial comfort level 1 out of 10, can't recommend, feels about as pleasant as intentionally hitting a wall with your head repeatedly, while chewing on glass..
I spent quite some amount of time learning how to measure from a reference, but there is always the risk of getting lost in detail  Now I expect myself to be able to draw at a certain halfway consistent quality from a reference, and the outcomes I produce with this method are naturally abysmal and disappointing without end.
On the upside, once I am done drawing the few lines, that I can remember, I can still switch back to the reference and compare results. And the comparison does give me a lot of feedback about my methods of measurement. For the sake of clarity, with measurement I mean about the widest range of techniques possible. Learning how to draw an abstract paperdoll of a human body absolutely is just one more method of measurement, like is getting known to tyoical proportions, and I guess this way of "measuring from memory" doesn't even make sense without some of those available.

What I hope to achieve with the attempt is to one day just go to a park and draw from people, that are actually moving around, and still capture the scene, without either having to first unalive them to make them hold still, or at least having to take a photo, and then using my bad photo as the reference instead of the sight before me.

It is quite a hard check against trying to include too much detail. It's just astonishing how hard it is to even remember faintly the positions of the main masses or major limbs after staring at a reference for a few seconds, although in my mind, before I start drawing, I feel like I remember all of it perfectly. Then the pen hits the paper, and all of that confidence bursts away in a heartbeat once I cut myself off the physical reference and rely on the clarity of the "reference" in my memory.

I think this feedback has impact on what I draw when I draw olschool with the reference "on", as I get refocused on essentials over details. I generally dislike the terms "flow" and "stiffness", because they are usually highly ambiguous, lead to a metric ton of misunderstanding and make beginners just attempt dumb stuff to produce "better flow". But I think the question of essential versus detail is valuable.

Like, it is the undeniable quality of each poem ever, that it is astonishingly easy to remember, given the length and complexity of the text, as poems are in essence a mnemonic device developed by pre-literate cultures. Making visual art "stick" to memory is dependent on creating mnemonic techniques, and I think lot of the misguided talk about "flow" and "stiffness" is really aimed at achieving that.

Well, 'weird' is a very good and useful description for a problem many artists face. We start drawing something, and it feels all good while we do it, but when we look at the finished result, it feels like it doesn't meet our expectations.
So, if "weird" means, "did not meet expectations" what were those expectations to begin with? Are those really ours? Did we really have them clearly when we put out first line on the paper? Why and how did we fail them then?

The list of artists I have shown above has 2 commonalities for which they are famous: 1. They come with really easy to follow instructions (draw a circle of that size, draw a line in that direction... and 2. They produce results that match cultural expectations.

If you came from manga, then you already trained one such type of expectations. If you already looked into Loomis, then you already know a second set. These guys above offer yet more sets, focused on different aspects.

If you really feel, that the problem with your stuff is that it looks "weird", then the obvious answer is, learn copying one of these. If you really master reproducing one of this techniques, your portraits will be the opposite of weird. To the point, that they will be downright boring. If you want to make money by offering drawing as a service, boring isn't bad, it's reliable quality. But ofc it also stifles artistic growth.

So, if your aim is not to overcome your "weird", but to master it, and get to your own type of weird, don't treat any of these tutorials as gospel. These are toys, and you aren't done playing with them, until you can break them apart and reassemble them. Loomis measures the head differently than mangaka do, so does everyone of these people. And they are different to each other, but none of them is simply right.
You can end up with "same-face-symptom" after grinding Loomis heads the same way you can have "same-face-symptom" after dedicating yourself to manga. Bridgman has another perspective as well, but he doesn't overcome the problem, that while learning to draw exactly that way may overcome the problem of "weird", as you will be able to exactly produce the outcome you expected, if you just repeat it enough, you can't overcome "boring" that way.

Once you are familiar with a few of them, (and you know 2-3 by now) and so you know a few methods of measuring the head, you can try to compare how they approach the subject differently. You might not have the confidence to say "mangaka draw bad" (and it wouldn't be true), but you can now say, that mangaka eyes look different from Loomis eyes, bigger, and more childlike.
The next step would be to make the same comparison for the Inabatewil abstraction of the human head. If you don't think about any of those big cultural icons, and just start putting marks on paper, you are bound to follow some pattern, as human beings just aren't random.
You know manga, you know Loomis, you know Bridgman, observe Inabatewil, until you can compare what you naturally do, with what these others do, so you can make more informed decisions about what you want to do.

And then off course, you can compare any of those abstrations to references and will realize, that none of them really "naturalistically" fit "the reality". Because actual human faces are always weirder than the abstraction, unless the photographer and the make-up artist already put a ton of work in to normalize them.
Your drawings will stop feeling 'weird', once you settled the decision of how you expect the outcome of your drawing process to look. And weird outcomes are just the sign, that you are clearly not settled yet, clearly still experimenting, and there is only so much you can learn from other in this phase.
Best advice, stay weird, until someone starts throwing money at you for reproducing the style you developed for yourself.

Drawing portraits from "imagination" basically means, that you are following a pattern of construction for a human face, often called an abstraction. Either one, that you completely came up with yourself, just from randomly associating your experiences of drawing with the experience of looking at a reference, or that you completely learned an abstraction someone else originally came up with by heart, and most likely than either of those, a mix of both.
You probably organize your attempt of drawing a face by some vaguely culturally known pattern you saw someone else do, or you interpreted an abstraction someone else came up with along the lines of your own experience.

If you feel like you are doing a lot of practice, but your experience doesn't seem to develop the way you want it to, I would suggest you try to train one of the popular abstrations for a bit, so you have a basis to decide how much of them you want to include into your own style.
Here are examples of the most popular I am aware of. With a bit of search engine fu you should be able to discover free tutorials for each of them on the webs: Andrew Loomis, Frank Reilly, Henry Bridgman, Steve Huston, John Asaro,
or even japanese industry style, which sacrifices natural proportions for style, and is mostly shared by a lot of japanese artists, look for Manga or Anime style.
https://line-of-action.com/forums/users/190807/replies?page=4
I am a bit lazy with the links. This link should lead to old posts from me, and if you scroll a bit down you should see examples of the results these different artists are going for. Generally Andrew Loomis is the first go-to, but you might already know him. If you already know Loomis heads, but feel, like they aren't doing it for you, you can check out those other artists. Becoming familiar with more than one of them also makes you more familiar with the entire concept of using abstractions for "drawing from imagination", which can help you you reflect on your own way of drawing faces and heads.

Generally, these are all artists, that come towards drawing faces from lines. There are obviously completely different approaches towards portrait painting, like approaching from value, but unless someone insists to read me write about that, i won't, because I barely know, that they exist.

The difference is, that you expect a lot more from your human figures than from your animal figures, because you know them much better. 

You are more familiar with details, so it is harder to abstract from them. If your drawing of a frog makes the audience clearly recognize some sort of a frog, then that is satisfactory. With a human, you know how to distinguish a specific gender, a body type, an action, a gesture, individual traits, specific anatomical properties and functions, down to the level of micro-expressions. And because you intimately know those distinctions, it is far harder to abstract from all of them to simplify into geometric shapes.

If you want more "flow" in that regards, you don't have to just "practice" perfectly combining precision and expression, you also and foremost have to start negotiating with your own sense of taste, how much precision you are willing to sacrifice for the sake of expression. 
Like in, really experimenting with pushing the boundaries, by drawing in a way that leads to unpleasant results, because it feels like cheating or overly simplistic. That way you get both a sense of how far you are willing to go into one direction, and what the aesthetic costs are of going into that direction as far as you can bear.

I had the experience of trying to follow some very gesture centric artist's tutorial. At first I thought, that I did not understand their ideas. But when I looked closer and tried to match stroke for stroke how they approached a specific reference, I became aware, that my problem was not so much a lack of skill, but a lack of willingness to sacrifice the amount of detail, they gleefully ignored. 
They are clearly proficient artists, but I rather stick to my specific "stiffness" than to adopt the amount of generic typification they produce, because I am still in love with these details, and with observing, examining, and portraying them.

What you did followed a clear idea, and you worked diligently enough to not distract from the idea. You traced the shape of the model with clean simple lines, adding contour lines to indicate volume. You abstracted a lot of details, like the specific clothes, the minute curvatures produced by fat tissues and muscles, and the lighting in the reference, to produce an abstract grid. Even if you wouldn't have underlayed it with the photograph, the grid you developed would give a very exact impression of one way a human body can relax, while laying back in a lazy or contemplative mood on a chair.

This is certainly a very valid way to simplify the reference you worked from.

I can't tell you whether you "nailed" it, as I am not entirely certain, whether you have a specific goal in mind going forward, but doing this exercise must have deepened your understanding about the human body. I have my own specific taste in how to develop a figure, and I watched and read enough explanations from other artists about their methods, so I could try to compare your approach to the reference to the way someone else might do it. I am just not certain, whether that would help or hinder you on your path.

TL;DR I find your result beautiful.

OK, first of all, content warning: wild speculation from some old dude on the interwebs, who so far earned less than 50 bucks in total by selling his own art.

Now, after that is out of the way, let's talk about things I heard from other people, that sounded convincing to me, and that I long decided to start with soon, but keep pushing away. When you say figure/gesture drawings I assume you are mostly talking about drawing from reference. Which is a good starting point, but if you want to be able to get towards most of the stuff, that you are talking about, you need to develop further into drawing from imagination.

That might sound strange, because isn't that what you did anyways, before you started drawing from reference(?), and no, it isn't exactly the same. The difference is, that you have to learn to draw from imagination at the same level of precision, as if you would be drawing from reference. The way to get there, is to first learn how to modify the images you drew from reference. So, you learned all about gesture, and the way the human body is constructed, and you can prove it by really nicely copying this photo? Level up, try again, but this time from a different point of view. The model on the reference is from side view, how would it look in the same pose in 3/4 view, in front view, in back view, drawn from above? To do that, you really need to up your construction game. That hip is here, that chest is there, so if I turn them by x degree, they would be....?

A different modification, OK, this model has their arms raised, how closely can you picture the same model, but with lowered arms? How much would that impact shoulders, muscles, would it have influence on the torso?

Or, same model, same pose, but now the light is from a different direction.

If you can freely rotate and modify your models, then you are at a stage, where you can react to someone handing you a photo and telling you: I want you to draw this guy, but jumping like a frog. I want you to draw that gal, but beating up a boxsack.

Which, I would imagine would be a lot of the actual job descriptions you might professionally get. Unless you approach the job from the other side, and want to become the person who tells other people to draw that guy, but jumping like a frog, but then you have to learn the fine art of storytelling, how to write a scene, how to structure a scene, what makes a good narrative, etc, and then you will probably get better advice on other sites.

To me, the biggest hurdle is to go back to the beginner stage, and start to focus on structure, structure, structure. Instead of practicing to draw as pretty as I can, really trying to use 3-d shapes, that are simple enough for me to rotate. like boxes, tubes, and the occassional sphere, and reconstructing all references with that stuff. then trying out if i can turn them around.

Maaaaybe you knew all of that already, then I have to ask for your forgiveness for blathering. About actual art careers, I can't seriously answer you any questions, as I have less of one than you at this point.

OK, to respond to popular demand, let's describe a practical example. Today nothing came up, so I don't have a scan to illustrate. For example: in principle I am doing a series of gesture study. A model comes up, I find her smile really intriguing, it captures my attention. In principle a facial expression isn't part of a gesture study, but it is what caught my eyes. I see, that the way she holds her head frames her expression in an interesting way, so I definitely want to show her head, neck and shoulders. Her hands seem also important. Her legs on the other side, yes, they are there, and she is a model, so they aren't ugly per se, but they don't do a lot for me that supports her smile, and I just don't extend the drawing beyond her waist and focus on just her head and upper toso instead.

Deciding not to include her hips and legs off course is already a massive simplification of the drawing, but it isn't deducted from general principle, but from following my own sense of beauty. Someone else, or even me on an other day, may have decided to only draw her feet, and it also would be a valid decision. The various technical drills we are practicing also introduce us to more things to spot, and to better understand what we are seeing.

Maybe the long curving line from her neck along her torso along the hip towards her knee might be something that catches my eyes on another day. The interesting way she twists her shoulders against her hips, etcetera, and the longer I spend on the drawing, the more I will try to include those details, too. In a 30 second sketch, I have to decide for one of the possibilities, and focus on that, and when I continue the drawing, that first impression generally should stay my guide, .... unless I realize during drawing, that a switch in focus would really enhance the result. But that is in practice the rare occassion, and I trained my habit to stick with the first impression long enough to generally trust my brush to find the right path over the paper on its own.

Just in case this helps anyone understand better what I want to say, this is my 30 minutes session from today: I think I found OK lines, but as I am currently between two shifts, have been distracted a lot, and barely managed to get in minimum practice, I did not expect top quality from me.

Maybe another illustration: this should be a link to my sketchbook for winter 2019/2020. By that time I barely knew anything about technique, I think I finished the major drawabox lessons with the homework. I watched a few youtube-videos and saw someone mention intersting shapes, and demonstrate how he copied an interesting detail from a photo of a roof. I found the words "interesting shape" intriguing, and wandered the streets trying to find some simple and interesting shapes to draw: https://www.artstation.com/artwork/eaPaBD

OK, let's make that a triple post then. @Idon'tknow. I told you before, that I find your questions always quite deep and worthy to consider. And the reason for that is, that your problems heavily resonate with me. I sometimes base parts of my answers on assumptions, which can in itself be considered quite transgressive and impolite, but I want to assure you, I am not attempting to pathologize you, and I am not trying to psychoanalyse you through the screen. Things is, I feel a lot of resonance within myself to the worries and problems you express, and I think that they are indeed tied to the general process of producing art, and likely shared to some extent by all artists. Just, you seem to to experience these specific problems more pronounced than other people, which gives you the courage to repeatedly raise them to debate in public.

I have yet another suggestion for what you could try to overcome the problem, this time tied a lot closer to the specific topic, of "how to break subjects down into simpler shapes and how to make my shapes more interesting and dynamic".

One more assumption, I am sorry, but, since you spend a lot of time and effort into becoming a better artist, it seems quite safe to assume that you know the experience of just looking at something, and becoming aware, that it is beautiful. This basic experience is the key to understanding how to assess and apply all that theory. And, becoming aware that something looks beautiful is, like a lot of other stuff you practised, basically a trainable skill. So, the suggestion, start a drawing session with references equaly timed at 5 minutes or so. Don't touch a pen, just look at a reference and try to decide IF you find it beautiful and WHAT makes it beautiful. When you have come to a conclusion, use the forward button to skip to the next reference.

If you spend more time looking at things, and contemplating, THAT they are beautiful, in a practicing artist like yourself, this will eventually trigger the curiosity to understand, WHY they appear beautiful to you, and how you could best express that beauty to other people. Now, beauty can be expressed in words, but unless you are a poet, your words will likely be to bland and boring to truely show that beauty to other people. As an artist, the natural instinct is to take a pen and piece of paper, and draw that beauty.
To discover, that idea to be harder than it appears at first thought. And go look for instructions. But frankly telling someone how to draw is like telling someone how to drive a bike: hard to put into words, even for the best teachers.
The techniques that can be taught all start from the premise, that the artist already sees something beautiful and wants to share it. Gesture, composition, the study of perspective and of shapes, these are all just attempts to describe beauty in a way that is practically useful. If you worry so much about understanding this concepts, that you lose sight of your own intuition of beauty, they become impossible to understand and apply correctly.

Finding the simplest interesting shape isn't a problem of geometry, it's first of all a problem of self observation, phenomenology, if you want. If you look at something, while experiencing beauty, which shapes is your attention drawn to? Teachers, tutorials and instructions can give general guidelines what attributes such shapes can generally have, either universally, or within our shared cultural tradition, like symmetry, clear asymmetry, simplicity, etc.... but if you try to interprete such an instruction without keeping in mind, that it is just a suggestion for you how to express the beauty you experience, than you won't understand it. If your result looks less beautiful to you than before, than either you got something wrong, or the instruction is just wrong for you, and you should try another.

A final word about proko courses: you don't have to pay anything for any of his basic courses, and they are generally complete in conveying the idea. Indeed, if you pay for the premium course, you might be dissapointed, how little extra you get for that. It's basically like a "buy me a coffee" button or a patronage fee, which just gives a few extra perks to show gratitude, only that stan apparently isn't interested in administering 5 bucks at a time.
If you find something interesting, but are short on money, just do the basic course for free.

A follow up suggestion, to make it ultimately easier for you to execute that first instruction. From your forum posts so far, I am almost certain, that you picked up a really bad habit: once you have finished a piece or a lesson, you most likely look at it, judge your result, search all the possible flaws and spiral into desperation about how you failed yet again to improve even the slightest bit.

You should conciously practice to break that habit. When you have finished a piece or a lesson, make an effort to find everything that went right, every idea you started with, that now shows in the result, every aspect of the drawing, that you do like, everything you managed to do better than in previous attempts.

Be fair to yourself and your circumstances, if you otherwise had a horrible day, you can't be surprised, that you didn't produce peek results, instead congratulate yourself to have found the strength to repeat practicing anyways.

Say that stuff out loud, or even better, write it down, to make sure it gets rooted in your memory. Do that every time you finish either a lengthy piece or a daily practice session, until it becomes routine.

If you find it hard to be kind to yourself and your own results, you can practice that by giving feedback to others. Look at their results, find anything you can possibly like or recommend about it, and tell them first, even if their skills are way less developed than yours. Don't focus on spotting flaws, or pointing out what went wrong, unless you see a simple and practical way, they could easily improve some aspect of their drawing. If you do, explain it kindly, after you are done mentioning everything that is already good in their attempt.

Idon'tknow, I think your problem at this point isn't, that you know too little about drawing theory or that you have to watch yet another instruction video. The problem is, that you misunderstand what theory can do for you, and that you are stuck to it like to holy scripture.

I have seen you post quite a lot over time, including the results of entire sessions. I remember one session you posted. Most of your results where extremely simplified, and frankly a bit boring as a result, but there were a few, where you seemed to have slipped up, and veered away from sticking extremely close to any instructions, and those were a league above the rest in quality, but with handwritten comments almost excusing yourself for their existence.

You try to read drawing theory like an algorithm, just a long and complicated series of simple commands, that you have to learn by heart and then faithfully execute to reproduce the intended outcome. That is not how it works.

Drawing theory is just there to point you towards relationships between lines and forms and what-have you, that you might otherwise miss. The repeated practices you do are not a mystical ritual, that will some times in the future allow you to press a button, raise a numerical stat inside you, and thereby grant you access to the follow-up set of instruction for advanced characters. The reason you repeat that stuff so often is, to make it come automatically while drawing, without having to think about it.

Here is one instruction from me to you, but I am afraid, given what I have seen so far about your relationship to yourself, you might find it hard, almost impossible to follow. Chose one subject that you really find beautiful, and just try to draw it as good as you can. DON'T try to apply anything you learned or read or seen, don't put yourself under pressure of timing or how to "correctly" simplify stuff or anything, just be confident and focus on the drawing itself. If you become unsure, whether you have done something right or wrong, don't fret about redoing it or anything, just accept it for the time being as what it is and continue developing the drawing from there on. If you feel like your drawing doesn't look close enough to your initial idea, don't fret about it, accept your preliminary results and continue working from them as best as you can, until you feel the piece is finished.

The amount of time you have spent consuming theory, analyzing theory, doing repetitive practice to hone your skills according to theory, it has already taken root and shows results. Stop panicking about whether you draw correctly or wrong, just draw, and your experience will show anyways, you don't have to constantly monitor yourself, to make sure you keep injecting enough of it into your process, or to repeatedly judge every mark you have put on paper to prevent yourself from heretically deviating from orthodoxy.-

You did a good job sticking to the tutorial. Trying to apply someone else's ideas to your own drawings helps you find new eyes for drawing, and begin to formulate new ideas about how your drawings work. And yepp, gesture drawing lessons are a bit of a riddle, as if you look at different teachers, they all propose different approaches and priorities, but still have something intangible in common.

As a self-taught artist, you'll never get as much feedback or firm orientation as you would like to have. You just select one thing and stick to it, and watch yourself to understand what changes. Sometimes you will experience rapid and satisfying improvements, at other times you will feel like you are stagnating, or even feel like your skills diminish, because doubts about what you have thought about establshed rules come about and challenge your confidence.

To me the 30 seconds/1 minute/5 minute/10 minute split of LoA's lessons feels like a setup for experimentation. You develop a foundation basically in the first minute, but the 30 second split forces you to develop it as a sequential process. "So, I will always observe/measure/focus this first, then this second to build the foundation" And then the 5 minute pieces are the proof of concept, where you discover what worked and what did not work.

OK, first one, and I don't know how important that is for you, but what you are doing is more a mix of gesture and form than pure gesture study. I do the same, so that isn't a horrible flaw, just an observation and a bit of nitpicking about language. A purely gestural analysis would be an almost exclusively linear study of the tensions and balance in the figure. Which in a puristic form often doesn't even look so good, because it doesn't convey form or perspective. If you are interested in separating gesture and form, Michael Hampton does it a lot, but be warned, he also purposefully deviates a lot from observation, and his results seem sometimes rather losely inspired by reference than drawn from reference.

A very specific point on the kneeling male figure in the second image, and that is about the position of the head. It is quite distinctly above the shoulders, which looks awkward. This specific model, and others with that body type, often has his head sitting rather low between the shoulders, with the neck more diagonal or even leaning towards horizontal rather than straigt up vertical, and the face extended forward over the chest. I also think you drew the hip too high, shortening the upper body and giving the legs a strange rabbit-like tension.

The trick with the videos is mostly to chose documentaries or video essays, that you don't have to watch. You mostly listen to people talking about various things, either stuff that interests you, or stuff that doesn't actually interest you a lot, but is beautifully said, like a lot of video essays on contemporary art and/or philosophy. I mentioned looking at AI-images of Psytrance videos, but that is only, when I do repeated lessons over several hours, and then the looking happens only in the breaks between those lessons.

Well, your solution IS quite accurate. As is your observation, that youtube tutorials tend to use references that make it easy to apply the technique the makers want to demonstrate.

I think the answer is, if you are most interested in a perspective depiction of, in this example, a hip that is almost vertical to your point of view, then the stickman method isn't the perfect tool. You CAN still use it, and I actually think you got the most out of it in that instance, it just won't produce a very satisfying result (as witnessed by the fact, that you express your dissatisfaction). Tools have limits, and you can push those limits a bit with a lot of practice, but you can't make them disappear.

You are "supposed to draw" it to best encapsulate your desired artistic effect, and if one tool doesn't cut it, switch to another tool.

Frankly, your lines do actually decently indicate volume. Which makes it a bit hard to come up with the one silver bullet tip, that is guaranteed to blow you to the next level.

The big youtubers always push towards more simplification, although I must say, it drives a tear into my eyes to ignore for example Jorges neat muscle contours to achieve it. In case you need a reminder, here is a very short clip, that lists 3-D primitives for mannequinization:



I while ago proko had this idea to completely express a pose (sans limbs) with only a box for the hip, a box for the chest and a box for the head. Sounded silly at first to me, but when I tried it out it indeed gave me some more insights about the orientation of the masses. You might want to try it, too.

As far as I have understood the idea, the way this simplification stuff "adds volume" isn't about changing so much about the lines you are currently producing when drawing from reference (apart from simplifying them and thus losing a lot of detail), it is more a preparation to go beyond the reference and allow a basis for manipulation. Like in, draw a pose simplified enough that you can reproduce the same pose, but from a different angle. And that is actually painfully hard.