Want advice in how to learn pixeling

[Uhfgood]

I will use all your advice, and like I said I'm not disregarding anything but I'm asking a specific question.  You know an obvious progression which would probably govern what I do is resolution.  I could start with 8x8 images and then move up to 16x16 and 32x32 and 48x48 and so on until I'm at a certain resolution.  So these are specific progressions I could start with.  Do I start out with house hold objects, or tiny landscapes.  Am I being clear at all?  I feel like I must be miscommunicating.

wrong. the resolutions don't differ in terms of difficulty. They difference of them lies in how difficult it can get to illustrate things with fewer pixels readable. The bigger stuff it the longer it takes to render with pixels, but that has more to do with mechanical work than with difficulty.

So to address what you said here I was simply giving an example of a progression, not specifically suggesting I do this particular progression.  The example is clearer because there are steps 8x8, 16x16 and so on NOT that I'm asking should I do resolution, I'm asking what types of pieces should I work on that go from easier to more difficult.  You did mention taking a reference photo and sketching that 10 or 20 times, and that human faces were more difficult.  So that's kind of part of an answer.  That's sort of what I was looking for you said "try something like that because it doesn't have any faces", as I wanted a list of things to draw (or in this case pixel, but it's still applicable). 

In any case thanks for your time, and I apologize for the trouble.

What I really wanted was something like this article - http://www.gamedev.net/page/resources/_/technical/game-programming/how-do-i-make-games-a-path-to-game-development-r892 -- In it Geoff Howland lays down a path to development by having you complete specific games.  First he says tetris, then breakout, and then further.  This is what I'm trying to figure out with regards to pixel art.

Wow, don't ever do this. Don't ever recommend this to any new programmer - you will put them back so many years if they try do things this way.
But this does bring some nice analogies.

The article IIRC isn't recommending a someone who never coded before to jump in and make tetris.  He was suggesting different games to make in order of (in his opinion) complexity.  So if you know programming but want to start making games, tetris is a good one to start with because of lack of AI, and so forth.

So now that I've explained that, you may ask more specifically. Do you want us to tell you how to learn traditional art before delving into pixel art, or are you already capable of doing traditional art? or do you want us to tell you which kind of perspectives are easiest and what kinds of games to copy to kind of get an idea of how to make generic pixel art (really slowly)?
If you just want the latter I really doubt people here are going to be willing to help, because they take a lot of pride in not only their work, but also their advice.

I'm capable of doing traditional art, and have done some in the past.  I'm not what you would call an accomplished artist, but I'm not entirely a beginner.  I'm not asking for either of these and apparently, again, I did not communicate what I was asking very well.

I can't make it any clearer than "What do I make, in steps easier to more difficult.?" not "how do I make *x*".  There is some mistake I made in how I worded my original message.  When I said 'start from scratch' I didn't actually mean start back at the beginning when you know nothing about traditional art.  I meant to learn concepts what specific pieces would you make in order, from easiest to hardest?

In any case I won't bother you guys any more since there's no point when you don't understand what I want help with.

I apologize for taking up your time.

[Joe]

Well that's interesting. I didn't think that's what you were asking because there's no need for that kind of artificial progression. It's really very simple, you have this whole set of traditional skills, and a much smaller set of pixel skills to learn, and an order to learn them in. I think we all agreed on that.

Subject order is largely irrelevant to artistic growth. If you want to do shape studies you could get the same value from a face, a flower, or the landscape Cyangmou posted. The order of subjects doesn't really matter, it's really a personal preference. Are you planning on making a lot of graphics containing people? Practice with people. All I would say on that matter is to not stick to one particular subject, a lot of people tend to get "comfortable" and stagnate.

It is not much more difficult to draw a human face, rather, it is more apparent that a human face is incorrect than a plant. Both are just drawing what you see.

Subject order is irrelevant. Concept order is what matters, and much more important, which is why each answer addressed it.

[Uhfgood]

Well that's interesting. I didn't think that's what you were asking because there's no need for that kind of artificial progression. It's really very simple, you have this whole set of traditional skills, and a much smaller set of pixel skills to learn, and an order to learn them in. I think we all agreed on that.

Subject order is largely irrelevant to artistic growth. If you want to do shape studies you could get the same value from a face, a flower, or the landscape Cyangmou posted. The order of subjects doesn't really matter, it's really a personal preference. Are you planning on making a lot of graphics containing people? Practice with people. All I would say on that matter is to not stick to one particular subject, a lot of people tend to get "comfortable" and stagnate.

It is not much more difficult to draw a human face, rather, it is more apparent that a human face is incorrect than a plant. Both are just drawing what you see.

Subject order is irrelevant. Concept order is what matters, and much more important, which is why each answer addressed it.

At least you understand what I wanted, however, subject order might be relevant.  Take for instance this progression, humanoid cartoon characters -> realistic humanoid comic book characters -> actual people -- all using photo references.  Since the cartoon characters are the most simplified versions of actual people, it should be less difficult than doing the realistic comic book characters which should be less difficult then drawing actual people.  For me this would be the case at least.  But with each step I should become better or more efficient at it and then soon the actual people won't be that hard anymore.  Since you're trying to draw what you see this progression would also be helpful with that, since cartoon characters are already abstracted and you're not using the symbol system quite as much, and focusing more on the actual lines... then when you got proficient with that, moving on to comic book characters would do the same thing only now I would be able to learn proportion and more realistic anatomy.  By the time I got to actual people the leap wouldn't be so big.

What do most people do when they first start out, they start out as kids copying their favorite comic strip/cartoon characters, then when starting to get in their teens they start copying comic books... by the time they get to proper art classes they're already pretty good artists and use the proper art classes for refinement.

I also apologize for being lengthy in here but I wanted you to understand why I was asking what I asked.

[32]

I think that method of progression is a total waste of time. You want to start with the real people and work backwards otherwise all you're learning is what other people worked backwards to. Your focus in regards to people should be construction; what shapes are people made of and how do they join together. Linework is an expression of a 3 dimensional understanding of an object and it's hard to learn that from a cartoon drawing.

I drew cartoon people for a long time before I started doing life drawing and I don't think that knowledge helped all that much. However drawing the real people made my stylised drawings infinitely better. I think your best bet is to take photographs of real people and draw them any way you like. The important part is looking as much as it is drawing.

[pistachio]

From what I can get out of your (pretty vague) goals you want some kind of comprehensive progression list like this but for pixel art? That's actually sound and it's kinda interesting that you hardly see one set down for pixelling specifically... Ok, we're basically nerds in the middle of a niche... But there are already Kickstarters out there trying it out.

Personally before I pool my money into a Kickstarter or until The List gets made, I'd dump all the knowledge from free resources into a roughly contextualized order, maybe off of some practical introspective notes I took, and pull that knowledge up as I needed it (because it's a free country, man!)

It's usually pretty damn satisfying to conquer your weaknesses instead of keeping them vague and esoteric, in art or work or life or whatever it is. Only with rare exceptions, it's how you move forward. So in that way finding the right, specific tutorials and the right way to use them is a great idea.

Best way to know if you've used them right, is to ask a straightforward guy with experience who points out the problems

I'll respond to more tomorrow.

[Ambivorous]

Wow, don't ever do this. Don't ever recommend this to any new programmer - you will put them back so many years if they try do things this way.
But this does bring some nice analogies.

The article IIRC isn't recommending a someone who never coded before to jump in and make tetris.  He was suggesting different games to make in order of (in his opinion) complexity.  So if you know programming but want to start making games, tetris is a good one to start with because of lack of AI, and so forth.

I see. Well in that case I'd have to say just go straight to Galaga.
The thing is that you are making the game in tiny bits either way. It's not like someone just sits down and codes Galaga from scratch in one try and then hits compile. First you'd just try put an image on the screen, then you move that image, then you add in player controls, then collision, and then finally AI. At every point of that after you've added controls it's a version of one of the other games (and people find it surprisingly fun to play a game that they themselves have made regardless of how pointless or unpolished it is).

The main reason for suggesting any of those games is that they're 2D scrollers which is why they're so easy to make.
You need first year, university maths just to add that extra dimension and that's not even including learning about good collision detection and occlusion - it's just putting a model on the ground. Hell, even adding rotation to a 2D game already means your maths knowledge needs to be decent, but again you can do it in small parts and that will make the progression smoother.

So now that I've explained that, you may ask more specifically. Do you want us to tell you how to learn traditional art before delving into pixel art, or are you already capable of doing traditional art? or do you want us to tell you which kind of perspectives are easiest and what kinds of games to copy to kind of get an idea of how to make generic pixel art (really slowly)?
If you just want the latter I really doubt people here are going to be willing to help, because they take a lot of pride in not only their work, but also their advice.

I'm capable of doing traditional art, and have done some in the past.  I'm not what you would call an accomplished artist, but I'm not entirely a beginner.  I'm not asking for either of these and apparently, again, I did not communicate what I was asking very well.

Actually to me it sounds like you want the former, which is good.

As an aside don't apologise about wasting time or whatever it is that's happening. People just won't reply if they're actually done helping you and mostly there are gruff, jaded members replying (because of how many people come through and ask questions and then just leave without learning anything and thinking everyone hates them), so if you feel like you're being attacked, or upsetting people - that's perfectly normal.
I'm just putting down stuff that I've learned in the hope you will find useful information in it. It's also good for me to write these things down for my own sake, so it isn't wasted no matter what.
Moving along.

Learning will be different per person, but I will give you an order and some reasons and hopefully you can just take what you can out of it and do what you must. If I still haven't understood you correctly, then you'll just have to wait until someone else does.

(To start off, always draw realism first then move on to cartoony. Also, throughout all these practices never use the eye dropper tool on your reference image because this destroys the purpose of the exercise.)

Draw pictures of realistic faces from photos/real life. Lots of faces. I'm talking 10 faces a day kind of thing. Never erase. Never be ashamed.
As was mentioned your brain can see errors in faces really easily because it's trained to recognise them, so if you want to learn how to copy proportions well faces are your best bet (and then hands, then bodies).
Why mustn't you erase or be ashamed? Simply because your mistakes are an important lesson. For one it shows you how your brain interprets what it sees. Often this results in eyes that are too big (your brain puts heavy stock in eyes) and other deformities because that's what your brain focused on most and so they appear bigger to you (it might not be eyes for you). REMEMBER THIS. Later on you'll be taking advantage of this effect to create your very own style (style for many artists being the slight differences that are created as a result of this cognitive bias). The style will come naturally if you just let your brain do its thing.
So then seemingly entirely contradictory to that we now want to iron out those small imperfections. Actually not iron, more like just going over it with your hand because we don't want it entirely flat (or you'd lose your style and your art would be boring). This will come naturally once you see your mistakes and begin fixing them. Eventually you will be drawing realistic looking faces and they will have a special something you've added unknowingly.

You will need to learn values to shade realistically and give things depth. Depth is especially important in pixel art because it is a 2D medium, so all depth has to come from correct representation of light.
To learn value you do value studies. Simply take a picture and greyscale it and then try to copy it as accurately as possible (again take a picture of something real: bodies, faces, landscapes with nice mountains/hills - something with a nice amount of large, easy to identify shapes).
Don't spend long on each picture and zoom out/blur your eyes so that you get the overall shading correct and don't get bogged down by details.
Now compare it to your reference image. You'll notice the same sort of imperfections as with sketching faces. These are your style. Flatten them out a bit, and unlike the sketching when it comes to value you can actually iron them out entirely if you want.

Now that we have those down, let's work on colours. You can go straight into pixel art to learn colours if you like.
Again just pixel something from real life. Take a picture and scale it down to some tiny size (128x128 or something) and have the full version open next to you, and then try to create the image that is as small as the 128 version, but with as much detail as the full image (obviously this is impossible, but you are trying to learn how to represent things at this point more than copy them exactly).
You'll use the smaller image to gauge what colours to use more or less. You could also rather to thumbnail studies to get colour right and then go into this pixelling of scenes, which ever works better for you.

Alright now you are the master of copying realistic images of all types! Well done.
NOW comes the part where you can bring back those imperfections. You don't HAVE to look at any other artists's work, but you can if you want.
Your brain was giving you hints at how it recognises things with those imperfections earlier, so let's just use those to make our images unique and recognisable. Easy!
Draw cartoon faces and shade in a cartoonly manner (one colour for where light hits and one for shadow, or something like that). You will find this easy because your brain was already doing it and now you have mad skills.
Any medium is fine for this, but rather don't use pixel art just yet.

Finally, you are now the jedi-wizard master of representing real life things with believable less-real things.
Just yolo straight on in to any pixel art project. Anything you have been drawing you will be able to pixel and if you want to pixel something you don't know how to draw, learn to draw it first.
Once you've drawn something a bunch of times you will be able to pixel it with such ease that you'll wonder why you ever thought pixel art was difficult.

Further you will need to hone your skills.
Redo cluster studies because this time you will understand why they are so valuable.
Experiment. Never be happy with your current level of skill or you will stagnate. Try new things.

And that's about as much information as I can give with regards to what order to try doing anything. If you're already able to do any of those parts then just do the others, maybe redo a few things in case you've picked up bad habits.

[Uhfgood]

From what I can get out of your (pretty vague) goals you want some kind of comprehensive progression list like this but for pixel art? That's actually sound and it's kinda interesting that you hardly see one set down for pixelling specifically...

Exactly like that, but it doesn't have to be *that* comprehensive but in order of possible difficulty (of course different people see that differently which is fine, but it's good to have something to start from).

I see. Well in that case I'd have to say just go straight to Galaga.
The thing is that you are making the game in tiny bits either way. It's not like someone just sits down and codes Galaga from scratch in one try and then hits compile. First you'd just try put an image on the screen, then you move that image, then you add in player controls, then collision, and then finally AI.

A note about the games thing, I've completed a few so I know what I'm doing there (including the aforementioned Tetris).  I was using the article as an example of what I was looking for.  As far as going straight to Galaga, the point of the article was creating a progression of games to create that added on features.  For instance Galaga has waves in patterns, and a kind of AI.  Tetris does not.  In fact it doesn't have a lot of things so it's generally simpler to implement.  2d collision is one of the fundamentals of game development because you don't have a game if you can't have two objects interact with each other.  If you've read the article, he next recommends breakout because the collision is more advanced, and adds some extra ball deflection stuff, and level layout, then he recommends pac-man because of the AI of the ghosts (simple ai but probably a good recommendation)... And so on.  It's somewhat about scope because doing games with a smaller scope first means you'll gain confidence in completing games.  Any way there are other games you could choose and it could be a different set of games that do roughly the same thing.  Maybe you want to add in Galaga or space invaders in between breakout and pacman.  Because the waves of enemies roughly correspond to the rows of bricks and then there is some basic ai there.  Point is, there's a gradual expansion of scope.  Going back to pixel art, I couldn't make something like this - http://pixeljoint.com/pixelart/4328.htm until I can do something like this - http://pixeljoint.com/pixelart/97251.htm -- Extreme examples I know but hopefully you get the idea I'm trying to convey.

All of this to get down to this:

I'm capable of doing traditional art, and have done some in the past.  I'm not what you would call an accomplished artist, but I'm not entirely a beginner.  I'm not asking for either of these and apparently, again, I did not communicate what I was asking very well.

Actually to me it sounds like you want the former, which is good.

Draw pictures of realistic faces from photos/real life....

You will need to learn values to shade realistically...

Now that we have those down, let's work on colours...

Draw cartoon faces and shade in a cartoonly manner...

Finally, you are now the jedi-wizard master of representing real life things with believable less-real things.
Just yolo straight on in to any pixel art project. Anything you have been drawing you will be able to pixel and if you want to pixel something you don't know how to draw, learn to draw it first...

Further you will need to hone your skills.

Redo cluster studies...

This is exactly what I'm looking for.  That didn't sound like it was very difficult for you to come up with, so how could I have worded my original question differently in order for you to have given me this answer in the first place? :-)

[Ambivorous]

Actually to me it sounds like you want the former, which is good.

Draw pictures of realistic faces from photos/real life....

You will need to learn values to shade realistically...

Now that we have those down, let's work on colours...

Draw cartoon faces and shade in a cartoonly manner...

Finally, you are now the jedi-wizard master of representing real life things with believable less-real things.
Just yolo straight on in to any pixel art project. Anything you have been drawing you will be able to pixel and if you want to pixel something you don't know how to draw, learn to draw it first...

Further you will need to hone your skills.

Redo cluster studies...

This is exactly what I'm looking for.  That didn't sound like it was very difficult for you to come up with, so how could I have worded my original question differently in order for you to have given me this answer in the first place? :-)

Phew, glad we got that sorted. Honestly I would say your opening post defined it perfectly fine given what we know now.

Where I think the confusion came in is that to me it seemed I gave the same answer in that most recent post as in my first post. The only difference (to me) is that I was telling you what to do in the most recent one and saying what I would have done in the first.
I don't like the idea of telling people what to do and how to do it because the same thing doesn't work for everyone. For example, that feeling you speak of in your game design where you finish little projects and begin to work on progressively bigger ones. I would not have thought of that as valuable because it holds no weight to me. I like the idea of being able to grow my game as I fancy, so if I made Tetris I would become quite unhappy once I'd finished because I could not do any more.
So I just gave you a brief summary of what I'd have done and figured you'd ask for more if you wanted it.

All said and done - that list I gave is exactly what I'd do and therefore what I would recommend. I urge you to head over to The Daily Sketch and begin your journey into learning pixel art.
What happens there is that we are supposed to post art every day, and the idea is that if you are showing people your work daily you are more likely to complete it. It should also be noted that when someone consistently does it others tend to follow more, so don't be shy and certainly don't hold yourself back from asking questions, because the guys here will help you out with no argument if they can see you are taking it seriously.

[Uhfgood]

Thanks Ambivorous!

[Ai]

If Ambivorous's post was the kind of thing you were looking for, then I'd recommend Vilppu's Drawing Manual. It's made up of 12 stages in a clear progression. similar to what Ambivorous was saying, but goes into more detail and also tries to give you a big-picture view of how to structure art.

That has no direct relation to pixel art, of course, but pixel art isn't that complex; most of the notable techniques are just 'optimizations' rather than things to learn about art.

All said and done - that list I gave is exactly what I'd do and therefore what I would recommend. I urge you to head over to The Daily Sketch and begin your journey into learning pixel art.
What happens there is that we are supposed to post art every day, and the idea is that if you are showing people your work daily you are more likely to complete it.

Huh, I didn't know that was the idea. I've been despairing over managing to post there like.. ever, since I have a huge backlog to work through. Maybe it would be okay to post some of the stuff from July that I've refined in the past few days.