[Comic writing/Colouring] Downward Spiral.

[Chairtastic]

After much hooplah, I have gotten permission to post this stuff up in this hizzy.  Danman and I have partnered up on a comic, check Danman's thread for some of his stuff.  I serve as an auxiliary writer, and grammar checker, while learning how to colour with GIMP.

Below you will see an example of my colouring.



Basic, isn't it?  I'd like some help with improving it, and learning some shortcuts so I can send comics along to Danman for posting on schedule.  Please restrict your comments to the colouring, and not the art iself.  If you want to do that, go to Danman's thread, which you will find here.

I would also like some critiquing of my comic writing ability.  My work in Danman's comic:  Asocial Skills, starts here onward.

I don't expect the advice to be gentle, but I'd like it to at least be helpful, if at all possible.  But then, I myself am prone to rampant douchebaggery....

First Order of Business.
Downward Spiral.

EDIT**
Oop, almost forgot.  All images are copyright to their drawer, Danman.
Liatai is copyright to Liatai.
49 Techmaster Glitch (or whatever its called) is copyright to techmasterglitch.
Meanyowl is copyright to Meany.

[Tapewolf]

Personally I'd have cleaned up the lines a bit more myself.  There seem to be one or two white patches, but you seem to have avoided the first trap of colouring, which is trying to do a flood-fill on antialiased lines.

I'd say you were receiving the line-art in or converting it to straight B/W, yes?  And then either flood-filling or manually painting it in?  Is that correct?
The two most common techniques seem to be either doing what you're apparently doing, or hollowing out the line-art and painting onto a layer underneath.  The latter allows you to work with antialiased lines, but I find the former to be quicker and easier, at least so far.

Oh, here's a question.  If you're using the flood-fill technique, are you cutting the characters out of their backgrounds?

IOW, does the line art - in GIMP - look like this:


http://www.project-future.org/stuff/things/niall_flat.png

..or this..?


http://www.project-future.org/stuff/things/niall_layered.png

I think one the single most important things to get to grips with is working with layers.  I don't know whether you're familiar with them or not, but if not, that's where I'd concentrate my research.

[Turnsky]

Personally I'd have cleaned up the lines a bit more myself.  There seem to be one or two white patches, but you seem to have avoided the first trap of colouring, which is trying to do a flood-fill on antialiased lines.

I'd say you were receiving the line-art in or converting it to straight B/W, yes?  And then either flood-filling or manually painting it in?  Is that correct?
The two most common techniques seem to be either doing what you're apparently doing, or hollowing out the line-art and painting onto a layer underneath.  The latter allows you to work with antialiased lines, but I find the former to be quicker and easier, at least so far.

Oh, here's a question.  If you're using the flood-fill technique, are you cutting the characters out of their backgrounds?

I think one the single most important things to get to grips with is working with layers.  I don't know whether you're familiar with them or not, but if not, that's where I'd concentrate my research.

tapewolf, tapewolf, tapewolf. chances are i'll come off as either an elitist, a douchebag, or both, BUT, you are off the mark. seeing as we've been having discussions about lineart all over the place lately, i'll also come across as slightly miffed.

first off, aliasing, antialiasing, forget about that, aliased lines are "automagically" antialiased when you compress down in the 'save for web' feature or whatever the heck gimp has. If your lines are aliased when compressed, something's wrong with your program or processes.

Second, while flood fill is a common technique, and "paint behind the lines" another, there's a certain degree of finesse for both. However, it benefits from clean lineart that only benefits further when the person inking it gets it right the first time. To ink is to not sketch, sketchy lines make for a poor visual product, and an absolute pain to clean and color.

and lastly, transparencies on backgrounds and layers. layers are good, i like layers, making the space behind the characters transparent however, is just something convoluted and adds another step to the whole process which i find counter productive and in the end serves no overall purpose.

to illustrate further, allow me to show you two images. i shall just post the links, as they're actually quite large.
http://www.ariannia.com/images/stuff/ffc368flathigh.jpg
here we have the flat colors for page sixty three, see the aliasing on the lines? the background for the record is on my color layer along with all the characters and such. which is just the one layer that's below the main lineart layer.

and now we have the flat colors in the same size as it would appear in production.
http://www.ariannia.com/images/stuff/ffc368flat.jpg
and behold, the jaggies are gone, like cola from a patch of carpet in a shamwow commercial.

now, i know i'll prolly be chewed out by someone because of my tone, but i'm tired and really, me and critique don't do civil. I got taught by having knowledge hammered into me. God knows it should work for some out there also.

[Chairtastic]

Thank you both for popping in.   To answer one of Tape's questions; yes, I get them in black and white, haven't a clue how to convert an image to black and white, and Danman actually showed me a trick where the line art is duplicated, converted to alpha in a new file, then squished onto the colouring when its done; is that colouring behind the lines?

Another thing.  I'm new to GIMP.  I know how to create new layers, and again, my skill is limited to the above trick, and draw straight lines, however I have no idea what things like aliasing, or antialiasing are.  Perhaps you could elaborate?

And Turnsky...not at all sure what you're trying to say, as again, I have no knowledge of aliasing, transparencies, or any of that.  Terribly sorry.

[Turnsky]

oh okay, a bit of a clean slate.. fair 'nuff.

first off you might wanna see whether Danman can provide you with slightly thicker lines to work with, that might improve things a tad looks wise.

[Tapewolf]

first off, aliasing, antialiasing, forget about that, aliased lines are "automagically" antialiased when you compress down in the 'save for web' feature or whatever the heck gimp has. If your lines are aliased when compressed, something's wrong with your program or processes.

I think you've missed what I was trying to say.  What virtually everyone does at first, myself included, is scan something in in greyscale, flood-fill it and wonder why they get white fringes where the two meet.  As it happens I was going to say what you've said here, but got distracted.

making the space behind the characters transparent however, is just something convoluted and adds another step to the whole process which i find counter productive and in the end serves no overall purpose.

I beg to differ - it depends entirely on how the background is assembled.  If the background is present in the line-art, then as you say, no - don't do it, that's a waste of time.

If however, your line-art contains just the figures and the background is being constructed digitally by the colourist - which is what DMFA has often done - then I would construct the backdrop on a different layer and do the transparent background thing on the character layer.  IMHO it can be a big advantage because it gives you freedom to reposition the characters if that becomes necessary.  You also get the bonus of being able to re-use the backdrop on another panel if that is appropriate.

It depends what you're trying to do.

[Nyil]

So, while these two are going at it, I think I'll go over what I do for coloring digitally

1.) Clean up the lineart - it honestly just looks like you colored the pencils instead of inking it in and THEN coloring it, because the lineart looks a little messy.
2.) Thicken those lines up. There's several styles in which you could do that, but basically, you create a new layer, and trace around the draft under it with a thicker brush, ignoring superfluous lines.
As I've mentioned, there's several methods for doing this. The first is to just draw them in by hand. Personally, my hands aren't that stable, so I tend to sketch them in. http://www.furaffinity.net/view/3786637/ is an example of this.
Second, you could use the GIMP equivalent of the pen tool if it has one - it's basically a vector based tool that will give you very even results. Vectoring is actually one of my favorite things in the world to do, as it gives very interesting results.
http://www.furaffinity.net/view/3655398/ is an example of art with vectored lines.
The third thing, isn't really a method of drawing lines, but a way of making the lines more interesting, so to speak. That way is, basically, to 'shade' the lines, by making dark shadows appear according to where the light source is and such - http://www.furaffinity.net/view/4335907/ is how that looks.
3.) Shading. Pick a light source or two or three or four if you're feeling brave, and shade. Shading can make images really pop from a page, and can really add life to a piece of art.
4.) Backgrounds - they make the image less boring to look at. You don't need anything super complicated. If you'll notice, in all three of the images I posted, there is SOMETHING going on in the background, whether it's just a big flower or a shadow to show that your characters aren't floating around in space.

Hope that didn't seem too mean, sometimes people take my critiques far more seriously than they should Anyhow, good luck with coloring!

[VAE]

*folds ears backwards*

Sadly, no, he wasn't using pencils - my lineart isn't exactly clean - in stuffs like my current avatar i did quite a lot of fixing the lines in GIMP before i colored it in....

what i gave him was the stuff thresholded to remove the remains of pencil and the squared paper squares (i use old paper from high school for drawing these days ) and then shrunk to about twice the width of the "final product resolution" (so about 500 width per panel, given i use 250 per panel in the finished stuff), that should answer tape's question.

[Turnsky]

*folds ears backwards*

Sadly, no, he wasn't using pencils - my lineart isn't exactly clean - in stuffs like my current avatar i did quite a lot of fixing the lines in GIMP before i colored it in....

what i gave him was the stuff thresholded to remove the remains of pencil and the squared paper squares (i use old paper from high school for drawing these days ) and then shrunk to about twice the width of the "final product resolution" (so about 500 width per panel, given i use 250 per panel in the finished stuff), that should answer tape's question.

there's problem one.

get plain white paper, the variety most likely used in photocopiers.

Plain. White. Xerox Paper.

edit: this is just 'in general' (and not directed at you in particular, but think of it as 'general advice to all') but i do think it's something everyone forgets, is that before folks start to worry about image editors, pixels, tablets, and so on, there should be consideration taken with the most basic of tools to actually get things right at that stage before even contemplating getting into other things.

those basic tools are:

Paper. Pencil. Eraser.
Paper should be white, clean (no lines), and can be bought by the ream of 480-500 sheets relatively cheaply if you go the copier paper route, and really that's all you need to begin with.
Pencil, whether mechanical or wood, should be just HB, nothing fancy, and again, relatively cheap.
Eraser, just a plain, white, vinyl eraser does the trick nicely.
Clipboards are largely your choice, but i do recommend one that's quite sturdy.

[VAE]

*folds ears backwards*

Sadly, no, he wasn't using pencils - my lineart isn't exactly clean - in stuffs like my current avatar i did quite a lot of fixing the lines in GIMP before i colored it in....

what i gave him was the stuff thresholded to remove the remains of pencil and the squared paper squares (i use old paper from high school for drawing these days ) and then shrunk to about twice the width of the "final product resolution" (so about 500 width per panel, given i use 250 per panel in the finished stuff), that should answer tape's question.

there's problem one.

get plain white paper, the variety most likely used in photocopiers.

Plain. White. Xerox Paper.

hmm , i still probably have some left from department scrap papers.
i'll see if it makes matters better, though the above can get rid of the grid rather well...

[WhiteFox]

Meany: As for the colouring here... it looks pretty clean to me.

The quickest way, I have found, to add colour to lineart goes like this:

1) Take the layer with the grayscale (Or B/W, but I find grayscale has better results) line art. On the Layers panel  there should be a drop-down menu labeled "Mode" (In GIMP. The same goes for Photoshop, but the menu isn't labeled). Set it to Multiply. Nothing will visibly change at this point.
1.5) Do any needed touching up or editing on the lines. Erasing can be done with the paintbrush using white, or with an eraser. It really doesn't matter. This step is optional, and probably on danmans side of things anyway.
2) Make a new layer for colour, and put it underneath the line layer.
3) Outline regions of colour with a pencil tool, then flood fill the outlined regions with a paint bucket.

The upside is that the process doesn't require any editing of the lineart, at all. You can colour pencil sketches (This portrait of Felix was an inked drawing, shaded with pencil, and given colour in photoshop. In this pic of Darkshine, I preserved a fair bit of the preliminary pencil lines and used them to add texture to the armour). The lines don't have to be closed shapes, they don't have to be B/W, and smudges wont slow things down. Cleaning up the lineart is recommended, but not necessary.

Having to outline the colour areas before flood filling them takes a bit of time, but the fact that they're beneath the line-art means that one can do this quickly; the lines will hide any minor errors. It takes less time then methods that require you to touch up the linework extensively.

As for adding shadows... There's a lot to study when it comes to how light works, and everyone has their own techniques for shading an image.  I would reccomend doing a lot of reading up, a lot of experimentation, and damn the torpedos. Share the results, we'll comment.

One suggestion; shadows depend on the shape of the object they're falling on. Take a look at how shadows fall on three dimensional shapes, and how the shadow tells you what shape it is. (This suggestion goes out to Aisha and Tapewolf, too)

People tend to use a large, soft, Airbrush or a Gradient Tool. I don't, because it makes a very, very consistent fade from one shade to another. Shadows are not regular shapes, and they do not always fade at a steady rate. Still... most other methods are manually intensive, or very technical (My method is horribly manually intensive: First, I cel-shade an image using two colours (Sometimes three or five). Then, I use an airbrush or pencil tool set to 50% opacity, and an eye-dropper. I-Drop one shade, blend it into the adjacent shade with the brush, I-Drop the blended shade, blend it into the adjacent shade, wash, rinse repeat. Lots of work, but it gives me lots of control).

I don't expect the advice to be gentle, but I'd like it to at least be helpful, if at all possible.

The funny thing is... people can be jackasses whether they're right or not. Some people will become jackasses because they think they're right, when in reality all they have is a very strong opinion. When you get right down to it, jackasses are jackasses, and being right doesn't make being jackassery any more acceptable. Being opinionated isn't the same as being right, either. There are polite ways to make constructive comments, and critique does not equal criticism.

Also, a lot of people will get HIGHLY opinionated about petty things, like techniques and materials (I'll admit, I've done it myself, to my regret). My advice would be to try out as many different methods as possible, and see how you like them. Keep what you like, change or drop what you don't. If people make a big fuss, it's not worth stressing about.

[Turnsky]

alternatively: http://machall.comicgenesis.com/info/art.html a handy-dandy tutorial on how Ian did it back in the day.

[SquirrelWizard]

When I look at the pic, it seems like vlad's arms are either too long or his legs are too short. When a person has their arms down by their side, the tips of their fingers usually end near the middle of their thigh. as to how you can resolve it... well you can get one of those little wooden maniquins, or you can get a book on anatomy for artist, or finally you can draw from people around you (you dont even have to do fine details, just get the general body parts in the correct places and proportions)

Also, I think your perspective is off. From what it looks like, the bigger person is staring down the little guy, while the little guy meekly waves back. If this were the case, then both of their shoes would be resting on the same plane. Since this is a shot from the side, the way the shoes are set up, it looks like Vlad is further into the background than his large buddy.

Its been touched on, but there are several places on the pic where you have some wandering lines. Touch up before coloring should catch that.

[Chairtastic]

Thank you all for your advice.   I hope my implementing some of the techniques you mentioned will improve the colouration.

However, alot of you have been mentioning the lineart itself, beyond cleaning up and such.  I do believe that talk belongs in Danman's thread, since it is his art and he could benefit most from it.  Please and thank you.

And now to find GIMP's tutorial again and research all these fancy words like layers, and transparency you lot keep kicking around. 

[LionHeart]

Layers are your friend when it comes to colouring. When in doubt, add a new layer.

[VAE]

When I look at the pic, it seems like vlad's arms are either too long or his legs are too short. When a person has their arms down by their side, the tips of their fingers usually end near the middle of their thigh. as to how you can resolve it... well you can get one of those little wooden maniquins, or you can get a book on anatomy for artist, or finally you can draw from people around you (you dont even have to do fine details, just get the general body parts in the correct places and proportions)

Also, I think your perspective is off. From what it looks like, the bigger person is staring down the little guy, while the little guy meekly waves back. If this were the case, then both of their shoes would be resting on the same plane. Since this is a shot from the side, the way the shoes are set up, it looks like Vlad is further into the background than his large buddy.

Its been touched on, but there are several places on the pic where you have some wandering lines. Touch up before coloring should catch that.

1) Yeah, that sucks and is one of my problems... though my fingers DO reach quite a bit further, due to the fact my spine isn't the straightest
since i generally draw poses off myself it just might be the reason   

2) It probably is... one of the things i doesn't get right often (read nevar)... i probably needed to elongate some feature of the large guy and forgot about rel. positions....

[Chairtastic]

Well folks, here's the next spectacular failure at colouring coming out of my sweatshops.  See below.



This image is a hodgepodge of differing styles.  Before compression it was four layers, one for the ink, one for widespread colouring, and two for effects.  I cleaned the ink up as best I could, and removed the text from this image, because it was kinda in the way.  The Meanyowl was done mostly under a style perscribed by Danman and Liatai.  The chest was a mixture of a multitude of small stripes of colour that were then blurred together to result in a mottled look.  As you can see, that turned out as well as could be expected. 

Liatai (the lab-coated woman) and Danman (the man running away) were both done partially in a style suggested by WhiteFox,  where the portions to be coloured were first traced with the desired colour, and then filled.  If you notice the smidgens of colour in their hair, and Liatai's ears/mouth, that was done with an airbrush with lowered opacity...or at least I think so because my memory of way back then is fuzzy. 

The monitor was done fairly simply, colour filled some areas, then got to work on the glowing lense, which was done by fuzzy selecting the lense, Growing it twenty pixels, filling a light blue, and Blurred, then lowered opacity.  I also experimented with shading here, using the Ink tool, on the same layer as the glow for the lowered opacity, and tried to clean it up.  Failed miserably.

Comments?

[WhiteFox]

The only spot where the flat colours look off are around meanyowls neck and belly; there's a bit of white halo there. Everything else looks pretty clean.

I can't make out any shadows on danman or Liatai, tho...  Might be the contrast on my monitor or something.

(quick sketch-o-gram) Pardon my mess:


1) The spots/patterns on the owl don't look like they conform to the shape of the surface. Perspective would squish the dots into ovals, and the direction of the streaks on the neck/belly would change direction depending on what angle the surface was relative to the view.

2) One thing to watch out for is marks that look like they were done with a brush. They were done with a brush, but we don't want them to look like they were done with a brush. Sometimes, all it takes is a second stroke to tweak the shape. (The little diagram shows what the second stroke would look like; the marks on the back were done this way). This would also work for shadows, like on the monitor's ring.

Monitors Glowy Ball: The colour of a bright light is usually very pale. Try dropping the saturation to about 15%. If it's bright, it'll project light onto surfaces nearby, or make highlights on reflective surfaces.

4) A quick rundown on how I do highlights. Paint a hotspot, switch the brush to 50% opacity, and start blending. on every pass, I use the eyedropper to pick up colours and blend them further. The third step is after 3-4 passes. It should be noted that the hotspot should stay bright.

The highlight on the monitors ring was done in two strokes of the airbrush: one for the hotspot (at 100% opacity), and a second one for the arc (at 50% opacity). The highlights above/below the light were done in 2-3 passes, mostly scribbling.

5) Shadows falling on a torus are funny things. On one half, the shadows will be on the inside edge of the ring, and on the other half, they'll be on the outside edge of the ring. The diagram shows why: if one took a cross-section of the ring, the light would fall on both halves from the same direction. (It should be noted that in reality, one side of the ring would cast a shadow on the opposite side. I forgot to add this, Je'm excuse)

There's a dark portion on the top edge of the monitor that goes over the line. Using a selection mask can help with that sort of thing: Use the Fuzzy Selection tool to select the colour you're painting on, and the brush tool wont paint on anything that isn't selected. If what you're trying to mask has a range of colour, like the meanyowls neck and belly, Fuz-select the white and Invert the selection, or lasso the area and use the Fuz-select to de-select the white.

6) and 6b) Adding bold shadows usually requires one to picture what's being lit in simplified shapes (Note to danman: this also helps when drawing something in perspective, or at an angle). Most of the these are cylinders, spheres and blocks. (Liatai's shoulder was treated as a quarter of a sphere). Danmans shoe is a squared off cylinder plus a wedge.

Dan and Liatai are not being lit by the same light source: The little circle-and-triangle diagrams indicate the angle of the lighting for each of them.

6c)  This is one example of how shadows can show the 3D-shape of an object (Even though the lighting is coming from the wrong direction in this case  ). Danman's foot looked kind of flat, like it was viewed from the side rather then at an angle, so I added shadows in such a way that it looked turned  towards the camera.