Layers and opacities in pixel art.
Printed From: Pixel Joint
Category: The Lounge
Forum Name: Diversions
Forum Discription: Get to know your fellow pixel freaks. Chat about anything to do with video games, comic books, anime, movies, television, books, music, sports or any other off topic bs you can think of.
URL: https://pixeljoint.com/forum/forum_posts.asp?TID=765
Printed Date: 09 September 2025 at 4:09pm
Topic: Layers and opacities in pixel art.
Posted By: DrunkyDuck
Subject: Layers and opacities in pixel art.
Date Posted: 25 August 2005 at 1:38am
I was just wandering what you all think about using layers and opacities in pixel art for making glass, water or transparent objects.
A big part of pixelling is in making your own customized palette, so should this workaround be considered cheating?
I mainly ask your opinion on this because it's recommended by some pixel tutorials.
Personally, I prefer to see it done manually.
------------- Reaching out to embrace the random.
Reaching out to embrace whatever may come.
I embrace my desire
to swing on the spiral of our divinity
and still be a human.
|
Replies:
Posted By: Saboteur
Date Posted: 25 August 2005 at 2:07am
Manually is better, but a lot of people don't have the time to do it so. I don't think it would be considered cheating... Then again, what do I know?
------------- "I was minding my own business and walking across a pebbled path, and a Duck started giving me the business."
|
Posted By: iSTVAN
Date Posted: 25 August 2005 at 3:00am
Indeeed! It certainly is cheating as far as I'm concerned. It is an interesting topic of conversation though. Where do you draw the line with pixel art? (Excuse the pun).
I consider certain things within the realm of pixel art. I often fiddle with the contrast or brightness of ALL colours upon completion of a hand pixelled work, or in order to select colours during or before you've started. However, any reduction of colours, or special effects that actually alter the placement or arrangement of hand pixels. I consider transparancy or layering devices to fit under this second category.
I can't really see how someone could consider my techniques to breaking the laws of hand pixelling as it deals with the image as a whole and is only used upon completion of the actual pixel piece. Its no more cheating than the fill tool!
However opacities seem to be a technique which skips the hard ships of blending colours. I can understand how layering techniques could be used to select colours, but not to actually create form or subject matter such as water or glass. THAT would be skipping. It is often very obvious aswell, not very effective unless they go to great lengths to hide it, whereby you're probably better off just knuckling down and doing it by hand!
I wonder is there is any official stand on this matter...
|
Posted By: inkspot
Date Posted: 25 August 2005 at 3:11am
I do all manually in MS Paint. Used to and i truly think that this is the right thing. Its true pixelart. people didn't have all those layers and opacities in the olden days before and just look at those classics.
With photoshop, there came in those idiots who cannot even draw a proper house. people are used with automation, and automation is not your creation, especially when the program automation is not your creation.
Well, from that point, you can now argue that its not even your creation if you use usual Paint like program, since that automates too. And then on, it turns really philosophical. Just to create simple pixelart, as your creation, you have to create a computer and programs in it...
Ok, now, forget what I just said. That was some bad philosphy. Pixelart was born with simple automations and in programs (or not?), but it didn't have any opacy layers and stuff.
Ok, now to the point. What I want to say is that actually doesn't matter if you use some automated opacy or not, but you should know the basis of drawing and basis of opacy first. You just don't learn if automation does everything for you. I don't even use layers as i am trying to be more precise on my work and try not to make mistakes. I don't say i am pro, I am very much beginner and that is my vision of proper learning.
I am just sick and tired of those punks who come and claim they are beginner dadada and they don't even know how to draw a house from children class level.
Usually those are guys who come and ask for tutorials because they are too laizy to find some out for themselves and usually they don't even look at the program if somebody is not telling them to do so...
I am laizy too, but usually I try to figure things out by myself, it takes time but at least I know why and what I am doing. in fact, i am too laizy to read all those tutorials (especially all kinds of tutorials about program use) and i prefer to find things out piece by piece, manually.
I have used automation and I have got screwed by it. I hate that opinion that photoshop is best for pixelart, I mean, just look at it, too many functions and windows and automations and layers and stuff, it messes up. Keep things simple and you win!
|
Posted By: Niss
Date Posted: 25 August 2005 at 5:25am
That is indeed an
interesting topic! I remember beeing really strict about what is proper
pixel art technique and what is not when I started pixelling. That was
because I didn't knew better at that time. Truth is the term
Pixel Art is confusing. It describes a special technique (like air
bursh or oil painting) AND it hints that it is some art style (like
cubism and surrealism).
Technical
stuff: To define pixel art technically is pretty easy (1px brush, no
auto-AA, no other auto-shape-tools, manual color selection (guess you
didn't know that eh?) and maybe some other stuff). So... but it is not
like some people decided to develop a new technique (although it might
seem so to "noobs") and rattled the proper limitations down (and yes a
technique is most easily defined by it's limitaions). The technique was
a result of the hardware details of earlier computers (and damn those
times where more fun computer-wise !
). A low resolution meant that you had to create your graphics with
only a few pixels and at the same time you only had a few colors. So
it's obvious that things like auto-AA and tools for complex or even
simple shapes didn't make sense - not that there was a program like
photoshop at that time (I'm talking about pre-DeluxePaint times and
also pre-"high" color 256 palette times).
From this technique resulted a unique look much later known as pixel
art (yes this term is pretty new). You should realize that this was
simply the only possible form of digital graphics then. Later with 256
colors and higher resolutions there appeared first "non-pixelart"
graphics on computers including scanned and res/color-reduced pics
(which still looked good. Have a look at beneath a steel sky. The bg's
are scanned-in drawings from dave gibbons!). Also progs like deluxe
paint had functions like auto-AA and different shapes.
Style:
While it is difficult in digital media to speak of art styles or
periods because it is all new and fresh I strongly believe that there
is something like Pixel Art. Like I described above it there resulted a
special look from the limitaions of the technique and developed for
itself. Also since a few years it experiences a rennaissance - again
because of technical limitations. This time from cell phones, Handelds
and digital signs like on busses/subways.
In my opinion you should know the history of the art form you create
pieces in otherwise you can't really judge your own pictures and their
effects on others (it's exists in a context you know even if you don't
want it to). That means have a close look at old games, demo scene
pics, icons and all old digital graphics...
Oh sorry drunkenDuck for hijacking your thread Just flowed out of my mind...
|
Posted By: nvision
Date Posted: 25 August 2005 at 6:19am
Meh...not another one of these "pixel purist" threads!
As far as I'm concerned, I think as long as there is some level of individual pixel manipulation, it doesn't matter whatever other tricks you throw in. If it makes your work look better, and is more efficent, then bravo for you. If you're producing commercial work, only the end product matters, and not the means of production.
The only reason that things like these were not used in traditional games is that most of the technology didn't exist. If Leonardo da Vinci were around today, I guarantee you he'd be a ground-breaking 3DS Max artist, or utilizing some other technology to it's fullest potential...
|
Posted By: Niss
Date Posted: 25 August 2005 at 6:31am
nvision: I agree it doesn't make sense to limit oneself just because of
"purist arguments". I just wanted to show that those purist ideas
developed from hardware limitations and not some abstract phenomenon.
What tools you use should really depend on what you want to create and
not because you are becoming hot when you're seeing pixels/pencils or
brushes or whatever
|
Posted By: sedgemonkey
Date Posted: 25 August 2005 at 8:37am
For me pixel art is about exploring iconic representation, dithering, hatching, shading, limited color palettes and making every pixel count. Any tool to help with color selection seems completely valid in pixel art to me. It just saves time with having to use the damn fill bucket over and over again.
|
Posted By: Wannahlakujuu
Date Posted: 25 August 2005 at 8:59am
I can't really do any of that, just because all I really have is mspaint, and if you can do that stuff in mspaint (and if you can then I didn't know) The only other program that I have is Paint shop pro, and I don't know how to do anything other than reduce color depth and stuff, but I only use it for coloring drawings, and doing lineart from drawings. Other than that, I use mspaint, even if I did know how to do all that stuff, which I could if I wanted to, I wouldn't use it, because I like to do everything by hand.
|
Posted By: pixelblink
Date Posted: 25 August 2005 at 3:39pm
I personally use opacity in some of my pixel pieces to add "dirtyness" and shadows, etc. That's not to say I don't enjoy a good purist piece every now and then. As an artist, I don't like to have limitations. As a future job opportunity though, it is excellent practice to limit your colour selections as many paid jobs require it (ie: cell phones, video games, etc).
To repeat myself: it is good to be able to restrict your palletes. You gain more knowledge and skill this way as people like Ryan-gfx, zi-, and others have. Full respect to them for doing what they do and how they do it.

-------------
|
Posted By: Omegavolt
Date Posted: 25 August 2005 at 10:33pm
Anything can be made pixel by pixel. Its all about knowing your colors and contrast and AA. If youre making this image for a game or something, where time is of the essence, then hey, use all the tools in your arsenal. But if its art, and considered 'pixel art', then using something other than pixels is kinda cheating. ;)
|
Posted By: sedgemonkey
Date Posted: 26 August 2005 at 11:30am
Meh, I guess I like the "style" more than the rigid definition of technique. Call me a traitor. 
|
Posted By: DrunkyDuck
Date Posted: 26 August 2005 at 11:44am
Hey, I'm no purist. If you're working on a game (especially a one you dislike), ofcourse you would use anything to get the job done faster. But I'm guessing most people here make their sprite art as a hobby, growing as pixellers. And as Pixelblink said:
You gain more knowledge and skill this way
------------- Reaching out to embrace the random.
Reaching out to embrace whatever may come.
I embrace my desire
to swing on the spiral of our divinity
and still be a human.
|
Posted By: Cryssy
Date Posted: 26 August 2005 at 3:29pm
To me it depends. I can be a pixel purist and other times I am not.
Some of my larger pieces I do in a program that gives me layers. It
helps me esspecially if I decide on new details I need to squeeze in
there.
As far as opacity it really honestly depends how long I have been
working on a piece and if I can't get it right without it. Sure I
admire others that can do it, but I do not think less of myself as an
artist for using opacity...I still have to pick the color and place the
shape, and that to me is about 8/10s of good art is knowing where
things go to make everything look right. Just because you use opacity
doesn't put it in the perfect spot. 
|
Posted By: PixelSnader
Date Posted: 27 August 2005 at 4:41am
i think layers arent and opacity is cheating
more from a stylewise view, because when you use opacity, you are letting the machine decide which colour should be put in a certain place, thus taking away some of the colouring style the artist has
and layers are to me just tools to make long animations faster
eg, when i in paint need to make a cloud move left in 5 frames and have another animation in 10 frames, in paint i got to make eacht frame manually, though in things like Gale i only have to make a certain amount, since layering will fix the rest but it doesnt take away from what i intend to do with the pixels
its kinda like the paint bucket, sure i could do every pixel manually, but if i want the whole screen to be the exact same colour, whats the point? there is a difference though with things like the spray can: a tool that lazy b#stards use as a substitute for dithering
edit: allso stylewise, its great that those demosceners can make such realistical images, but to me they look more like filtered CG's of photos. NOTHING, ABSOLUTELY NOTHING BAD about sceners, but to me pixel art is about big pixels (like 3x or at least 2x on a normal monitor) and making it look like youre playin things like the SNES again even though its nice to sometimes let go of those limits and just use over 100 colours on a piece
-------------
▄▄█ ▄▄█
▄█▄ ▄█▄
|
Posted By: d-p
Date Posted: 27 August 2005 at 3:51pm
Im seeing things like nvision.
There should be enough space for individuality, Im using ps a lot but I
worked with ms paint for a very long time as well. And yes surprise I
can draw an iso house by myself.
ps is not a tool for "idiots", its a powerful tool for pixelart but not
only for it. As some already said, you need more than a good tool to
create a convincing piece of pixelart. I accept purist who just work
with paint or similar progs, but some of those purist dont accept any
other way of working than their own. I really get bored by these people
telling ps users that they'll get to hell. Even using opacity means
that you place a pixel two times at least - when working on a good
result -, layers just helpers for tyding if someone needs it. Does such
work harm purist pixelwork, pixeled with the 1px brush and with strict
color restrictions? I think not at all. There should be more
acceptance, I think. Or there should be more and harder rules for
submitting pixels here like: no pieces larger than .., no programs
like..., just the 1px brush, no line, no fill bucket, no working after
nine, not more colors than... just my 2 cents.
|
Posted By: blue
Date Posted: 31 August 2005 at 9:42am
I don't quite understand why it's alright to use opacity to select
colors - and then draw over what the opacity did exactly, essentially -
but not to just use opacity in your image. If you have one flat color
over another flat color, slightly opaque, it looks...exactly the same
as it would if there were no opacity. (; Basically, it's using a tool
to come up with the same result.
Progress, progress, progress! We should take advantage of the tools at
hand. It's like working with all the colors you can. Unless
you're specifically making a restricted color piece...use as many as
you need to make it look good.
On that note, though, I don't really use opacity. Mostly mix my own colors, all that good stuff. (;
and a question for the 'pixel purists' XD
Is it also wrong to use the fill tool? The shape tool? This isn't manipulating pixels, either...
|
Posted By: 1ucas
Date Posted: 03 September 2005 at 10:48pm
To me, here's the deal:
If you use layers and transparency, the color count will most likely skyrocket. Then, two things might happen:
1) In the end, you'll get less than 256 colors. This means you can save it in a GIF pallete without problems. PNG is perfectly fine as well.
2) You'll exceed 256 colors. This means you can't really save it as GIF because the format won't allow the pallete to be saved intact. This IS a problem, because the program will HAVE to remove colors for you when it saves as GIF. And there are two ways for this to happen:
2.a) The program will simply take a few colors out and make color bands for you. This WILL ruin your pixel work, and will probably create the illusion that you did a masterful use of the pallete adding weird colors to several peculiar places. It ruins your true work, IMO, and therefore it can't be considered entirely your work. Nothing differs this from going "Reduce color depth -> N colors (no dithering)". It is the program applying a FILTER, and just because it wasn't you who explicitly applied it doesn't mean it's any better.
2.b) The program will take a few colors off and dither for you. The method it uses for this (pattern, noise or diffusion) is completely irrelevant as well. It's just a filter to reduce the colors, and since it's not you who's dithering it's not your work. Therefore, it shouldn't be considered true, or at least pure, pixel work.
So, my point is: if you use layers, make sure you don't exceed 256 colors, and if you do, save it as PNG. This is pixel art, so every pixel matters. I think artists should always remember that, and letting the machine change pixels for you is actually letting it mess with your carefully done art. To me, that's not exactly cheating, but a lack of attention and care to the artwork, and a lack of respect to it and the other artists who will see it (in a certain level).
But, if you don't wanna have the trouble to check the number of colors before saving, just always save as PNG. You can't animate, but you can use transparency. (and don't come with the "aww, but not everyone can see it" bullsh*t. The format allows transparency, that's all you should care about. If the programmers of some software were too stupid to implement this on their program, that's their fault, not yours. That's how we force things to be implemented.)
AHEM! The only exception I'd give here is in case of animations, if they're really a fundamental part of the piece. The layer wouldn't be really necessary, still, but well, each artist will do what it can to get their things done.... But I'd rather if they used skill, not layers.
I'm not being a purist, though. I just like to zoom in and see how the person worked, because then I can study the technique and learn something from it, and something about the artist. But if I zoom in and see a computer-made diffusion dithering, it won't really do me any good, because well, how could I tell which pixels were intended by the artist and which were automatically and algorithmically placed by the program?
------------- http://toxicdump.org/labs/frameviewer/viewer.php - Online Animated GIF Frame Viewer
|
Posted By: neota
Date Posted: 05 September 2005 at 9:11am
1ucas -- Awesome, your post is very topical.
Originally posted by 1ucas
To me, here's the deal:
If you use layers and transparency, the color count will most likely skyrocket. Then, two things might happen:
1) In the end, you'll get less than 256 colors. This means you can save it in a GIF pallete without problems. PNG is perfectly fine as well.
2) You'll exceed 256 colors. This means you can't really save it as
GIF because the format won't allow the pallete to be saved intact. This
IS a problem, because the program will HAVE to remove colors for you
when it saves as GIF. And there are two ways for this to happen:
Okay, what you're saying is, don't let the program's automations mess
with your finished picture. I agree. My way of avoiding that is to do
it in advance myself :)
Some examples: * You can get completely awesome results from a
controlled quantization (check out the Sample-Colorize filter in Gimp
if you can, you have complete control over what colors are used, where,
how many..)
* 'Repeated dirty->resaturate' is a technique I am documenting in my
WIP Gimp-pixeling tutorial. It means the picture temporarily leaves the
defined parameters for it as you use a speed-technique such as
airbrushing, vectors or drawing modes. Then you put it back in range
(airbrush is good primarily for dithering, strangely.. I smooth out the
area then dither it to the two end colors and touch up. for colors..
Sample-Colorize is the weapon of choice for resaturating colors, with
absurdly little effort required to get a perfect result.).
So my solution to auto dithering/reduction/etc is to schedule and control it.
AHEM! The only exception I'd give here is in case of
animations, if they're really a fundamental part of the piece. The
layer wouldn't be really necessary, still, but well, each artist will
do what it can to get their things done....  But I'd rather if they used skill, not layers.
There are ways to use layers to do coloring directly (hint: layer masks).
If you're smart, you can restrict the resultant amount of colors quite absolutely.
I'm not being a purist, though. I just like to zoom in and
see how the person worked, because then I can study the technique and
learn something from it, and something about the artist. But if I zoom
in and see a computer-made diffusion dithering,
NOTE: a diffusion dither is crap. However, a positional dither is often
fine (the commonest manually-done dithering is based on a 4x4 Bayer
matrix, Several paintprograms use a larger Bayer matrix.). Especially
if you can specify a custom dithermatrix, it can be used to good effect.
|
Posted By: 1ucas
Date Posted: 05 September 2005 at 12:51pm
It's not really a Bayer matrix. Bayer matrixes are for extracting RGB values with each item being an exclusive channel and sh*t.
There's a name for those dichromatic matrixes, but I'm not sure what it is...
And diffusion isn't crap, but it's really time consuming to be done by hand, and the results aren't that great.
------------- http://toxicdump.org/labs/frameviewer/viewer.php - Online Animated GIF Frame Viewer
|
Posted By: neota
Date Posted: 06 September 2005 at 9:02am
..
AFAIK, A Bayer dithering matrix looks like this:
082B
C4F7
3A19
E6D5
(where each letter represents the number at that location in the
matrix, and the number is at what threshold the pixel of color A will
become a pixel of color B.)
you can scale up the matrix simply; there is a 32x32 Bayer matrix implemented in the Gimp as of 3 days ago.
|
Posted By: Commodore
Date Posted: 09 September 2005 at 7:45am
Why make it harder for yourself if you can do it easier? Photoshop
doesn't automate things that much. In traditional painting you have
opacity and it is used alot. Why can't I use it on computer?
|
Posted By: Womped
Date Posted: 11 October 2005 at 3:13am
I, personally, always use around 8 layers per image I make. It's just
more efficient and i still place all the pixels normally; that is if
you don't consider the equivelant of the line tool as cheating .
And I do use opacity, but not very often.
In personal views I don't see opacity usage or layers as cheating, but opacity is ok to frown upon if overused.
------------- Maybe if you were less intelligent people would like you more?
|
|