日本語版 →

Planning Before You Start — How to Decide What to Draw

The biggest time sink in pixel art is not, as most people assume, the act of placing the pixels. It is this loop: you grab an image that looks roughly right, convert it with roughly the right settings, look at the result, and only then realise it is not what you had in mind. Because pixel art is built one square at a time, a change of direction halfway through erases every minute you have already spent. What a pencil sketch would fix with an eraser becomes half an hour of rebuilding in pixel art.

That is exactly why the planning you do before touching anything pays off. Planning here does not mean some specialised technique. It simply means deciding, in order, who the piece is for, what it shows, how large it will be, and how much time you are willing to spend on it. It takes five to ten minutes, or two minutes once you are used to it. Those few minutes routinely save hours of rework later.

This page walks through the thinking to do before you build a pixel art blueprint in Heartopia Design Studio — from choosing a purpose, to picking a subject, sketching a rough layout, settling the composition, checking the viewing distance, estimating the workload, and getting yourself unstuck when the piece stalls. Treat the tool operations as something that follows naturally at the end.

1. Decide the purpose first — every other choice follows from it

The first thing to settle is not the art style or the subject, but the purpose. Without a purpose, questions like "how much detail should this have", "how many colors am I allowed", and "what is the focal point" all float free with no way to answer them. Once the purpose is fixed, roughly eighty percent of the remaining decisions narrow down on their own.

It helps to sort purposes into three broad types.

Purpose How it gets seen What to prioritise in planning
Clothing you wear yourself In motion, from a distance, glimpsed for an instant Silhouette and color impression. Not detail — what matters is that "what color is that outfit" reads instantly
A piece to show off or display Stationary, viewed up close and at leisure Density of detail and overall finish. The fine work is the attraction, so there is nowhere to cut corners
A post to Everyone's Designs As a thumbnail in a list, compared side by side with other works Staying legible when shrunk. Immediate clarity outranks everything else

These three pull in subtly different directions. Clothing, for instance, is seen while moving, so intricate patterns are barely registered in practice. Something simple and strong — "a white base with one big red ribbon" — sticks in memory far better. A display piece, by contrast, gets a viewer who stops and looks, so detailed work translates directly into appreciation. And a gallery post has to catch the eye inside a small thumbnail sitting next to dozens of others, which makes color contrast and clarity of shape the deciding factors.

When you choose "create new" in this tool, you land on a screen for selecting a canvas type and precision (pixel count) matching the design frames in the game. With a purpose in hand, this screen holds no dilemma. If you know which in-game frame you want to use, pick the type that corresponds to that frame. If you have no plans to fit it into a game frame and simply want to make a picture, choose free canvas mode. Be aware, though, that a work made in free canvas mode has an arbitrary size and may not fit the in-game frames, so "am I going to use this in the game" is a question to settle at the very beginning. Leaving it vague and discovering after completion that the piece will not fit the frame is the most wasteful outcome of all.

Tip: Try writing the purpose down as a single sentence. If you can produce something like "an outfit with a summer festival feel that reads clearly from far away", your planning is already half done. If you cannot write that sentence, you have not yet decided what you are making. Pressing convert at that stage leaves you with no yardstick for judging the result, and you will end up fiddling with settings indefinitely.

2. Choosing a subject — what suits pixel art and what does not

With a purpose settled, the next question is what to depict. The thing worth knowing here is that not every subject in the world suits pixel art. Pick an unsuitable one and no amount of careful work will produce a result you are happy with. That is not a matter of skill; it is a matter of subject selection.

Only two criteria separate suitable from unsuitable

The test is simple: is the shape simple, and are the color boundaries clear. Pixel art works with a limited grid and a limited palette, so complex shapes break down into jagged outlines, and vague color boundaries collapse into flat, muddy blobs.

Five checks for choosing a subject

Once you have a candidate in mind, run it past the following five checks. Three or more hits means you can proceed with confidence. Two or fewer means you should either find another candidate or look for a way to simplify the subject — that is faster in the end.

  1. Can you recall its shape with your eyes closed? If you can reconstruct the outline from memory, that is proof the shape is simple. Everyone can picture "an apple"; nobody can picture "somebody else's back garden".
  2. Can you describe it in three to five colors? A subject you can sum up in a few words — "red fruit, brown stem, green leaf" — survives color reduction with its impression intact.
  3. Does it still work with shading ignored? Rendering three-dimensional shadows eats a lot of colors in pixel art. Choose a subject that stays recognisable without shadows and you save enormously on your palette.
  4. How thin is its thinnest part? Fine lines like a cat's whiskers or the frame of a pair of glasses vanish once they fall below one square. Roughly estimate what fraction of the whole those details occupy.
  5. Does it need lettering? Legible text in pixel art needs about five squares of height per character as a rule of thumb. Squeezing a long word into a small canvas is best avoided.

If you are working from an existing photo or illustration, this judgement feeds directly into the conversion result. This tool offers three conversion methods — Method 1 (Standard, based on perceptual color difference), Method 2 (Illustration / Simple, which raises brightness and contrast), and Method 3 (High Contrast, which emphasises contrast and saturation) — but no method will produce a clean result from a subject that is fundamentally too complex. "Choosing a subject that can work" beats "rescuing it with conversion settings" by a wide margin. The methods themselves are covered in detail in Image Conversion Tips.

3. Sketching a rough layout — biggest areas first

With a subject chosen, do not head straight for the canvas. Sketch a rough layout first. "Sketch" does not mean drawing anything skilful. Paper, a notes app, even squares and circles drawn with a finger on your phone will do. The goal is not a beautiful under-drawing but fixing the placement and relative size of the elements.

Procedure: place the large blocks in order

  1. Draw the outer frame. Sketch a rectangle with the same aspect ratio as the canvas you will actually use. Without it, the picture in your head will drift away from the real frame.
  2. Place the single largest element. The background color, the base color, the torso of your subject — whatever occupies thirty percent or more of the picture. Once that is fixed, the remaining space becomes visible.
  3. Place the second largest element. The subject's head, a large pattern, the symbol you want to emphasise. By this point seventy to eighty percent of the picture should be filled.
  4. For details, just note "something goes here". Eyes, a mouth, small ornaments — a circle in the right place with the word "eye" next to it is enough. Drawing details at the rough stage anchors you to them and makes it hard to fix the overall balance afterwards.
  5. Step back and look. Stop, view the whole thing, and check whether the balance between the large elements feels off. If it does, erasing and redrawing at this stage costs you thirty seconds.

There is a reason for the "biggest areas first" order. The eye registers large blocks of color first, then shape, and only last the fine detail. In other words, a work's first impression is essentially decided by its largest elements alone. Start from the details and you end up deciding this most important part last, which is how you get a piece that is intricately worked yet somehow lacks impact.

4. Settling the composition — where the focal point sits, and the empty space

Composition is the step that tightens up the rough layout. There are complicated theories about it, but on a canvas as small as a pixel art grid, these three points are enough to make things hold together.

(1) Put the focal point at centre, slightly high

Placing the focal point dead centre is stable but a little monotonous. The standard fix is to shift it slightly above centre. People unconsciously perceive the upper part of a picture as "heavy", so a focal point placed a little high draws the eye naturally, while the empty space left below settles the whole composition. For clothing designs in particular, thinking about the height at which the eye tends to land pays off.

(2) Do not be afraid of empty space

The most common beginner move is to fill open space by adding small patterns. The result is a noisy picture with the focal point buried in it. Pixel art is displayed small, so the more elements you add, the less each one registers. Think of empty space not as "somewhere I have not drawn yet" but as a device for making the focal point stand out. Thirty to forty percent of the picture filled with a single flat color is about right for readability. Better still, the more empty space you have, the easier the building work becomes. It is one of the few points where the design benefit and the labour benefit line up perfectly.

(3) Use symmetry and asymmetry deliberately

A left-right symmetrical composition is orderly, stable and reads as neat and deliberate. It suits heraldic designs, geometric patterns and formal clothing. There is a practical bonus too: with symmetry you build one side and mirror it, which halves the actual work.

Asymmetry, on the other hand, creates movement and personality. A single patch on one shoulder, a signature in the bottom right corner — designs like these have a liveliness symmetry cannot match. Asymmetry is harder to pull off, though, so a good compromise is "symmetrical by default, broken in exactly one place". The overall order holds, and the one broken spot naturally becomes the accent.

5. Checking the viewing distance — does it survive being shrunk?

The final and most important planning check is the viewing-distance test. Pixel art is always seen at a smaller size than the one you build it at. In the editor a single square may be the size of your fingertip, but worn as clothing in the game or shown as a gallery thumbnail, the whole piece is about the size of your thumb. Finish a piece without checking that gap and you get the classic result: impressive up close, unidentifiable from across the room.

Checking is easy and needs no equipment.

The pass mark is this: "what is this a picture of" should be answerable within one second. If it is not, the fix is not to add detail but to remove it. Cut elements, enlarge the focal point, widen the brightness gap between subject and background. Brightness in particular has an outsized effect: two colors of different hue but similar brightness blend into one when the image shrinks. Making the background light and the subject dark (or the reverse) can improve legibility dramatically. This line of thinking connects to Color Schemes and Palette Design as well.

6. Estimating the workload — know the finish time before you start

Round off your planning by estimating the workload. With an estimate in hand you can make realistic calls like "I only have an hour today, so let me go with this precision". Start without one and you tend to run out of steam midway, set it aside, and slowly accumulate a pile of half-finished pieces.

A rough formula for time required

Build time in pixel art can be approximated with the following.

Time required (minutes) ≈ total pixels ÷ pixels you can place per minute + number of color switches × cost per switch

It looks like a formula, but what you actually do is simple. First get the canvas's total pixel count (rows × columns). Then work out how many squares you can place in a minute. This varies a lot with experience, so the reliable approach is to finish one 10×10 block first and time it. If a hundred squares took five minutes, you now know your pace is twenty squares per minute.

To that, add the color-switching cost. Reselecting a color from the palette takes far longer than placing one square. The more colors you use, and the more they are scattered across the picture, the more switches you incur. This is largely why this tool recommends fifteen to twenty-five colors as a comfortable working range. A piece with fifty colors is not twice the effort of a twenty-color piece — it feels like three times or more.

Situation Effect on the estimate
Large flat background areas Much shorter. It proceeds like filling in, so it finishes faster than the pixel count suggests
Many colors scattered across the picture Much longer. Frequent switching makes it take longer than the pixel count implies
Left-right symmetrical design Shorter. Every decision on one side transfers straight to the other
Lots of fine gradients Longer. You have to distinguish between similar colors as you place them, which adds checking time

Tip: If the estimate exceeds the time you actually have, decide right now to either drop the precision (pixel count) one step or simplify the subject further. That is far cheaper than turning back after you have started building. Also, this tool's block work mode saves progress in units of 10×10 blocks, so you can plan to split the job up — "three blocks a day", for example. Count the total blocks and you can see how many days remain until completion, which makes motivation much easier to sustain.

7. Recovering when you get stuck, and the rights question for source material

However carefully you plan, the moment when something "just feels off" will arrive mid-build. Pushing on with the work at that point is usually the wrong move, because building while carrying the discomfort simply completes the piece with the discomfort baked in. Work through the following order instead. It moves from shallow causes down to deeper ones.

  1. First, shrink it and look. Very often the problem only bothers you because you are looking at it up close. If it holds up at the real viewing size, carry on.
  2. Next, suspect the colors. When the shapes are right but the impression is poor, it is usually the palette. Check whether too many colors are blurring things and whether the brightness gap is sufficient. Color reduction runs in two steps (removing the least-used colors, then merging similar ones), with manual "custom" reduction on top, so tidying up the palette alone often transforms a piece.
  3. Next, suspect the composition. Is the focal point too small? Have you crammed in too many elements? Look for a way to increase the empty space.
  4. Last, suspect the subject. If nothing so far has helped, the subject itself may not have suited pixel art. Go back to the five checks in section 2 and re-evaluate.
  5. If it still will not work, walk away. A few hours or a night's break makes your judgement startlingly clearer. Block work mode saves progress, so interrupting costs you nothing.

Rebuilding is not a failure in itself. If anything, noticing the problem partway through means you came out ahead. The design editing mode includes drawing tools such as a pen, along with an image insert function, so you can also take a finished blueprint as a base and revise it in places. There is no need to throw everything away — keeping what works and fixing only the problem areas is the most practical form of recovery.

Rights, and why to check them while planning

One last thing best confirmed at the planning stage: the rights to your source material. Realising after the build is complete that the material cannot be published is a heavy loss both in time and in morale. While you are still choosing the image to work from, confirm that it is a photo you took, a picture you drew, or material whose use has been clearly licensed.

Note: Please do not publish works that use images, logos or characters whose rights belong to others without permission. Published works appear in the gallery only after administrator review. The care required differs between enjoying something privately and publishing or sharing it. If you are considering a post to Everyone's Designs or issuing a share URL, keeping this in mind while you select your material will save you trouble.

Summary — five minutes of planning, hours of payoff

Here is the whole flow once more. Decide the purpose in a single sentence. Test the subject against the five checks. Lay out the rough sketch from the largest areas down. Settle the composition with the focal point's position and the empty space. Shrink it and confirm it reads in one second. Estimate the time required. And confirm the rights to your material. Seven steps.

All of it takes five to ten minutes, or a couple of minutes in your head once you are used to it. And the rework it prevents runs to hours. The actual operations in this tool — choosing a canvas type and precision, converting an image, adjusting noise reduction and color reduction, then building it out in work mode — proceed without hesitation once the planning is done. Think of planning as gathering up all the moments of indecision and clearing them out before the work begins.

Above all, pixel art made after planning is far more likely to get finished. Most pieces that stall half-built did not stall for lack of skill; they stalled because their maker no longer knew where they were heading. Decide the destination before you start walking, and your odds of arriving go up sharply.