The gold standard of digital painting is a purposeful image made with controlled pixels, understandable layers and an export that works in the real world. A beginner should choose the intended display or print format, make a strong value sketch, paint on an appropriately sized canvas and preserve the editable file. Software features matter only when they improve the art or its reliability.
This digital painting reference explains raster art, brushes, colour management, layer organisation, drawing tablets, stabilisation, safe reference use, licences and dependable export formats. A worked 2400-by-1800-pixel example turns an abstract canvas setting into a practical calculation. The guide uses official Krita documentation for concrete examples without assuming every painting application has identical features.
This chapter extends The Gold Standard through Visual Arts, Painting and Drawing. It also connects with Collage Making, Photography and Teaching Visual Arts.
Six decisions before opening the digital canvas
- Define whether the image is for screen, print or a particular teaching purpose.
- Choose a usable pixel size and clear composition before detail.
- Work with a small brush set and labelled editable layers.
- Check colour, permissions and any third-party assets.
- Save a native working document and a separate appropriate export.
- Reopen the export, test at real viewing size and retain a recoverable backup.
Digital painting chapters
1. Digital painting is an art practice, not a button
Digital painting uses software and input devices to build original marks, shapes, colour and visual relationships on an electronic canvas. It can emulate traditional media, explore impossible combinations or develop an illustration with a distinctly digital look. The standard is intentional communication, not possession of the newest tablet.
A beginner who uses a default brush to paint a clear three-value still life may have learned more than someone who downloads hundreds of presets without understanding light. The software supports decisions. It cannot choose the visual purpose of a picture by itself.
2. Start with the intended output
Ask whether the image will be viewed on a phone, printed in a book, displayed on a website or used as a classroom explanation. The target determines working dimensions, composition, colour assumptions and the final file. Designing for all destinations without priorities leads to avoidable mistakes.
For example, a small square social graphic and a large landscape poster do not share a natural crop. Choose a master composition and test the likely outputs early rather than discovering at export that important writing sits outside the intended frame.
3. Know the difference between pixels and physical size
A raster digital painting is made of a grid of samples called pixels. An image can be 2400 pixels wide regardless of its claimed paper size. Pixel dimensions determine the amount of raster detail available; printing resolution also depends on the intended physical dimensions.
Changing only a file’s pixels-per-inch metadata does not invent new image detail. If a 240-pixel-wide sketch is enlarged tenfold without suitable resampling or redraw, it may look blocky. Choose a reasonable size at the beginning according to the intended use and the device’s capacity.
4. Raster and vector graphics serve different problems
Most painterly brushwork is developed on raster layers, where colour is stored in pixels. Vector graphics describe shapes mathematically and can be resized differently. Some programmes combine both approaches; that does not make their data models identical.
A detailed textured portrait may suit raster brushes, while a crisp scalable emblem may benefit from vector construction. Digital painting should not be used as a catch-all label for every image produced on a computer. Explain the media accurately in the finished caption.
5. Use a workable canvas instead of the largest number available
Large canvases can consume memory, storage and processing time. Smaller canvases limit fine detail or print scale. Balance the intended output with the machine’s available resources rather than selecting an extravagant resolution to feel professional.
A practical exercise might begin at 2400 by 1800 pixels, but that is only an example, not a universal standard. Test brush responsiveness and available storage before committing to a long work. A stable setup can support concentration better than a huge sluggish document.
6. Make the canvas ratio match the composition
The aspect ratio describes width relative to height. A 2400-by-1800 canvas is 4:3, while many displays use other ratios. Cropping or stretching a painting to fit an unrelated format can change the visual meaning.
Sketch a bounding box for the important subject and preserve a useful safe margin around it. If the work must be adapted to several destinations, prepare deliberate alternative crops and inspect each one instead of allowing automated centre cropping to remove meaningful details.
7. Use familiar brushes first
Begin with a simple round brush, an eraser and perhaps a texture brush. Explore what changes when size, opacity, pressure and spacing vary. A brush preset is a system of settings, not a magical shortcut to drawing skill.
Try painting one cylinder using only a basic brush before applying a simulated watercolour or oil texture. If the cylinder’s light and dark relationships are unclear, a more complicated brush merely makes the problem look decorated.
8. A stylus can help, but it is not mandatory for every task
Many artists prefer pressure-sensitive styluses because they can vary marks more naturally than a mouse. Krita’s official manual explains how tablet drivers and pressure support interact. The actual result depends on the device, operating system, application and settings.
A beginner without a pen can still practise composition with a mouse, trackpad or touch screen when the technique allows it. Choose inputs according to comfort, access and task, and avoid claiming that expensive equipment is required for original art.
9. Separate input failure from drawing failure
When pressure does not change the brush, the issue may be device configuration, driver support, brush settings or an application’s input mode rather than the user’s technique. Verify simple test strokes before changing the entire drawing workflow.
The diagnostic sequence is straightforward: test the stylus with a basic brush, confirm pressure is enabled, check app-specific instructions and, if available, use a device test utility. Only then decide whether the gesture itself needs more practice.
10. Layers preserve options when used deliberately
Digital layers allow elements to be separated, hidden, reordered or revised. Krita’s documentation compares them with transparent stacked sheets. A practical beginner structure is sketch, base colour, shadows, highlights and notes when needed.
More layers are not automatically safer or better. Hundreds of unnamed layers can slow review and confuse future editing. Group related elements, use descriptive names and flatten only a separate export when that is appropriate. The working file should preserve the original decisions.
11. A non-destructive workflow has limits
Selections, masks, adjustment layers and copies can protect parts of the image while you experiment. Their availability and behaviour vary across applications. Even a non-destructive edit can become confusing without names and a saved checkpoint.
Before a risky change, create an incremental version and explain what is being tested. Keep the earlier state until the new decision is verified. Unlimited undo is less valuable than a clear history of why an important change was made.
12. The first painting should start with values
Make a small grayscale thumbnail of the subject. Identify light, middle and dark regions and position the focal area. If a composition is unclear in grayscale, more colours often will not fix it.
Imagine a digital image of a market stall. A hundred food items may create colour, but the important sign, vendor and counter should still be legible from a thumbnail. Clear large forms give viewers an entry point.
13. Silhouette checks are quick and revealing
Fill the main subject with one temporary solid colour. Ask whether its shape is recognisable without texture. This can reveal a confusing pose, crowded object grouping or weak separation from the background.
Restore the full painting and adjust the largest relationships first. A few well-placed edits can matter more than dozens of highlights. Software makes silhouette tests easy, but the value comes from the artist’s interpretation.
14. Colour picking is a tool for comparison
Digital colour values can be sampled, stored and reused, helping the artist keep a coherent palette. But the sampled pixel may reflect lighting, compression or a source image with its own colour profile. It is not automatically the object’s ‘true’ colour.
Pick a limited palette and label intended roles such as background, skin or fabric, highlight and accent. Choose relative contrast deliberately. A very saturated colour may compete with the focal point even when every individual RGB value is technically valid.
15. Understand colour profiles before exporting
Colour management helps applications interpret image values consistently across devices and outputs. Krita’s documentation explains working spaces, display profiles and ICC conversion. A numerical RGB triplet does not guarantee identical appearance on every screen.
For ordinary web delivery, sRGB is a pragmatic default workflow in many circumstances. Wider gamuts, HDR or specialised print processes can require additional care. Follow the application’s and intended printer’s instructions, and test before promising a precise match.
16. A monitor is not a universal colour reference
The screen’s brightness, gamut, calibration and viewing environment affect perceived colour. A painting that looks delicate on one display may appear dark on another. Avoid adjusting it entirely on the basis of one unusually bright or tinted screen.
Check an export on a second reasonable device and inspect contrast at a typical viewing size. For colour-critical professional work, use appropriate calibration and proofing. For a beginner illustration, understanding the limitation is the important first step.
17. Opacity is not the same as layer order
A fully opaque upper layer can conceal what lies beneath; a partially transparent one combines visually with lower layers. A layer moved behind another may appear to disappear, although its pixels still exist in the document.
Use a simple red circle above a blue rectangle to explore three situations: opaque overlap, fifty-percent opacity and reversed layer order. Compare the results before applying the lesson to skin, foliage, clouds or other complex subjects.
18. Brush blending should support form
Smudge and blending brushes can soften transitions, but overuse may make every region equally muddy. Decide which edge describes a sharp material boundary and which shadow genuinely softens. A digital painting still depends on observation.
Make a before-and-after screenshot of one object’s edge. If the softened version loses important form, undo or reduce the effect. The paint engine is not the judge; it merely offers another option.
19. Selections and masks reduce accidental changes
A selection isolates part of an image, while a mask may restrict the visible effect of a layer or adjustment. Use them when they clarify work, especially around complicated silhouettes. But do not create a maze of dependencies that is difficult to edit later.
If a student cannot explain why a region changed, revisit the layer and mask arrangement. Good technical organisation should lower cognitive load and free attention for artistic decisions.
20. Smoothing can be an accessibility option
Krita documents smoothing and stabilisation settings that can reduce unintentional variation in a stroke. This can help with long curves, difficult line control or some physical needs. It also changes the responsiveness and character of a brush.
Try a line with no smoothing and another with a stabiliser, then compare the result at actual viewing size. Choose an appropriate setting rather than treating all hand movement as an error that software must erase.
21. Assistants are guides, not substitutes for perspective
Many drawing applications offer rulers, ellipses, symmetry tools or perspective assistants. Krita’s documentation explains such aids. They can support reliable geometry when an illustration needs it, but the artist must still choose the viewpoint and make the scene coherent.
Paint a simple street view with a horizon and two receding edges, then check whether people and objects plausibly share the same space. A perfectly straight line cannot rescue a physically inconsistent design.
22. Texture should not hide a weak structure
A photo-texture overlay, noise filter or simulated paper grain can make a digital image feel tactile. Yet texture can also distract from the subject and conceal errors in lighting. Compare the image with and without the effect at the intended final size.
If the composition is clearer without the texture, remove it or restrict it to areas where the material quality matters. The gold standard is a reason for every treatment, not a library of fashionable visual filters.
23. Use reference images with permission
Artists may study references, but downloading somebody else’s illustration or photograph does not automatically allow tracing, redistribution or commercial use. The relevant permissions depend on law, context and licence conditions.
Keep a source list for third-party images, brushes and textures. Creative Commons licences vary, and many require attribution; some restrict derivative use or commercial distribution. For a public portfolio, being able to identify a permitted input is part of professional competence.
24. Original work needs an honest workflow statement
If a painting combines hand-drawn marks, stock photos, generated elements or another artist’s assets, describe the combination accurately. It is possible to create a meaningful composition with several media, but readers should not be misled about what was made and how.
A school or professional assignment may have specific rules on external assets, AI assistance and permitted tools. Check those rules before submitting. The ability to explain authorship and process is a more durable skill than pretending an image arose from one heroic brushstroke.
25. Save the editable working document
An editable native file can preserve the layer stack, masks, guides and other programme-specific capabilities. The Krita manual describes KRA as its native working format. A flattened PNG or JPEG cannot be assumed to keep all those editing options.
Create sensible incremental files at major milestones. For example, save an initial sketch version, a value-blocking version and a final colour version. File names should be understandable to someone who did not watch every hour of the process.
26. Exporting is a separate design decision
PNG is lossless and supports transparency; JPEG uses lossy compression and is often convenient for photographic or painterly gradients. Krita’s manual explains their different strengths. Repeatedly saving over the same lossy JPEG can degrade image quality.
Keep the original editable file. Export a copy with the dimensions, profile and format suited to the destination. Reopen the exported file and check legibility, cropping, transparency and colours before declaring it ready for publication.
27. The file size follows more than width and height
An uncompressed image’s pixel count helps estimate basic memory, but actual files also depend on layers, bit depth, compression, masks and application data. A single 2400-by-1800 RGBA image at eight bits per channel carries approximately 17.28 million bytes of raw pixel data.
Three such layers have a simple raw-data subtotal of 51.84 million bytes, before overhead and ignoring compression, caching and other factors. This is an educational estimate, not a promise of exact disk or RAM usage. It shows why very large projects can strain equipment.
28. Backup is part of artistic reliability
Devices fail, documents become corrupt and files can be accidentally overwritten. An export on social media is not a complete backup of an editable project. Keep at least one separate recoverable copy and test that it opens.
A practical beginner habit is to save to a known project folder, make version checkpoints and retain a second copy on another appropriately secured storage location. Avoid sharing raw files containing private source photographs without considering access permissions.
29. Control privacy in screenshots and screen recordings
A digital painting demonstration may show notifications, personal file paths, names, reference photos or account information beside the canvas. Before recording a tutorial, close unrelated windows and remove sensitive material.
Children’s work and images require appropriate consent before online publication. A recording that helps one learner should not expose another participant’s private information.
30. Use a realistic work rhythm
Long uninterrupted screen sessions can cause discomfort and reduce concentration. Arrange a comfortable workstation, adjust display settings and take appropriate breaks. Avoid claims that one exact chair angle or universal break schedule solves every physical need.
A good session includes time for looking at the actual subject, making a thumbnail, painting, saving and reflecting. Time away from the brush can support better decisions because the artist returns able to notice what the picture needs.
31. Accessibility can be designed into the toolset
Some artists benefit from larger interfaces, remapped shortcuts, stabilisation, alternative input devices, larger brushes or saved colour palettes. The most useful adjustment depends on the person and the software available.
Evaluate the learner’s ability to communicate their intention instead of imposing one mandatory stylus grip, speed or application. Technical support should remove unnecessary barriers while preserving the opportunity for original visual decisions.
32. Build a feedback log with causes and next tests
‘The exported text became unreadable at phone size’ leads to a typography test. ‘The portrait looks too flat’ suggests checking value structure. ‘My pen pressure stopped working’ calls for a device test, not another anatomy lesson.
Store an annotated screenshot beside each correction. A student who can identify an actionable cause develops independent problem-solving, while one who repeatedly hears only ‘make it better’ gains little.
33. A four-session digital painting pathway
Session one chooses output size, tests a brush and builds a five-value thumbnail. Session two blocks the main shapes in named layers. Session three adds purposeful colour, edges and texture while recording source assets. Session four saves the native working file, makes an appropriate export and checks it on the target device.
The plan works for children, secondary students and adult beginners when the software and complexity are adjusted appropriately. Each session should end with a file that can be inspected and a statement of what was learned.
34. The gold standard is work that remains intelligible and reusable
The final image should convey a clear idea at the intended viewing size. The editable source should remain understandable, and the export should be reliable. Brush presets, filters and expensive hardware are optional means toward those ends.
A modest drawing with clean layers, honest assets and a tested export can be more professional than a spectacular image that cannot be edited, credited or used in its intended format. Good digital art combines creative judgement with practical stewardship.
Original 2400 × 1800 pixel-and-layer study
Create an original 2400 × 1800 pixel canvas with a 4:3 aspect ratio. Divide it into a three-column, three-row planning grid, making nine 800 × 600 pixel tiles. Use one tile for the focal object, another to test light against dark and a third to test whether the image still reads when reduced. Save three distinct stages in the native editable format. The figures below illustrate raw pixel data; they do not measure actual RAM usage or compressed file size.
| Element | Working | Result |
|---|---|---|
| Canvas dimensions | 2400 × 1800 px | 4:3 raster canvas |
| Pixel count | 2400 × 1800 | 4,320,000 pixels |
| Nine planning tiles | 2400 ÷ 3; 1800 ÷ 3 | 800 × 600 px per tile |
| RGBA 8-bit channels | 4 channels × 1 byte | 4 raw bytes per pixel |
| One raw layer | 4,320,000 × 4 | 17,280,000 bytes |
| Three raw layers | 3 × 17,280,000 | 51,840,000 bytes, ignoring overhead |
| Working file | Native layered format | Editable project and revisions |
| Publication file | PNG or JPEG as appropriate | Verify crop, profile and viewing size |
A 2400-by-1800 canvas is not a universal recommended size and a 51.84-million-byte subtotal is not a reliable file-size prediction: compression, layer types, masks, caches, colour depth and application settings change real requirements. Its educational purpose is to show why dimension and layer decisions matter before the student opens a dozen enormous documents.
Learning pathways: one art standard, different tools
| Participant | First purposeful exercise | Useful proof of progress |
|---|---|---|
| Primary learner | Three-value animal silhouette using approved software | Clear form and saved image |
| Secondary art student | Five named layers and an original digital still life | Explains value, revisions and sources |
| Adult beginner | Canvas and export test on two screens | Correct dimensions and reliable files |
| Learner needing adaptation | Suitable input device or stabilised brush | Independent meaningful choices |
| Experienced illustrator | Colour-managed editable master and two publication crops | Legible versions with dependable provenance |
Frequently asked digital painting questions
What is digital painting?
Making original painterly or illustrative artwork with digital brushes and tools. It is often raster-based but may incorporate other digital techniques.
Do I need a drawing tablet?
No. It can improve pressure-sensitive drawing, but many basic composition skills can be practised with other input devices.
What is the difference between pixels and PPI?
Pixel dimensions describe how many raster samples the file contains; pixels per inch relates them to a physical output scale. Metadata alone does not create new detail.
Are more layers always better?
No. Layers are useful when they support clear revisions, but unnecessary unnamed layers can complicate the document.
What is a native working file?
An editable format such as Krita’s KRA, preserving application-specific painting features more fully than an ordinary flattened export.
Should I export PNG or JPEG?
PNG is often useful for sharp edges, text and transparency; JPEG can be suitable for gradients and photo-like images when lossy compression is acceptable. Verify the destination’s requirements.
Why do colours look different on two devices?
Displays, colour profiles, calibration, brightness and viewing conditions differ. Use colour-managed methods when consistency matters.
How many pixels are in a 2400 by 1800 painting?
4,320,000 pixels. With 8-bit RGBA, one raw layer is about 17.28 million bytes, before additional data or compression.
Can stabilisation help someone with shaky lines?
Some digital painting applications offer smoothing or stabilisation; test its benefits and trade-offs for the particular artist and brush.
Can I use images or brushes downloaded online?
Only under appropriate permissions and licence terms. Keep source records and credit creators when required.
Does a JPEG retain the editable layer stack?
Ordinary JPEG export is flattened and lossy. Preserve a separate working file for future changes.
Is a beautiful export proof of an honest portfolio process?
No. A trustworthy portfolio also describes authorship, relevant assets and the actual workflow.
Official software, file-format and reuse references
- Krita Manual: Introduction to Layers and Masks
- Krita Manual: Colour Managed Workflow
- Krita Manual: Drawing Tablets and Pressure Sensitivity
- Krita Manual: Freehand Brush Smoothing and Stabilisation
- Krita Manual: KRA Native Working Format
- Krita Manual: PNG and JPEG Export Formats
- Krita Manual: Saving, Exporting and Incremental Versions
- Creative Commons: Reusing CC-Licensed Content
Continue into the wider eduKate arts ecosystem
The gold standard: Make the picture legible and original, keep its working structure understandable, handle colour and reused materials honestly, and deliver a file that really works where it is meant to be seen.