Learn / Guide
Custom Cursor Blurry in Windows: DPI, Size and Asset Checks
Compare blurry custom cursors against a built-in scheme. Check pointer size, display scale, multi-size assets and application-specific differences.
A blurry custom cursor can come from the available image sizes, the design's own soft edges, or how an application displays it. First compare the custom pointer with a built-in Windows scheme at the same display scale and pointer size. Do not change screen resolution or assume every soft edge is a driver fault.
Windows cursor files can contain several images at different sizes. Microsoft documents that the system selects a matching image; it also notes that cursor sizes change in steps rather than at every DPI value. A project's “HiDPI” label is therefore not a guarantee of identical sharpness at every size, scale or application boundary.
1. Write down the two size settings
- In Settings → System → Display, select the affected display and note its Scale percentage.
- In Windows 11, open Accessibility → Mouse pointer and touch and note the pointer Size position. Windows 10 uses Ease of Access → Mouse pointer.
- Note the custom package name, release and Windows size variant.
- If you use more than one monitor, record which screen shows the softness instead of treating the setup as one scale.
Pointer size is adjustable independently of display scale. A larger pointer accessibility setting can request more image detail even when the display scale remains unchanged. Conversely, changing display scale affects much more than the arrow, so it should not be the first permanent change for a pointer-only complaint.
2. Compare one variable at a time
| Comparison | What it helps distinguish |
|---|---|
| Custom scheme versus a built-in scheme at unchanged settings | Package appearance versus a broader display or application issue |
| One custom size variant versus another from the same creator | Artwork proportions and supplied size coverage |
| Normal arrow versus Busy or Working animation | A static-file issue versus animation-specific assets |
| The same pointer over the desktop and an application | System scheme appearance versus an application-selected cursor |
| The same task on each connected display | A screen-specific pattern worth reporting, not proof of a particular DPI bug |
Use the scheme selection guide for the comparison, keeping your existing scheme name and extracted files. Restore the original selection after gathering the observations if you prefer it.
3. Read the Windows size coverage, not the Linux list
BreezeX documents separate Windows Regular, Large and Extra-Large groups. The visible artwork occupies different proportions of its image canvas. Its Xcursor pixel-size list describes another platform's format, so that list alone does not establish what a Windows download contains.
Check whether the creator supplies the Windows variant appropriate for your intended visible size. Renaming a file, enlarging a small bitmap in an image editor, or choosing a larger download preview cannot add original detail. Prefer creator-supplied multi-size assets or an official larger variant rather than a converted file from an unrelated mirror.
4. Separate intended softness from unsupported promises
The Comix Windows port is unusually explicit about its limits. Its README says the original artwork was designed for larger cursors, describes softness at the smallest pointer setting, and separates measured display-scale rows from assumed ones. It also documents narrower image-size coverage for animated roles.
Those are maintainer observations about that project, not DeskSpruce testing or a general formula for all Windows releases. Microsoft and a project's measurements can describe different size details; do not turn one package's table into a universal compatibility claim.
When to report the issue
If one package or one role is consistently soft, report the release, role, display scale and pointer size to its maintainer. If only one application differs, include that application's version and whether it uses its own cursor. Screenshots can document what you see, but record the actual settings too; a scaled screenshot can itself alter apparent sharpness.
If built-in and custom schemes both show a broader display problem, cursor-file replacement alone has not isolated the cause. Preserve those comparison results rather than repeatedly installing more packs.