The short answer
Use Keyboard Ghosting Test to hold real combinations together and inspect the keys your browser receives simultaneously. Test the combinations you use, not just the biggest group you can press. Missing events may be blocking, rollover limits, reserved shortcuts or focus problems; they do not automatically prove a damaged switch.
NKRO and 6KRO describe reporting capability in a particular context. A browser sees events after firmware and OS processing, so its observed maximum is a lower bound for your attempt, not certification of an electronic specification.
Practical checklist
- Check individual keys with Keyboard Test before simultaneous combinations.
- Choose a preset and use its exact physical codes, including left-side modifiers.
- Start, keep the combination area focused and hold all expected keys together.
- Release fully between attempts and compare the strongest simultaneous set.
- Avoid reserved browser or OS shortcuts; Escape stops and Tab leaves the area.
- Compare documented connection or firmware modes only when supported; never infer unrestricted NKRO from one maximum.
Three different ideas behind the word ghosting
In ordinary gaming discussions, ghosting often means a key did not register while other keys were held. Strictly, phantom ghost input is an extra key that was not pressed. Blocking is suppression of an ambiguous combination, while a rollover limit describes a restriction on simultaneous reporting. These terms describe different behavior, and a missing browser event alone cannot determine which mechanism is involved.
A keyboard matrix arranges switches along shared electrical paths. Certain combinations can create ambiguity depending on the design, and firmware may avoid unwanted phantom input by blocking another key. A browser cannot probe those circuits or observe the keyboard’s internal scan. It receives the event stream that the device and operating system chose to expose, so this guide distinguishes what you observed from a proposed electronic explanation.
What 6KRO and NKRO usually mean
6KRO generally refers to reporting six non-modifier keys simultaneously in a particular operating or connection mode. Modifier keys may be represented separately. NKRO means n-key rollover, a broader capability to report many simultaneous keys, but practical behavior still depends on interface, firmware and application handling. Manufacturer documentation matters when a specification is important to your task.
Testpartout counts all currently delivered held identifiers, including modifiers, so its number is not directly a standardized 6KRO certificate. Observing seven keys in a chosen combination does not demonstrate every possible seven-key group, every connection mode or unlimited rollover. Conversely, observing fewer keys may reflect how you pressed them or a reserved shortcut. The careful statement is that this browser observed at least a particular simultaneous count during this run.
Start with individual-key evidence
Before testing rollover, use Keyboard Test to press and release keys individually. Confirm that the expected physical code arrives and examine the produced character separately. A key that does not appear by itself requires a different investigation from one that disappears only in a group. Repeating keydown while a letter is held is normal OS repeat behavior, not evidence that simultaneous count increased.
Mechanical feel, debris, remapping software and layout can all affect a practical typing complaint. Follow the manufacturer’s cleaning instructions rather than opening switches based on a browser trace. If a key works in a plain OS app but not in the browser, focus or reserved behavior may be the next step. A key that works here but not in a game may involve bindings or an overlay instead of rollover.
Use a real simultaneous attempt, not a union of taps
Keyboard Ghosting Test starts only after you press its Start button and focuses the combination area. Hold the expected group together, pause briefly and release fully. Its preset summary retains the strongest single simultaneous matching set. Pressing W, then Shift, then Space without overlap cannot complete Sprint + Jump because these separate observations never formed the expected held combination.
Repeat slowly and watch which key drops when another is added. Change one key at a time so the observation remains interpretable. Held state is keyed by event.code when available and repeat events replace the same key rather than inflating count. Keyup removes it. Blur, visibility loss and Stop clear held state because a release event can be lost when focus changes; stale keys should not be mistaken for phantom hardware input.
Choose gaming combinations that matter to you
Sprint + Jump uses W, left Shift and Space. WASD Stress uses W, A, S, D, left Shift and Space. FPS Left and FPS Right combine W with A or D, left Shift, left Ctrl and Space. These are optional practical examples, not universal gaming standards. A combination your game never needs is less useful than the actual movement and action keys you hold together.
MOBA uses Q, W, E, R and Space; Rhythm uses A, S, D, F, J, K and L. Free Test Mode allows another deliberate group without an expected list. The pattern of combinations can matter more than one maximum because different matrix paths and firmware decisions can behave differently. A keyboard accepting one group does not establish unrestricted acceptance of all groups with the same number of keys.
Physical codes, characters, modifiers and Fn
event.code identifies a physical-position convention, while event.key is the produced value influenced by layout, Shift and input methods. The presets use codes, so a non-US layout can show a different character while the expected physical code is present. The page shows both to make that distinction explicit. Left and right modifiers are separate codes; a right Shift will not silently satisfy a preset labeled left Shift.
Fn is frequently handled internally by the keyboard without an ordinary browser event. Media keys, layer changes and special firmware actions may also be processed before a webpage can see them. A missing Fn tile does not prove a faulty key. Modifier flags can provide context, but they do not grant access to firmware state or certify that an application interprets the group identically.
Reserved shortcuts and keeping control of the page
Ctrl, Alt, Meta, Tab, function keys and their combinations may invoke browser or operating-system commands. The tool deliberately does not promise to defeat all reserved shortcuts. It avoids trapping navigation: Tab can leave the area and Escape stops the test. Do not attempt combinations likely to close tabs, switch applications or trigger system actions as if their missing events were reliable hardware evidence.
The combination area must retain focus to listen. Selecting a preset or clicking Stop can move focus, and returning to the area may be necessary before another attempt. If the page clears held keys after a focus change, that is an intentional safety behavior, not a sudden rollover failure. Compare in an ordinary application and consult documented shortcuts when the same combination has different results across contexts.
Connection modes, follow-up and privacy
Some keyboards offer documented firmware or connection modes that change reporting behavior. Wired USB, Bluetooth and a dedicated wireless receiver can have different paths; a hub or adapter adds another variable. Compare only supported modes and keep the combination and application similar. Do not flash firmware or change undocumented settings based on a browser number. A manufacturer’s own recommended diagnostic may be appropriate if a claimed feature is essential.
Keystrokes, codes, combinations and counts remain in local page memory and are not sent to analytics or stored as a raw log. Do not type passwords or private text even in a local test. Only consent-controlled start and complete events may leave through the analytics helper. Persistent group-specific symptoms deserve a precise description and manufacturer follow-up rather than an unqualified claim that the browser certified or disproved NKRO.
Frequently asked questions
Does a maximum of six prove 6KRO?
No. The browser count includes modifiers and depends on the attempted combinations and delivered events.
Can repeated keydown inflate simultaneous count?
No. A physical code is represented once while held.
Why does Fn not register?
Many devices process it internally before standard events reach the OS or browser.
Is a missing preset key definitely defective?
No. Check simultaneous timing, focus, exact physical modifiers and reserved shortcuts before suspecting hardware.
Keep the follow-up specific
Explore Full Device Check for a local diagnostic checklist and Video Call Check for meeting preparation. Read methodology for browser limits and privacy for actual data handling. If this guide disagrees with the current tool, contact Testpartout with a correction without private device identifiers or sensitive data.