10.3 · Settings and Preferences

Capture and OCR Settings

Capture settings control the region selection overlay and post-capture behaviour. OCR settings control the Tesseract engine used for text extraction.

Capture settings

Setting	Default	Description	UI location
`ShowDimensions`	`true`	Show live `W x H` pixel dimensions next to the selection rectangle	Settings → Capture
`PlaySound`	`true`	Play a shutter sound when a capture is taken	Settings → Capture
`CopyToClipboard`	`true`	Automatically copy the captured image to the Windows clipboard	Settings → Capture
`OverlayOpacity`	`50`	Darkness of the dimmed-out area around the selection (0 = clear, 100 = opaque). Range 0-100	Settings → Capture
`OpenEditorAfterCapture`	`true`	Open the annotation editor on the captured image immediately	Settings → Capture
`DefaultDelay`	`0`	Default pre-capture delay in seconds. Used when the toolbar delay combo is not changed	Settings → Capture
`DelayOptions`	`[0, 3, 5, 10]`	Values offered in the delay combo box	JSON only

Notes

OverlayOpacity affects only the region overlay — it has no effect on fullscreen capture.
Disabling OpenEditorAfterCapture returns you straight to the toolbar; the capture is still added to history.
DelayOptions can be changed by direct JSON edit to add, for example, 15 or 30 seconds.

Tip

Tip — Turn PlaySound off when recording screencasts so the shutter does not land in your audio track.

OCR settings

Setting	Default	Description	UI location
`Language`	`"eng"`	Tesseract traineddata language code. 29 languages ship in `tessdata/`	Settings → OCR
`AutoCopyText`	`false`	Copy extracted text to the clipboard automatically after the OCR popup opens	Settings → OCR
`PreprocessImage`	`true`	Apply contrast / threshold / denoise before OCR. Improves accuracy on screenshots	Settings → OCR

Notes

Language codes follow Tesseract conventions: eng, fra, deu, spa, ita, por, rus, jpn, chi_sim, chi_tra, kor, ara, and others.
You can chain languages by separating with +, e.g. "eng+fra", if both traineddata files are present.
The OCR engine is lazily initialised — the first extraction after launch takes 2-5 seconds longer while Tesseract loads.

Tip

Tip — Leave PreprocessImage on for screenshots of UI text. Disable it only when extracting from photographs where the preprocessing would over-sharpen.