# Control OCR and image preparation

The OCR Settings that decide how scans become searchable text — the master switch, advanced image preparation's speed-for-accuracy trade, the barcode engine choice — and the restart habit that makes changes stick.

Product: content-central · Versions: 7.x · Audience: system-administrator · Time: 15 minutes · Last verified: 2026-09-05

Canonical: https://help.ademero.com/content-central/capture/control-ocr-and-image-preparation

**At the end of this guide you'll know which switch governs each part of the scan-to-searchable pipeline, the one trade-off setting worth testing on your own documents, and the restart habit that guarantees changes actually take effect.**

Everything lives in **Administration** > **System Settings** > **Capture Settings** (in the **Capture & Conversion** group), mostly under the **OCR Settings** subcategory.

## The OCR Settings

| Setting | What it does |
| --- | --- |
| **Enable OCR and auto-rotate** | The master switch: captured images get text recognition and upside-down pages get righted. Off, scans arrive as unsearchable pictures — on is the answer for almost everyone |
| **Use advanced image preparation** | The trade-off worth testing — its own description states it: "Enabling this option may slow capture speed, but boost accuracy for documents w/ color backgrounds." Shaded forms, colored paper, highlighter-happy offices: try it on |
| **Enable 2 pass OCR processing** | A second recognition pass for accuracy at more processing cost |
| **Store OCR in database** | Marked "Not recommended" by its own description — leave it unless support says otherwise |
| **Barcode engine** | Which engine reads barcodes during capture: the **Ademero Barcode Engine** ("the engine Content Central has always used" — every existing zone and QCard was set up against it) or **Cortexa BarVision**, built to recover barcodes "the current engine cannot read — faint or heavy print, skew, speckle and damage." Struggling barcodes are the reason to switch; test your QCards and [zones](/content-central/capture/read-data-from-fixed-zones-and-barcodes) after |

## The neighbors that shape image capture

In the **Image-Capture Settings** subcategory alongside: **Enable blank-page removal** (with **Remove pages smaller than (bytes)**) — the duplex-scanning essential that drops empty backsides; **Generate PDF/A-1b PDF documents** for archival-format shops; capture throttling and processing-thread counts for tuning server load. And under **PDF Electronic Capture**, **Use existing text in PDFs** skips re-OCR when an uploaded PDF already carries a text layer — "Improves capture performance," accurately.

## The restart habit

After changing settings here, **restart the Ademero Content Central Capture Service** (on the server, in Windows Services). Some of these settings apply live; at least one — advanced image preparation — quietly doesn't until the service restarts, and the settings page won't always tell you which is which. One restart after a settings session costs a minute and removes the whole question.

**Success check:** capture a test scan after your changes — it arrives searchable ([search finds its text](/content-central/search-and-find/why-text-you-can-see-isnt-searchable)), right-side-up, minus its blank pages. For the advanced-image-prep decision, run the *same* difficult document through both ways and compare the extracted text — your documents are the only benchmark that matters.

## What's next

- [Make already-captured documents text-searchable](https://help.ademero.com/content-central/search-and-find/make-already-captured-documents-text-searchable)
- [Read data from fixed zones and barcodes](https://help.ademero.com/content-central/capture/read-data-from-fixed-zones-and-barcodes)