When your Sparkasse Leipzig statement arrives as a scan or a photo of the page rather than a clean export, the numbers on it are just pixels until they are extracted. A statement that was scanned or photographed adds one more layer: the engine runs OCR, then verifies every extracted figure against the running-balance check. Anything that the scan may have blurred or distorted is caught against that arithmetic, so a smudge does not silently become a wrong number.
When a Sparkasse Leipzig statement you need to convert spans more than one page, Convert·Into merges every page into a single transaction table rather than treating each page as its own file. Longer statements that run several pages are handled as a single continuous table. The engine stitches the transaction table across page breaks, re-checks the running balance at every page boundary, and flags any row where the printed balance and the calculated balance diverge. You review only the flagged rows, not the whole file.
If the running balance on your Sparkasse Leipzig statement does not add up row by row, that is exactly the kind of mismatch Convert·Into is built to catch rather than pass through silently. The running balance is always verified against the statement's own printed closing figure. The engine steps through every transaction line, recomputes the balance, and compares the final calculated figure to the one printed at the foot of the statement. If the two numbers match, the extraction is consistent end to end; if a row does not add up, that row is surfaced for review while the rest of the file remains intact. Once you resolve any flagged rows, you export the reconciled data to Excel, CSV or QBO. Every column maps cleanly: date, description, debit, credit and balance land in separate fields with no merged cells and no split rows.