# Troubleshoot import and conversion Diagnose a file missing from the picker, a failed import, an unreadable scan, a stalled or refused Smart Score conversion, or wrong converted notation. > Type: Troubleshooting · Status: Stable · Platforms: iPadOS 17+, iPadOS 17+ · Last reviewed: 2026-07-18 ## First, separate import from conversion Import and Smart Score conversion fail in different places and are fixed in different places. - **Import** — `Import files…` or `Scan a score` in the library — copies a file onto the device and creates a piece. If the piece never appears, or its pages are unreadable, the problem is the source file or the scan. - **Smart Score conversion** — `Convert this PDF` in the reader top bar, or `Toolbox` (wrench) > `Convert to Smart Score` — reads an existing PDF piece and builds a structured notation version from it. It runs on the conversion server, so it can fail for account or network reasons even when the import was fine. Fix import first: conversion can only read the page images that import produced. ## A file is missing from the picker `Import files…` opens the system Files picker. Page Turner shows what the Files app shows; it does not keep its own connection to any storage provider. 1. Open the file in the iOS **Files** app first. If it is not there, enable and sign in to its storage provider inside Files, and let the file finish downloading. 2. Reopen `Import files…` in the library. If the file is visible in Files but cannot be selected in the picker, its type is not importable. The picker accepts: - PDF - Images (a photo or screenshot of a score) - MusicXML — `.musicxml`, `.xml`, and compressed `.mxl` - MEI (`.mei`) - MIDI - `.sheetturner` packages shared from Page Turner - `.imslp` bundles (IMSLP downloads; the title and composer inside are applied automatically) > **Limit** A generic ZIP archive cannot be imported. The only zip-based containers the picker accepts are `.mxl`, `.sheetturner`, and `.imslp`. LilyPond `.ly` source files are not supported. ## Import fails or produces an unreadable score A failed import reports the source system's wording — for example `Import failed: …` or `Couldn't import the scan: …`. Page Turner does not diagnose the file beyond that message, so check the source directly: 1. Import one file instead of repeating the whole batch, to find which file fails. 2. Open that file outside Page Turner — in Files or in the app that produced it — and confirm it opens, has content, and shows its pages in order. 3. For a cloud file, confirm it is fully downloaded to the device. A placeholder that has not finished downloading has no usable content to import. 4. For a PDF that asks for a password elsewhere, remove the password before importing. The import error will not name password protection as the cause. 5. For a scan, retake blurred, cropped, rotated, or shadowed pages with `Scan a score`, and confirm no page is missing. If the import completes but the title or composer is wrong, do not import the file again — a re-import creates nothing better. Long-press the piece in the library, choose `Properties`, and correct the metadata there. See [Metadata](/docs/library/metadata). ## Smart Score conversion does not finish Conversion needs a network connection and a signed-in account. Start it with `Convert this PDF` in the reader top bar, `Toolbox` (wrench) > `Convert to Smart Score`, or the `Create Smart Score Layer` step in the Trainer — each submits the same kind of server job for the open piece. While the job runs, the Trainer's `Create Smart Score Layer` step shows a progress card with a `Stop conversion` button: - `Stop conversion` stops waiting for the job and clears the card. The server pass may still settle on its own. - Closing the Trainer panel does **not** stop the job. It keeps running, and the Trainer button shows a dot when it is done. If the job ends with a message instead of a Smart Score, read it before resubmitting — two of them are account states that resubmitting cannot clear: - `Sign in required — the free tier allows 10 conversions per account.` — sign in with your Apple Account in Settings, then retry. - `Conversion quota reached on this account.` — this account has used its conversions; retrying the same job gives the same answer. > **Important** A long score is parked before it runs: a `Large score` alert shows the page count and the estimated cost, and the job starts only when you choose `Convert anyway`. Choosing `Cancel` parks the job again — nothing has failed, and nothing is running. If the source is a skewed or photographed scan, improve the pages before retrying: `Toolbox` (wrench) > `Edit This Score` > `Dewarp & Clean`, or rescan with `Scan a score`. > **Note** `Dewarp & Clean` works only on a scanned or photographed PDF page. On an engraved score it declines with a message instead of altering the page. ## The result contains wrong notation A finished conversion can still contain a wrong note, a missed trill, or a missing octave line. Correct the passage instead of reconverting the whole score: 1. Open the converted piece and tap the wrench to open the `Toolbox`. 2. Tap `Fix a Wrong Conversion`. 3. Draw a box over the smallest complete wrong passage. Pinch with two fingers to zoom. 4. Tag what is wrong — wrong note, missing trill, missing 8va — then re-scan the passage or send it for revision. 5. Compare the proposed replacement against the PDF before accepting it. Full procedure: [Fix a wrong conversion](/docs/smart-score/fix-conversion). > **Limit** A successful conversion status confirms that a notation file was generated, not that it is musically accurate. Review the result before relying on analysis, fingering, or Follow.