Teachers end up doing the same three file jobs: turning paper into something a class can open, cutting one document into the parts different groups need, and getting a recorded explanation small enough to send. Each has a working order, and the order is the part that usually goes wrong.
On the paper side it is build, then split, then label. Building the PDF first means the split and the page numbers apply to the final document rather than to a draft. On the video side it is record, convert, then compress, because compressing a WebM and then converting it re-encodes the footage twice. Every step here runs in the browser, which matters when the material contains student names or an exam that has not been sat yet.
Turn paper into one handout PDF
A worksheet photographed on a phone is a set of images, not a document. Image to PDF puts the pages in order and writes one file with real pages, which is what a class portal, an email or a print queue expects.
Add the photos, drag the cards until the order matches the handout, then choose A4 or Letter, portrait or landscape, the margin, and whether each image should fit inside the page or fill it. JPG and PNG photos are embedded without re-encoding; WebP, AVIF, GIF and BMP are converted first, and a GIF contributes its first frame because a PDF page is a still image. There is no fixed image count, so a ten page worksheet is a single run, though every page is held in memory while the file is assembled.
Two limits are worth planning around. The tool has no perspective correction, so a page photographed at an angle stays skewed; Image Cropper can rotate in 90 degree steps and trim the frame, but it will not straighten a trapezoid. And cropping after the PDF exists means exporting the pages back to images, so trim the desk and fingers out of the photos before you build the file.
Image to PDF
Put worksheet photos in page order and export one handout PDF, with the page size and margins you choose.
Split, merge and label for different classes
Once the handout exists, class logistics become a matter of page ranges. The order is deliberate: split first, merge second, then number and mark, because a split run resets what counts as page one.
Split one paper into group sets
Split PDF has three modes: every page, custom ranges, and fixed size chunks. Type ranges as numbers separated by commas, for example 1-3, 5, 8-10, and each comma separated segment becomes its own output file. Pages are copied across as they are, so text stays selectable and links keep working.
The limit is encryption. An encrypted PDF has to be unlocked before its pages can be read, so remove the password first and split the unprotected copy.
Merge the term into one file
Merge PDF joins the parts into a single document in the order you drag the cards, with no fixed file limit and no recompression. Keep the merged copy separate from the split outputs, because merging a set that was just split will not reproduce the original unless the order is exactly the same.
Number the pages so a printed set stays in order
Add Page Numbers stamps a new text layer on every page. Set the starting number, a format that can include the total page count, the font size and the font, and skip the first page when it is a cover sheet. The original content is never flattened, so the text underneath stays selectable and searchable.
Mark each copy without changing the content
PDF Watermark applies text to all pages or a custom range, in a tiled or centered layout, with colour, opacity and rotation. Use it to mark a draft, a class set or a year group. It cannot open an encrypted PDF, so watermark before you protect a file, never after.
Split PDF
Cut a handout into single pages, page ranges or equal chunks, with a preview before you commit.
Pull text out of a worksheet
A scanned worksheet is a picture of a page, so selecting the text does nothing. Which route to use depends on what you already have.
If the pages are photos or screenshots
Image to Text accepts one image or several. Choose the language of the text and run recognition. English is built in, and eleven more languages including Chinese, Japanese, Korean, German, French, Spanish, Russian and Arabic can be picked from the language menu. The first run downloads about 6.5 MB of recognition engine and language data, and the browser caches it, so later runs start almost instantly.
If the material is already a PDF
PDF to Text handles both kinds of page. Pages that carry a real text layer are lifted out directly, which is instant and exact. Pages that are just a picture are rendered and recognised with OCR. The summary next to the result says how many pages came from each route, so you know which part deserves a closer proofread.
Where this fails is handwriting. The engine is trained on printed text, so a stack of handwritten answers will come back as a rough draft at best, and probably not even that. A typed worksheet, a clean scan or a screenshot is the reliable case.
PDF to Text
Reads the text layer instantly and runs OCR on scanned pages, all in the browser.
Record a lesson and send it
A recorded explanation is only useful if it plays on the device a student actually has. That makes format and size part of the workflow, not an afterthought.
Record the screen and your voice
Screen Recorder captures a whole screen, a window or a single browser tab, with optional microphone audio mixed into the recording. The output is WebM with VP9 video, with no watermark and no hard time limit. Plan for sessions of about thirty minutes or less, because the recording is held in browser memory, and use pause between sections rather than recording one long take.
Convert if the player refuses WebM
WebM plays in Chrome, Edge and Firefox. When it has to open anywhere, including Safari, an older school machine or a learning platform that inspects the file type, convert it to MP4 with H.264. The converter runs locally and shows live progress; a few minutes of footage normally finishes well under a minute on a desktop.
Compress for the class group chat
Video Compressor lets you set output format, quality, resolution and frame rate, so a 1080p recording can come down to 720p or 480p for a chat app. Clips under 200 MB and a few minutes long give the most comfortable results, because every frame is processed on your own machine. The first run downloads the FFmpeg engine of about 30 MB, and later runs reuse it from cache.
Add captions for students who need them
Add Subtitles burns SRT, VTT, ASS or SSA captions into the video. You can set font size, text colour, outline colour and weight, position at the bottom or top, and the vertical margin. Two limits matter. The renderer bundles a single Latin font, so non-Latin scripts are not covered. And the clip is fully re-encoded because captions are drawn onto every frame, so export takes roughly half the video length to the full length.
What these tools cannot do
- No handwriting recognition. OCR is trained on printed text, so handwritten answers and notes come back as a draft with mistakes, not a usable transcript.
- No perspective correction. Rotation in 90 degree steps and a crop frame are available, but a page photographed at an angle keeps its trapezoid shape.
- No grading and no question generation. These tools change file formats. They do not read answers, score them or produce a worksheet.
- The recorder is not an editor. It captures the screen and stops. Cutting a section out or joining two takes needs Video Trimmer and Video Merger, in separate passes.
- No non-Latin subtitle font. Burned-in captions are limited to one bundled Latin font, so CJK, Arabic and Cyrillic scripts are not covered.
- No learning platform integration. There is no upload to Google Classroom, Moodle or any other LMS. You download the file and upload it yourself.