Deleted Photos From an SD Card and Unsure What to Do First

I have until tomorrow evening to recover photos from an SD card, and I only have access to one computer and the card reader built into it. I deleted a folder by mistake while moving files, then noticed that 147 .JPG images were missing. I ejected the card right away and have not copied or shot anything else on it.

What should I do first to avoid making recovery less likely? Is it safer to create an image of the entire SD card before attempting recovery, or should I start with a read-only scan? Should the card remain locked during every step, and is there anything I should avoid doing on the computer?

An SSD can make a deleted file genuinely unrecoverable within minutes. That’s the detail I’d keep in mind before installing recovery software, downloading a guide, or continuing to use the computer as normal. The odds aren’t controlled only by how recently the file was deleted. They also depend on the storage device and what has happened on it since.

SSDs are especially unforgiving because modern operating systems commonly send a TRIM command after data is deleted. TRIM tells the drive that the freed blocks may be erased in the background. Once the SSD clears those blocks, recovery software has nothing left to find. That means a document deleted from a laptop SSD may already be gone for good later that same day.

Memory cards and USB sticks tend to behave more like traditional hard drives in this situation. They usually leave deleted contents sitting in place until another write replaces them. That’s why photos removed from a camera card can sometimes be recovered weeks later, assuming the card hasn’t been filled with new shots. Time matters, but new activity matters more.

Stop the writes before trying anything

My first move after an accidental deletion would be to stop using the affected drive immediately. Don’t download a recovery program onto it. Don’t keep browsing while deciding what to do. Even ordinary background activity, including operating system logs, can write new information into the space where deleted data is still sitting.

If the missing files were stored on the computer’s system drive, shutting the machine down is the safest option. Recovery should ideally be performed from another computer or through a bootable USB stick, with the affected disk attached as a secondary drive. That setup reduces the chance that normal system activity will overwrite the exact sectors you’re trying to preserve.

This urgency makes more sense once you know what deletion usually does. In most cases, the operating system doesn’t immediately scrub the file’s contents from the disk. It removes the record that points to the file and labels its occupied space as available for reuse. That’s also why a huge file can disappear from view instantly instead of taking several minutes to erase.

Until something else is written into that available space, the old contents may still be present. Recovery is therefore less about reversing a deletion command and more about locating information the operating system no longer tracks. I think of it as a race against the next write operation, not a race against the clock by itself.

What recovery software can actually save

Recovery programs scan storage for those untracked contents and try to rebuild usable files. If the deletion was recent and the drive has seen little or no activity since, the result can be close to perfect. In a good case, the software may restore not only the files but also their original names and folder structure.

Results become messier once parts of the drive have been reused. A recovered file may still open but contain damaged sections because some of its sectors were overwritten. In other cases, the contents survive while the record containing the filename does not. The software may then recover a working file under a generic name, leaving you to identify and sort it manually.

That distinction is useful because “recovered” doesn’t always mean “fully restored.” A scan can find plenty of files while only some remain intact. The condition of the underlying sectors decides the outcome, and no recovery program can reconstruct data that has already been replaced or erased by an SSD.

This is also why I wouldn’t treat recovery software as a substitute for backups. Recovery is an emergency measure with no guaranteed result. A simple automatic backup of the folders that actually matter changes accidental deletion from a stressful rescue attempt into a quick restore.

My verdict is simple: stop using the drive first, attempt recovery from somewhere else, and set up automatic backups afterward so you don’t have to rely on luck twice.

Do not let Windows “scan and fix” or format the card if it prompts you. Eject it, slide the card’s lock tab to the locked position if it has one, then put it back in. The built-in reader is enough. You do not need a second computer, but you do need free space on the computer’s internal drive.

@binaryotter8744lab is right about stopping writes, though I would make a sector-by-sector image of the card before doing a long recovery scan. Run the recovery software against that image and save every recovered file to the computer, never back onto the SD card. If there is not enough room for an image, scan the locked card directly and still use the internal drive as the recovery destination.

Since you have a deadline, verify the recovered photos rather than trusting the file count or thumbnails. Open several at full size, especially ones from the beginning and end of the missing folder. A JPG can be listed as recovered yet still be partly corrupted. Keep the original card untouched until you have copied and checked everything you need.

The SD lock tab does not guarantee write protection. It only tells the reader that the card should be treated as locked, and some built-in readers ignore it. Use it, but still reject every repair, format, or “scan and fix” prompt.

With a deadline, start with a filesystem-aware deleted-file scan. That has the best chance of restoring the folder structure, filenames, and dates. If it finds little, switch to a deeper file-signature scan limited to JPEGs. That second method may recover the pictures under generic names, so an apparent failure to find the original folder does not mean the images are gone.

A full card image is the safer route, but I would not burn most of the available time on it if the card reads normally, reports the correct capacity, and produces no errors. If it disconnects, slows to a crawl, or throws read errors, stop scanning it directly and image it first. Whatever route you take, point the output folder at the computer’s internal drive and leave plenty of extra space for duplicates and older deleted photos.

Don’t start a long recovery scan before checking the folder you were moving the photos to. The move may have copied some images before the deletion, so search the computer by file type and date first, then recover only what is actually missing.

Plug the computer into power and disable automatic sleep before starting. A recovery scan can run for hours, and some built-in readers do not reconnect cleanly after the machine sleeps. Create a destination folder on the internal drive now, then confirm there is enough free space for more than the missing folder because signature scans often return duplicates and older deleted images.

With a deadline this close, check the recovery program’s export rules before committing to a scan. Some programs will spend a long time finding and previewing files, then require payment before they save anything. Run the fastest deleted-file scan first and export usable photos immediately rather than waiting for every possible scan mode to finish. Sorting filenames and removing duplicates can wait.

@n​anoworks5487’s suggestion to search the computer is worth doing, but put a short limit on it. Check the intended destination, Downloads, Pictures, and any temporary folder created by the transfer program. After that, move on to recovery. Once files are exported, open a sample at full resolution and copy the entire recovered set to another location if possible. Keep the original card unchanged until after the deadline.

The scan results need to show pixel dimensions and file size, not merely a JPG count. Recovery tools can surface embedded previews that look convincing but are much smaller than the originals. Prioritize the largest versions, export them to the computer, and check their actual resolution before assuming the recovery worked.

The built-in reader is the weak link nobody’s flagged. Laptop card slots are often cheaper and flakier than a five dollar USB dongle, and a reader that drops the connection mid-scan can leave you worse off than when you started. If you’ve got any way to grab a basic external reader before tomorrow, I’d do that over trusting the internal one for a multi-hour scan. Past that, @nanocoder563 has the right order: filesystem-aware scan first for the real folder and dates, signature scan only if that comes up short. And I’d quietly drop the deadline stress. On a card, nothing degrades because the clock hit evening. The photos sit there until the next write lands on them, so the smartest thing you can do tonight is simply not touch the card again until you’re actually running the recovery.

A card sitting on the desk cannot overwrite itself; a card put back in the camera for a quick check may get new index files or a format prompt. Keep it out of the camera, use the computer only for recovery, and save every result to the internal drive.

Realistically, if you deleted a folder cleanly and haven’t shot anything new on the card, this recovers fine most of the time. Camera cards are about the friendliest case there is for undelete. So take the tension down a notch. The people warning you off SSDs and TRIM aren’t wrong in general, but none of that applies to a plain SD card sitting in a reader.

Where I’d set your expectations lower is the folder structure. A move operation is messier than a simple delete, because the file records were already in the middle of being shuffled when things went sideways. You might get the original folder back with names and dates intact, or you might get 147 files with names like DSC or IMG followed by numbers and nothing else. Either outcome is a win if the pixels are all there. Don’t judge the recovery by whether the folder reappears.

The point @rocketwolf3502 raised about embedded previews is the one I’d actually watch for, because it’s the trap that makes people think they succeeded when they didn’t. Cameras write a small thumbnail inside every JPG, and some tools happily hand you those thumbnails as if they were the photo. They open, they look right, and then you find out they’re a few hundred pixels wide. Sort your results by file size before you get excited. A real photo off a modern card is going to be somewhere in the multi-megabyte range, not a few kilobytes.

On the imaging debate between the earlier replies, I lean toward skipping it if the card reads normally and reports the right capacity. Making a full image doubles your time and only pays off if the card is dying or the reader keeps dropping. If either of those is true, image first, no argument. If not, you’re spending your deadline photographing a healthy patient.

One habit that saves people here: run the fast deleted-file scan, export whatever it finds, verify a handful at full size, and only then decide if you need the deeper signature pass. Treat each scan as a separate rescue rather than waiting for one perfect run to finish everything. And leave the card in the reader once it’s working. Pulling it out and reseating it between scans is how flaky slots decide to act up at the worst moment.