How to Reduce the Size of Large Photo Archives Without Losing Quality (2026 Guide)
· 6 min read · Tools
A large photo archive eventually runs into a drive that will not hold it or a cloud bill you would rather not pay. The instinct is to compress everything. That instinct destroys archives.
This guide covers how to reduce size meaningfully without damaging what you are trying to keep — and when compression is genuinely the right answer.
The rule that governs everything
Lossy compression is irreversible. A JPEG re-saved at lower quality cannot be restored. The detail is not hidden; it is discarded.
Which produces the only rule that matters:
Never compress your master copy. Compress derivatives.
Keep originals untouched. Generate smaller versions for the phone, for sharing, for the tablet — from the originals, alongside them. When storage gets cheaper, and it always does, you still have everything.
Start with what costs nothing
Before touching image quality, remove what should not be there. This is usually where most of the space is, and it costs no quality at all.
Duplicates
Archives assembled from several sources — phone, cloud downloads, social exports — routinely contain a substantial share of duplicates.
Run dupeGuru or a similar tool. This is the single highest-value step, and it is free in every sense.
Near-duplicates and bursts
Ten nearly identical shots of the same moment. Keep the best one or two. Burst sequences in particular can consume enormous space for little added value.
Review before deleting — sometimes the differences matter.
Screenshots and junk
Screenshots, photos of documents, receipts, memes saved from elsewhere, accidental captures. Most archives are perhaps a tenth junk, and none of it needs archiving.
Sorting by file size surfaces surprises quickly.
Video you do not need
Video dominates size. A single minute can outweigh hundreds of photos. Reviewing video first gives the biggest return per decision.
Then look at the video
Once the obvious is gone, video is where the remaining size lives.
Convert to efficient formats. H.265 (HEVC) achieves roughly comparable quality at meaningfully smaller size than H.264. The trade-off is compatibility — older devices and some browsers do not play it.
A sensible compromise: H.264 for anything you want to play anywhere, H.265 for bulk archive storage where you control the playback software.
Reduce resolution when it genuinely does not matter. 4K footage of a birthday party rarely needs to stay 4K. 1080p is a large saving and, on any screen you actually watch it on, indistinguishable in practice.
Tools: HandBrake with a graphical interface, ffmpeg from the command line. Both are free.
Always convert copies. Keep originals of anything irreplaceable until you have verified the converted files play correctly.
Photos: the honest numbers
Photo compression yields less than people expect, because modern camera JPEGs are already compressed.
Lossless optimisation. Tools like jpegoptim and MozJPEG re-encode without discarding information, typically saving a modest percentage. Safe, and worth running on a large archive.
Quality reduction. Dropping a JPEG to around 85% quality is visually near-identical and can halve the file size. At 70% you begin to see artefacts in gradients and flat colour. Below that, visibly.
HEIC conversion. Apple's format is meaningfully more efficient than JPEG at equal quality. Excellent inside the Apple ecosystem; requires an extension on Windows.
Resolution. A 48-megapixel photo viewed on a phone is using a fraction of its pixels. If an archive exists to be looked at rather than printed, downscaling saves more than any codec change.
Again: on copies, never on masters.
Where the real savings are
Ranked by return, for a typical archive:
- Removing duplicates — large, free, no quality cost
- Deleting junk and screenshots — large, free
- Culling near-duplicate bursts — large, free
- Converting or downscaling video — very large, small quality cost
- Lossless photo optimisation — modest, no quality cost
- Photo quality reduction — modest, permanent quality cost
Note that the top three cost nothing in quality. Most people skip straight to number six, which is the least effective and the most damaging.
The tiered approach
For archives that genuinely will not fit, tier by value rather than compressing uniformly.
Tier 1 — irreplaceable. People, events, anything you would grieve. Keep at original quality, in two places. Do not touch.
Tier 2 — worth keeping. Ordinary photos of ordinary days. Lossless optimisation, and video downscaled to 1080p.
Tier 3 — probably disposable. Screenshots, documents, memes, blurred shots. Delete, or compress aggressively without concern.
Most archives are perhaps a fifth Tier 1. Protecting that fifth properly while being ruthless with the rest solves the problem without loss that matters.
Cheaper than compressing
Worth stating plainly, because the arithmetic is often skipped.
An external drive of several terabytes costs roughly what a couple of restaurant meals do, once. Cloud storage at archive scale costs a few euros monthly.
Set against the permanent, irreversible degradation of photos you cannot retake, buying storage is usually the correct answer. Compression is for cases where space is genuinely constrained, not for saving a modest amount on a drive you already own.
What not to do
Do not compress the only copy. Ever. Whatever the space pressure.
Do not re-compress repeatedly. Each lossy pass degrades further. Compressing an already-compressed file is the fastest way to ruin an archive.
Do not delete originals before verifying. Convert, check the results play and open properly, and only then remove the source.
Do not trust "optimise storage" as a backup. Cloud services offering to keep low-resolution copies locally are managing your device, not archiving your photos. If the account goes, so do the originals.
Do not compress before organising. Sort and deduplicate first, or you will spend effort compressing files you were about to delete.
Organising a large archive first.
A practical sequence
- Gather everything into one place
- Deduplicate with a proper tool
- Delete junk — screenshots, documents, accidents
- Cull bursts to the best one or two
- Measure what remains. Often the problem is already solved
- If still too large: convert video, keeping originals of what matters
- If still too large: lossless photo optimisation
- Only then: consider quality reduction, on copies
- Back up what remains — two copies, one elsewhere
Most archives never reach step 6.
Summary
Reducing archive size is mostly a deletion problem, not a compression problem. Duplicates, junk and redundant video account for the great majority of wasted space, and removing them costs nothing in quality.
Compress only when genuinely necessary, only on copies, and never on the photos you would be devastated to lose.
If part of your archive is still inside a social platform, get it out at full quality first. Open your Snapchat export here.