November 10, 2026

Checksum Verification for Video Files: A Practical Guide

All articles

Checksum Verification for Video Files: A Practical Guide

Editor verifying a copied video file

Checksum verification is a way to prove that a video file you copied, moved or delivered is byte-for-byte identical to the original. You calculate a short fingerprint (the checksum) from the source file, calculate it again from the copy, and compare the two. If they match, the bytes are the same. If they differ, something changed and you copy again before you erase a card or send a master.

That is the whole idea. The rest of this guide covers what a match does and does not tell you, which algorithm to choose, how to build checksums into offloading and delivery, and what to do when the numbers disagree.

Posthive
Keep Video Projects Organized
Posthive brings version management, task tracking, and cloud connectivity together to help creative teams manage production workflows.
Visit Posthive

TL;DR

  • A checksum is a fingerprint of a file’s bytes. Matching fingerprints on source and copy mean the copy is identical, so a “copy complete” message alone is not enough.
  • Use a fast algorithm such as xxHash for offloads and MD5 or SHA-256 where recipients expect them. Always compare using the same algorithm on both sides.
  • Generate checksums at first offload, keep them in a manifest (ASC MHL is the format studios ask for), and ship the manifest with every delivery.
  • A match proves the file did not change. It does not prove the clip plays correctly, where it came from or who touched it.
  • On a mismatch, do not format the card. Re-copy from the source and verify again.

Table of Contents

What a checksum proves about a video file

A checksum is produced by running every byte of a file through a hash function. Change a single byte and the result changes completely. The Digital Preservation Coalition describes fixity checking as generating a checksum at one time and comparing it with one generated later, where a match indicates the bitstream has not changed.

For video work, that translates into one clear promise: the file on the destination drive has the same bytes as the file on the source. A TetherX help page on video checksums puts it the same way for downloads: a match confirms the file has the same bytes as the recording whose checksum was stored.

This matters more for video than for most files. Camera originals are large, they are often the only copy until you offload, and a single flipped bit inside a long-GOP clip can show up weeks later as a green block or a clip that will not open.

Why file size and “copy complete” are not verification

A finder window that shows the same file count and size feels reassuring, but it is a weak signal. According to Imagine Products, “copy complete” only confirms that the copy process reported completion. It does not prove bit-for-bit identity and it leaves no auditable checksum history. Only comparing checksums checks the actual byte sequence, which is why offload tools verify by checksum instead of trusting the copy dialog.

Camera makers say the same. ARRI advises using checksum-verified backups as a minimum standard for original camera data, and warns that simple file-copy methods can miss missing or corrupt files before the media is erased.

Why file size and "copy complete" are not verification

What a matching checksum does not tell you

A checksum answers one question: did the bytes change? Be precise about the limits, because they decide which other checks you still need.

  • It does not say what changed. The DPC notes that checksums do not tell you what changed, when or by whom, or how to reverse or repair a change.
  • It does not prove origin. TetherX states that a matching checksum alone does not establish the camera source, recording time, chain of custody or authenticity of a cloud record.
  • It does not prove the clip is healthy. If a camera wrote a damaged file, a perfect copy of it has a perfect matching checksum. Verify that the footage plays through before you wipe the card.
  • It does not contain the content. Emby’s guide to file checksums points out that a checksum can show input bytes changed, but it holds none of the missing original, so you cannot rebuild a file from it.

The DPC’s advice is to pair checksums with other measures: information security, logging and audit trails, multiple stored copies with fixity checking and repair, and backup and disaster-recovery processes. In practice that means a checksum is the first layer of a workflow, not the whole workflow. If you want the backup side, our guide to two verified copies first in a small-team video backup workflow covers it.

Choosing a checksum algorithm

You will meet four names again and again: MD5, SHA-1, SHA-256 and xxHash. For detecting accidental corruption in a video file, all of them work. They differ in speed, in how widely recipients accept them, and in how hard they are to forge on purpose.

Algorithm Type Typical use in video Notes
xxHash (64-bit or 128-bit) Non-cryptographic Fast offloading on set Very quick on large files, so verification does not slow the card turnaround
MD5 Cryptographic (legacy) Delivery lists, broad compatibility Accepted almost everywhere, fine for catching copy errors
SHA-1 Cryptographic (legacy) Older MHL files Still read by many tools
SHA-256 Cryptographic Archives and evidence Slower, stronger against deliberate tampering

Two specifications show what the industry actually uses. MediaHashList.org says the MHL specification is meant to ensure the complete source media is transferred without alteration, and that it uses SHA-1, MD5 and xxHash. Netflix’s data management guidance states that all copies of original camera and audio files should include checksum verification using xxHash64be, xxHash128 or MD5.

A simple rule: use what the receiving party names. If nobody names one, pick xxHash for speed during offload and SHA-256 for long-term archives or anything that could be disputed.

Compare like with like

A comparison only means something when both sides use the same algorithm over the same bytes. Emby’s guide stresses that a whole-file comparison must use the same algorithm and the same byte range on both sides, and that a filename or modification date is not automatically part of a content hash. So renaming a file does not change its checksum, and an MD5 value can never be compared with an xxHash value.

How to verify checksums step by step

Here is the workflow we would hand a small crew. It follows the practice that WITNESS recommends for archiving video: capture hashes early, such as at first offload, so you can show later whether files stayed intact.

  1. Offload with verification turned on. Use a copy tool that calculates the source checksum, writes the file, then re-reads the destination and compares. Do this to at least two destinations.
  2. Save the checksum report. Keep the list of file names and hashes in a text file or an MHL manifest next to the media, not in someone’s head.
  3. Confirm playback. Spot check the clips, or run a full scrub on important material, before the card is formatted.
  4. Verify at every hop. Whenever the footage moves (to an editor’s drive, to the cloud, to a colorist) recalculate and compare against the saved list.
  5. Verify the delivery. The recipient checks the files against the manifest you sent.
  6. Keep the report with the project. Store checksum reports alongside camera reports so you can re-verify archives years later.

Verifying a single file by hand

You do not need special software for a one-off check. On macOS or Linux, the built-in shasum -a 256 filename or sha256sum filename command prints a SHA-256 value. On Windows, certutil -hashfile filename SHA256 in Command Prompt does the same. Run it on both copies and compare the strings. For a folder of hundreds of clips, a dedicated verification tool is far less error-prone.

Verifying against a manifest

Verification tools can read a manifest and check every file in one pass. The TrueCheck manual describes calculating checksums (including XXHash-64BE and MD5 or SHA types), generating MHL files to seal a folder, and comparing results against an MHL or checksum text file. A successful run reports the number of files verified with a green completed status.

Checksums and the ASC MHL manifest

A loose text file of hashes works for a small job. Once footage passes through several hands, a structured manifest does the job better. ASC MHL, described by Imagine Products, stores hashes in a human and machine-readable XML format and labels each as original, verified or failed. That builds a chain of custody: you can see where a damaged or duplicated file entered the history.

Netflix recommends generating a unique machine-readable checksum manifest, using the ASC MHL format, and carrying it through subsequent transfers, including final delivery. If you work for a studio or streamer, that is likely your requirement, not a preference.

Checksums for client delivery

Delivery is where checksums quietly save projects. A client downloads a 40 GB master over a flaky connection, the file opens, and a week later the final export stutters at minute 31. A published checksum lets the client find that out in minutes, before air.

A good delivery package has:

  • The media files, named consistently.
  • A checksum manifest (MHL or a plain text list) stating the algorithm used.
  • A line telling the recipient how to verify, for example the command for their operating system.
  • A note of the version delivered, so the checksums match the right cut.

The same discipline applies to client review. If you want to see how production teams structure remote shoot client review practices, including how checksums fit between production, post and delivery, that is a good companion read.

Two more of our guides cover the surrounding process: securing masters with a preview to master workflow and a delivery receipt workflow so you have a record of what the client received and when.

What to do when checksums do not match

A mismatch is the system working. It caught a problem while you can still fix it.

  1. Stop and keep the source. Do not format the card or delete the original until a verified copy exists.
  2. Check the comparison first. Confirm both sides used the same algorithm and that you are comparing the same file, not an older version with the same name.
  3. Re-copy and re-verify. Copy again from the source, ideally to a different port, cable or drive, and recalculate.
  4. Try another route. If it fails repeatedly, suspect the cable, card reader, drive or the source media itself.
  5. Log it. Record which file failed and what you did. In an ASC MHL history, a failed entry means the current copy is not identical to the previous one, and the record is useful.
  6. Escalate if the source is bad. If the original itself reads inconsistently, stop using that card and contact the camera or media manufacturer.

Note that a checksum cannot repair a file. A second verified copy can.

Common mistakes

  • Verifying only once. A checksum at offload says nothing about a drive that fails in month three. Re-verify archives on a schedule. Our three copy archiving checklist gives a rhythm for this.
  • Mixing algorithms. A manifest in xxHash cannot be checked with an MD5 tool.
  • Checksumming after editing. A checksum belongs to one exact version of a file. Re-export it and the value changes, which is why version records matter. See video version control for creative teams.
  • Treating a match as proof of authenticity. Integrity is not authorship. Pair checksums with audit trails, as in our piece on a four phase audit trail.
  • Forgetting the paperwork. A hash list sitting on one laptop is lost with that laptop. Store it with the project.

Our perspective

We build Posthive around a simple belief: the checksum is only as useful as the workflow that carries it. Teams rarely lose footage because they did not know what a hash is. They lose it because the manifest lives on one drive, the approved cut is not obvious, and nobody can say which version the client actually received.

That is why we designed Posthive so the client keeps ownership of deliverables and versions, with version control, review, approval tracking and secure media delivery in one workspace. When the files, the versions and the approvals sit together, a checksum report becomes one more piece of a clear record instead of a stray text file.

We also think access should not be rationed. Our plans include unlimited free seats, so editors, producers and clients can all see the same delivery and the same history without anyone counting licenses.

Lorenz

Keep checksums, versions and delivery in one place

Checksums protect the bytes. A workspace protects the process around them: who reviewed which version, what was approved, and what was delivered. Posthive brings project management, version control, collaborative review, client approval tracking and secure media delivery into one cloud workspace for creative teams, agencies, freelancers and production companies. There is a free plan, Pro is $49 per month and Enterprise is $99 per month, and every plan includes unlimited free seats.

Posthive

See the plans on the Posthive pricing page, or look at how Posthive works for your team.

FAQ

Is checksum verification the same as file integrity verification?

In practice, yes. File integrity verification is the goal, and a checksum comparison is the usual method. You calculate a fingerprint from the original and from the copy, and a match means the file’s bytes did not change. It does not check that the content is good, only that it is unchanged.

Which checksum should I use for video files?

Use the one your recipient or studio specifies. Netflix lists xxHash64be, xxHash128 and MD5 for original camera and audio files. If no one specifies, xxHash is a fast choice for offloading, and SHA-256 suits long-term archives. Whatever you pick, use the same algorithm on both sides of every comparison.

Do I need to send a checksum with every delivery?

It is a good habit for large masters and for anything a client will archive. A manifest lets the recipient confirm the download is intact before they rely on it. Studio deliveries often require one, with ASC MHL recommended by Netflix as the manifest format.

Does a matching checksum prove nobody edited my video?

It proves the file matches the version whose checksum you recorded, so any change to that file would show. It does not prove who made the original, when it was recorded, or that the recorded checksum itself was not swapped. For that you need audit trails and controlled storage as well.

Sources