Batch Video/Photo Culler for Creators
And writes cull_plan.md instead of deleting a single file.
One pass over the card groups every burst shot within 3 seconds of the last one, finds every exact duplicate by SHA-256, picks a keeper in each group.
And writes cull_plan.md instead of deleting a single file.
$119, once. No subscription, no renewal, no per-shoot fee. There’s nothing to cancel, and your media never went anywhere to be kept.
The card, the couch, and four hours you won’t get back
You get home at nine. The card holds everything from a six-hour day, and most of it is the same photograph taken four times because you held the shutter down. Buried in there are the frames you’ll actually deliver.
So you start clicking. Arrow, look, arrow, look, reject, arrow. An hour in, your eye is gone and you’re keeping the third frame of every burst on autopilot, which is not a decision, it’s a reflex. And the shots you imported twice, back when you weren’t sure the first import took, are still sitting in the folder doubling the count.
The cull isn’t hard. It’s long. It’s the wall between the shoot and the edit, and it’s why the gallery goes out on Thursday instead of Monday.
How does it decide what’s a burst?
By name and by clock, not by looking at the picture.
- Point
MEDIA_FOLDERat the shoot. It walks the folder and every subfolder, inventorying.cr3,.nef,.arw,.dng,.heic,.mov,.mp4and the rest of the list. - It SHA-256 hashes each file to catch exact duplicates, skipping anything over
HASH_MAX_MB(200 by default) so a long clip doesn’t stall the run. - It groups a burst when files share a base name once the trailing sequence number is stripped,
IMG_0001next toIMG_0002, and were written withinBURST_SECONDSof each other. Three seconds by default. Then it picks a keeper and writes the plan.
What lands in the plan
Four files and two scripts, and none of them touch your media.
- Collapses
IMG_0001throughIMG_0009into one burst when they were written three seconds apart, so you make one decision instead of nine. - Keeps the largest file in each group by default, on the theory that the highest-resolution export is usually the biggest one. That’s a crude proxy, and the tool says so rather than dressing it up.
- Takes a
SCORES_CSVoffile,scorerows from any scorer you like, a local vision model, a cloud API, something you wrote on a Sunday, and uses real quality scores to pick keepers instead of file size. - Hashes out exact duplicates into
DUPLICATES.md, which is where the card you imported twice finally shows itself. - Writes
move_culls.shandmove_culls.ps1. They move culls into a_Culledsubfolder. They never delete. And they don’t run until you run them. - Never writes a single byte inside
MEDIA_FOLDER. It reads, and it writes its plan somewhere else entirely. - Reads raw and video in the same sweep,
.cr2and.dngalongside.mp4and.m4v, so a hybrid shoot doesn’t need two passes and two tools.
Who this is wrong for
For the working photographer or videographer who comes off a job with a card and no second pair of hands. One person, one machine, a few hundred to a few thousand frames a shoot. It is the wrong tool if what you want is a computer telling you which frame is sharp and whose eyes are open. It does not look at the picture. It reads names, timestamps, file sizes and hashes, and that is all it claims to read. If you already have an assistant who culls for you, you have the better system.
Photo Mechanic, Lightroom, and the part neither of them does
Both of those are better than this at what they’re built for, and it isn’t close.
Photo Mechanic is the fastest way anyone has ever built to look at pictures. Full-size previews render before your finger leaves the arrow key, and when the cull is a judgment cull, choosing between two good frames on expression, nothing else comes near it. Buy it. This can’t do that, and it isn’t trying to.
Lightroom is better at everything downstream of the decision. Develop, catalog, sync, export, deliver. Keep it.
Here’s the part nobody wants to say out loud: most of what’s on the card was never a judgment call. It’s nine frames of one moment three seconds apart, plus a duplicate set from an import you don’t remember. That’s arithmetic, not taste, and it’s the part that eats the evening. The arithmetic gets done first, by something that doesn’t get tired at 11pm, and hands you a plan. Then you open Photo Mechanic and go be a photographer about the twelve frames that are genuinely close.
Nothing leaves, nothing dies
There is no cloud dependency of any kind, which for a shoot under embargo or NDA is the entire argument. Your media gets read and never written to, never moved, never renamed. The only things created are CULL_PLAN.md, cull.json, DUPLICATES.md and the two move scripts, all under OUTPUT_FOLDER at 0600 permissions. Even the scripts only move culls into _Culled. Nothing gets deleted, by the tool or by anything it generates, so undoing a bad cull is dragging a folder back.
$119, and what shows up
Paid once. No renewal, no seat, no per-shoot charge, no upsell at 400 files. You get the culler, the plan writer, the duplicate finder and both move scripts. Setup is two paths in a .env file. The first CULL_PLAN.md is written on the first run, the same evening you get home, and hashing is the slow part, which is exactly why HASH_MAX_MB is there to cap it.
Security and privacy
Reads customer/creator media folders that may be large and personally significant (irreplaceable shoots). Security and non-destructiveness are designed in.
Principles
– Local-first. Nothing touches the network. No cloud dependency, no telemetry, no external calls of any kind.
– Never deletes. Never moves on its own. The tool only reads MEDIA_FOLDER (for inventory and SHA-256 hashing) and writes its own plan package under OUTPUT_FOLDER. The only thing that can move a media file is a generated move_culls.sh / move_culls.ps1 script, and only when the user reviews and runs it themselves.
– Least privilege. Reads MEDIA_FOLDER, writes only under OUTPUT_FOLDER.
Protections implemented (v0.1.0)
– Path-traversal defense — every output filename is sanitized (safeSegment) and every write is checked with assertWithinRoot(), so the plan package can’t escape OUTPUT_FOLDER.
– Fail-fast config validation — clear errors for missing/invalid MEDIA_FOLDER, OUTPUT_FOLDER, SCORES_CSV, BURST_SECONDS, or HASH_MAX_MB before any file is touched.
– Restrictive permissions — generated plan files (CULL_PLAN.md, cull.json, DUPLICATES.md, move_culls.sh, move_culls.ps1) are written 0600.
– Secrets discipline — .env and generated cull-plan/culled-media folders are git-ignored; no secrets or customer media in the repo.
– Bounded hashing — files larger than HASH_MAX_MB are skipped for duplicate-hashing (still grouped by burst) to bound memory and runtime on large video files.
– Review-then-run move scripts — move_culls.sh / move_culls.ps1 carry a commented header warning to review before running, use mv -n / -ErrorAction SilentlyContinue (skip rather than clobber on conflict), and only ever move files into a _Culled subfolder — never delete.
Known limitations / roadmap
– Burst grouping and duplicate detection are metadata/hash based; without a SCORES_CSV, keeper selection is a size-based heuristic, not true image-quality scoring.
– SCORES_CSV is a plug-in seam for an external vision-model scoring pass; this tool does not compute quality scores itself and does not bundle or call any vision model.
– No file-content inspection beyond hashing; the tool does not open, decode, or analyze media contents beyond size/mtime/hash.
– Exact-duplicate detection is content-hash based (SHA-256); it will not catch near-duplicates re-encoded or re-saved with different bytes (that’s what burst grouping + SCORES_CSV are for).
Reporting
Report suspected vulnerabilities privately to the maintainer before public disclosure.
Frequently asked questions about Batch Video/Photo Culler for Creators
What does batch video software for creators actually do?
Batch video software of this kind is built to solve one heavy problem, the sheer volume of footage and photos a creator brings back from a shoot. A single session can leave thousands of frames and long stretches of video, and picking the keepers by hand takes hours nobody has. The Batch Video/Photo Culler for Creators uses AI to cull and tag a large local library, so the likely keepers rise to the top and the obvious throwaways get pushed down, all without uploading a single file to the cloud.
The work happens on your own machine, pointed at a local folder of shoot footage or photos. The AI runs through the library, tags what looks worth keeping, and lets you review the tagged results and export your selects. That is the loop. You are not scrubbing through everything frame by frame, and you are not waiting on an upload for a library that might run into terabytes. For creators shooting high volume, keeping the process local is the only realistic option, because moving that much media to a hosted service is slow and expensive.
It helps to remember what all that footage represents. Every original clip and frame is your work, and under United States law copyright exists automatically in an original work of authorship the moment it is fixed, a point the U.S. Copyright Office makes in its overview, What is Copyright. Batch video software does not create or register that right, but it does help you find, tag, and keep control of the originals that carry it, which is a practical part of protecting the value of what you shot.
The tool is a one-time purchase, not a subscription, and it does not bill you monthly for cloud storage you do not want. That pricing choice matches how it works. Your footage never leaves your machine, so there is no hosted library to pay rent on. For a creator who already owns plenty of local storage, paying once for the culling and keeping the files at home is a straightforward trade.
Here is the honest framing. The culler does not edit your footage or make creative decisions for you. It narrows thousands of options down to a workable set of likely keepers and tags them, so the human review starts from a short list instead of a mountain. Aftershoot, Photo Mechanic, and Narrative solve pieces of this, and the differentiator here is doing the culling on huge local libraries where cloud upload is not practical. You can see how the tool fits photographers and video creators on the creators and agencies industry page, and judge whether local culling saves you the hours you currently spend sorting by hand after every shoot. The difference it makes is felt on the first shoot you run through it. Where a manual sort might eat an entire evening, the local pass hands you a tagged short list in a fraction of the time, and you spend your attention on the frames that actually matter. For a creator shooting week after week, that recovered time is the real product, and the culling is simply how it gets delivered.
Is my footage safe with this tool, and where does it live?
The short answer is that your footage lives where it already does, on your own machine, and the culler works on it there. Nothing is uploaded to a cloud service to be processed. For a creator, that is not a small detail. Shoots often contain people, and photos and video of identifiable people are personal information that deserves care, especially when a shoot involves clients, models, or minors. The FTC’s guide, Protecting Personal Information: A Guide for Business, frames the job as five steps, take stock, scale down, lock it, pitch it, and plan ahead, and keeping media local is a direct way to act on the lock-it and scale-down parts.
Batch video software that runs locally means you are not scattering copies of a client’s images across a hosted platform you do not control. Take stock becomes easier, because your footage sits in your own folders where you can actually account for it, rather than in some service’s storage that you would struggle to inventory. Scale down follows, because you are keeping one working library instead of duplicating it into a cloud you did not need for the culling in the first place.
This matters most when a shoot carries a duty of confidentiality. A wedding, a private client session, or a campaign under embargo all involve images the subjects expect to stay controlled. Keeping the whole culling pass on your machine means the sensitive frames are never handed to a third party just so an algorithm can sort them. The people in your footage stay out of anyone else’s storage, which is the plainest form of the protection the FTC describes.
It is fair to be clear about the limit. Local processing is a privacy posture, not a security product, and it does not replace your own habits, backing up your drives, controlling who can sit at your machine, and disposing of old media properly when a project ends. The tool keeps the footage on your device and adds no cloud copy of its own. What you do with the drives after that is still on you, and the FTC’s plan-ahead and pitch-it steps are a good checklist for it.
For a creator weighing this against cloud cullers, the trade is straightforward. You give up the convenience of someone else’s servers and you get to keep every frame under your own roof. Batch video software built to run locally is aimed squarely at the photographer or videographer who would rather not upload a client’s shoot at all. You can read how the tool keeps large libraries on your own machine on the Batch Video/Photo Culler product page and decide whether keeping your footage local is worth more to you than a hosted convenience. Weighed honestly, the case for local processing comes down to trust. You already trust your own drives with the footage. A cloud culler asks you to extend that trust to a service and its servers, for the sake of a convenience you may not need. Keeping the whole cull at home means there is one less party who ever touches your clients’ images, which is the safest arrangement of all.
How is the Batch Video/Photo Culler different from cloud culling tools?
The clearest difference is where the work happens. Tools like Aftershoot and Narrative lean on cloud processing, which means your footage travels to someone else’s servers to be sorted. The Batch Video/Photo Culler for Creators runs the cull on your own machine against a local folder, so nothing leaves your device. For a creator with huge libraries, that is not a preference, it is often the only workable path, because uploading terabytes of raw footage is slow, costly, and sometimes simply impossible on a normal connection.
That local approach ties directly to keeping control of your own work. Every original clip and frame is a copyrighted work the moment you shoot it, since copyright attaches automatically to an original work once it is fixed, as the U.S. Copyright Office explains in its overview, What is Copyright. Keeping the culling on your machine means your originals never sit on a third party’s storage during the sort, so the files that carry your rights stay under your control from capture through export. Batch video software that never uploads is one way to keep that chain of custody entirely in your hands.
Price is the second real difference. Cloud cullers tend to run on subscriptions, and some pair that with ongoing storage fees. The culler here is a one-time purchase with no recurring cloud bill, which fits a tool that uses your own storage rather than renting you theirs. Over a few years of steady shooting, paying once rather than monthly is a meaningful gap, especially for a creator whose library only grows.
There is a tradeoff worth naming. Local processing leans on your own hardware, so a faster machine culls faster, and a very old laptop will feel it. Cloud tools hide that behind their servers. What you get in return for using your own machine is that your footage never leaves it and you are not paying rent on storage. For high-volume shooters who already own capable gear, that trade usually favors staying local.
The plain summary is that this tool is built for the specific creator whose libraries are too big and too sensitive to hand to the cloud. It is not trying to beat every hosted culler on every feature. It is trying to do the culling well in the one place cloud tools cannot follow, on your own drives. If your shoots are small and you are happy uploading them, a cloud tool may suit you fine. If they are not, batch video software that stays local is the point. You can compare the approach on the creators and agencies industry page and see which side of that line your workflow sits on. One more practical note is worth adding. Because the tool owns none of your storage, there is nothing to migrate off later and no library held behind a subscription you have to keep paying. Your files were always yours and always local, so if you ever set the tool aside, your footage stays exactly where it has been the whole time. That independence is easy to overlook until a cloud service raises its price or changes its terms.
Does the culler help a self-employed creator keep records of their work?
Most photographers and video creators run as independent businesses, and that carries recordkeeping that has nothing to do with creativity. As a self-employed creator you owe self-employment tax and income tax, and you are expected to keep records that support what you report, a set of obligations the IRS lays out in its self-employed individuals tax center. Keeping your shoots organized is a quiet part of running that business, and it is easier when your library is culled and tagged rather than left as an undifferentiated dump of files.
Batch video software helps here indirectly but in a way that matters. When the culler tags and organizes a shoot into keepers and the rest, you end up with a library you can actually navigate by project. That makes it far simpler to see which shoot went with which client engagement, which supports the income and the business records a self-employed creator has to maintain. A tagged, organized library is a byproduct of culling that pays off well beyond the edit.
Be clear about what this is and is not. The culler does not track your income, calculate your tax, or file anything, and it does not pretend to. What it does is turn a chaotic pile of footage into an organized set tied to your projects, so that when you or your preparer need to reconstruct what work you did and delivered, the library is not the obstacle. The IRS guidance on self-employment is the standard you are keeping records against, and organized footage is one small input to that, not the whole of it.
The one-time purchase model fits the self-employed reality too. A working creator does not want another monthly subscription eating into thin margins, and the tool is priced to be bought once and owned. For someone counting business expenses carefully, a single predictable cost is easier to justify than an open-ended recurring bill, and it is one less subscription to reconcile at year end.
The realistic takeaway is that organization compounds. Every shoot you cull and tag now is a shoot you can find and account for later, whether you are pulling a portfolio, answering a client, or handing records to a preparer. Batch video software will not do your books, but by keeping your libraries tidy it removes one of the frictions that makes recordkeeping feel impossible for busy creators. You can see how the culler organizes large local libraries on the Batch Video/Photo Culler product page, and think about how much easier a tagged, project-based library would make the business side of your work. It is worth saying that good habits still matter more than any single tool. Organizing your footage does not replace tracking income and expenses, and the IRS guidance is clear that the records behind what you report are your responsibility. What the culler removes is the excuse that your library is too chaotic to work with at all. Start from a tagged, project-based set and the rest of the recordkeeping gets noticeably less painful, which for a busy creator is often the difference between doing it and putting it off.
How does the tool tag keepers, and does it keep my originals private?
The tagging is the working core of the tool. Pointed at a local folder, the AI runs through the whole library and marks the frames and clips that look worth keeping, so your review begins from a short list of likely selects instead of the full mountain. You then look at the tagged results, adjust what the AI got wrong, and export the keepers. The human still makes the final call. The culler just does the brutal first pass of narrowing thousands of options down to a handful worth your attention.
Because every original you shoot is a copyrighted work from the moment it is fixed, as the U.S. Copyright Office notes in its overview, What is Copyright, keeping those originals under your own control matters. The tool tags and sorts your library in place, on your machine, without shipping the originals anywhere. Your raw files, the ones that carry your rights, never leave your drives during the cull. Batch video software that works this way means the sorting never becomes a reason to hand your originals to a third party.
Privacy of the originals is a direct result of that design. The AI reads your footage locally to tag it, and no cloud copy is created in the process. For shoots that involve clients or private sessions, that is the difference between a tool that respects the confidentiality you owe your subjects and one that quietly duplicates their images onto someone else’s servers. The originals stay yours, in every sense, from capture through export.
There is a fair limit to state. The tagging is a strong first pass, not a substitute for your eye. The AI surfaces likely keepers, but you decide what actually makes the cut, and it will occasionally flag something you would reject or miss something you love. That is expected, and it is why the review step exists. The value is in how much it narrows the field, not in pretending to have taste. Treating it as a fast sorter rather than a final judge is how creators get the most from it.
Put together, the capability is simple to describe and hard to live without once you have it. A local AI pass that tags keepers across a huge library, keeps your originals private on your own machine, and hands you a short list to finish by hand. For a high-volume shooter, that is hours back after every session with no cloud upload and no monthly bill. Batch video software built around a local tagging pass is aimed at exactly that creator. You can read how the tagging and export flow works on the Batch Video/Photo Culler product page. The best way to think about it is as a tireless first reader for your footage. It never gets bored halfway through a ten-thousand-frame library, and it never rushes the end of a long sort the way a tired human does. You bring the judgment, it brings the stamina, and between the two of you a shoot that used to swallow an evening becomes a short, focused review. That balance is what keeps creators coming back to it after every session.
| Attributes | Value |
|---|---|
| Deployment |
On-Premises |
| Platform |
Windows |
| License Type |
Perpetual |
| Billing Period |
One-Time |
| Target Customer |
Creators & Agencies |
| Free Trial |
No |
Frequently bought together
-
AI Grant / Proposal Writer (local RAG)
$39.00 -
AI QA / Citation Layer for AI outputs
$39.00 -
AI Receptionist That Acts
$95.00 -
AI Support Inbox for Microteams
$39.00 -
Batch Image / Brand Asset Generator
$65.00 -
Bulk Document AI Processor / Batch PDF Splitter+Binder
$249.00 -
Client Portal + AI Concierge / Micro Portal Builder
$39.00 -
Construction Takeoff Lite / Drawing Annotator
$65.00







