Should photos and video be in the same folder?

In the same project folder, yes, in separate subfolders. They are edited in different applications with different tools, so mixing them in one directory makes both harder to work with, while separating them into different projects breaks the connection to the shoot they came from.

Why do photo and video organizing tools not work together?

They evolved separately. Photo tools built around catalogs, RAW development and ratings; video tools around timelines, proxies and timecode. Metadata standards differ too, so photo tools rely on EXIF and IPTC embedded in the image, which has no equivalent that all video formats support.

How do I keep stills and clips of the same subject connected?

Shared vocabulary. The same subject tags and the same date-and-project folder naming across both, so a search for a tag returns both the stills and the clips. On macOS, Finder tags work identically on a JPEG and an MP4, which makes them the natural common layer.

Do HEIC photos from an iPhone cause problems?

Not for organizing. HEIC is widely readable now, and it keeps EXIF like any other format. Some older tools and some delivery targets still want JPEG, so keep the HEIC as your original and export JPEG copies when a client needs them, rather than converting the originals.

How to organize photos and video from the same shoot

One folder structure for stills, clips and audio, why photo-first and video-first tools fight each other, and how to keep a shoot together across all three.

7 August 2026 6 min read

Look at what actually comes off a real shoot. A card of video from the main camera. A few hundred stills, some from a proper camera and some from a phone. An audio recording of the interview, and another twelve minutes of ambience. Maybe drone footage.

Now look at how most people organize it: the video gets a careful folder structure and a naming convention, the stills get dumped into a folder called Photos, and the audio ends up somewhere nobody remembers.

Six months later, the video is findable and the rest is gone. Not deleted, just unfindable, which amounts to the same thing.

Why mixed media falls apart

Three structural reasons, worth naming because they explain why willpower alone does not fix this.

The tools are tribal. Photo software grew up around catalogs, RAW development, ratings and non-destructive edits. Video software grew up around timelines, proxies, codecs and timecode. Each treats the other’s media as a second-class attachment. Neither wants to be the system of record for both.

The metadata standards diverged. Stills have EXIF and IPTC, embedded in the file, read by everything, an actual working standard for a few decades. Video has no equivalent that every format supports and every tool respects. So the photo world assumes metadata travels inside the file and the video world assumes it does not.

The volumes are lopsided. 800 stills and 40 clips is a normal ratio, and it means any browsing view that mixes them is dominated by photos. So people separate them, and then never look at one of the two piles again.

One structure, three lanes

The fix is a project folder that treats all three media types as first-class, with parallel lanes rather than one merged pile:

2026-08-14_ClientName_Project/
  01_Video/
    A7SIII/
    Drone/
  02_Photos/
    Camera/
    Phone/
  03_Audio/
    Interview/
    Ambience/
  04_Transcripts/
  05_Project/
  06_Exports/

The principles:

The shoot is the top level. Not the media type. You remember jobs, not file formats.
Media type is the second level. Because that is where the tools diverge.
Source is the third level. Which camera, which phone, which recorder.
Exports stay separate from originals. Always, in every project, without exception.

This survives being handed to another person, which is the real test. Someone who has never seen your system can open this folder and understand it in five seconds.

The common layer: tags and names

Structure separates. What ties the three lanes back together is vocabulary.

Use the same subject tags across all three media types. If a video clip is tagged harbour and golden-hour, the stills from the same hour get the same tags, and so does the ambience recording. Then a search for harbour returns all of it, which is exactly what you want when you are building anything.

On macOS this is easy, because Finder tags work identically on a JPEG, an MP4 and a WAV. That is the one place where the file system is more capable than the specialist tools: it does not care what the file is. A tag vocabulary that works for B-roll works unchanged for stills.

Use the same date-first naming. 2026-08-14_Harbour-Sunrise_001.jpg and 2026-08-14_Harbour-Sunrise-Wide.mp4 obviously belong together, in any list, in any application, forever.

Where photos need their own handling

Three things stills need that clips do not.

RAW and JPEG pairs. If you shoot RAW plus JPEG, keep them together, same base name, and decide once whether the JPEGs are throwaway previews or deliverables. Half-kept JPEG copies are a recurring source of “which of these is the real one”.

EXIF is genuinely useful, use it. Camera, lens, focal length, aperture, time, and often GPS. It is embedded, it is standard and it survives copying. Do not strip it in a batch export unless you mean to, and do think about whether GPS should stay in files you send to a client.

Volume triage happens earlier. With 800 stills, culling is the first pass, not an afterthought. A selects subfolder or a select tag on the first pass is the difference between a usable set and an unopened folder.

Where audio needs its own handling

Audio has the opposite problem: too few files, all invisible. No thumbnail, no preview, nothing to look at in a grid.

Two things fix most of it:

Transcribe anything with speech, save the transcript as a text file next to the recording, and it becomes searchable by content.
Describe anything without speech. Type of sound, environment, whether it loops. Twenty words in a sidecar file.

Detail in organizing audio recordings, including the specific case of a sound library that should still be usable in ten years.

The one pass that covers all three

Here is the argument for handling mixed media in one operation rather than three.

Every one of the three lanes needs the same thing: a description of what is in the file, written somewhere the operating system can search. A photo needs it as much as a clip does, because IMG_4821.HEIC is exactly as opaque as C0087.MP4. A recording needs it most of all.

Historically you needed three different tools, because describing a photo, describing a video and transcribing audio were three different technical problems. They are converging: the same kind of model that describes a video frame describes a photograph, and speech recognition handles the audio track of a clip and a standalone recording identically.

Which means the sensible shape of an ingest pass is: point one process at the whole project folder, let it describe everything, name everything, tag everything and transcribe what has speech, then file each media type into its lane. One pass, three lanes, one vocabulary.

That is the practical version of the argument in AI video asset management, applied to the mess a real shoot actually produces.

A checklist for the next shoot

Create the project folder before you copy anything, with all six subfolders.
Copy every card and every recorder to its lane. Verify. Do not format anything yet.
Run one description pass over the whole project folder.
Cull the stills. Ten minutes, ruthless.
Apply the shared subject tags across all three lanes.
Read the transcripts before you open a timeline.
Back up the whole project folder, then format cards.

Steps 3 and 5 are the ones that used to be a full evening. That is the part worth automating, and it is why doing it for all three media types at once matters more than doing it well for video alone.

Cliptag handles video, photos and audio in the same drop on macOS. Photos get described, named and tagged, including people you can name yourself. Recordings are classified by type and transcribed when there is speech. Video gets the full pass. Everything is filed by media type and date, with Finder tags across all three. On-device mode on Apple Silicon is free and unlimited, and on the cloud plans four photos count as a single analysis.