Help · Getting Started
Your first named gallery.
Three things to learn and you're delivering named files: index your headshots, pick the roster MatchID runs against, and review the matches. Here's how each one works.
Build the local face index MatchID compares against. No index, no matches.
Jump to indexing →The browser panel is where you select photos and the roster MatchID names them against.
Jump to the browser →Confirm the suggested name against your roster. Names land in IPTC and EXIF on the way out.
Jump to review →Before you start
Why headshots come first.
MatchID is face recognition. It works by comparing the faces in your photos against reference faces it already knows. Those reference faces come from the headshots you index. Skip this step and there's nothing to compare against, so nothing gets named.
Think of indexing as introducing the players to the app once. You point it at a folder of named headshots, it reads the strongest face out of each one, and it stores a small math fingerprint of that face on your Mac. That's the index. The original headshots and the fingerprints stay local. We can't see them.
You index a roster once. Reindex when athletes change, no quota. After that, any folder you process against that roster gets named automatically.
The one thing new users miss
A roster with zero indexed headshots can't match anyone.
Selecting a team in the Roster Library gives you names and jersey numbers for manual tagging. It does not give MatchID faces to recognize. For face recognition to work, that team needs headshots you've indexed. The app warns you when a selected roster has none.
Step 01 · Indexing
Index your headshots.
Open the Subjects page for a team, then click Index headshots… A drawer slides in from the right and walks you through four short steps.
- Category. Pick where the new subjects live, and set a
Base Person IDlikebvol. The app auto-numbers each one (bvol01,bvol02…). - Pick files. Drag in a folder of headshots, or browse to them. Each file becomes one subject. The strongest detected face is enrolled.
- Review. Check the auto-suggested names and IDs. Edit any row before indexing. A status column flags anything with no detectable face.
- Index. Click
Index N subjects. A progress bar runs while the app builds the index. Done.
Display names backfill from the file's XMP metadata when it's there, so a clean headshot folder mostly fills itself in.
Review the auto-suggested IDs and names. Edit any row before clicking Index.
| Preview | Display Name | Person ID | Status |
|---|---|---|---|
 | Camille Dubois | bvol01 | Created |
 | Grace Lin | bvol02 | Created |
 | Lucia Santiago | bvol03 | Created |
| team_huddle.jpg | bvol04 | No face |
The Index Headshots drawer, Review step. Names and IDs auto-filled, one flagged with no detectable face.
Step 02 · The browser panel
Pick the folder. Pick the roster.
The browser is three columns. Folders on the left, your photos in the middle, and the Process panel on the right. The Process panel is where MatchID gets told what to do.
#14
RAW#7
#9
RAW
#17
RAW#19
#23
RAW
#21
RAW
RAWThe Browse view. Folders left, photos center, the Process panel right with the roster MatchID will run against.
Selecting the roster is the step that matters most. In the Process panel's MatchID tree, check the team whose headshots you indexed. That tells MatchID which faces to compare against. The hint line confirms how many subjects are indexed, like 1 team · 12 subjects · 12 indexed.
Name the job, give it a Job ID, then click Process all N →. The app detects faces, matches them against your indexed roster, reads jersey numbers when a face isn't legible, and stamps the confident ones automatically.
Working a folder with subfolders? Turn on Cascade to subfolders so they inherit the roster you picked instead of you setting each one.
If you see this warning
Roster has 0 indexed headshots.
It means you picked a team MatchID has no faces for. Open Subjects → Index headshots… for that team first, then come back and process. Until then, face recognition has nothing to match against.
The Roster Library still helps here. It gives you the names and numbers to tag by hand. But automatic face matching needs the indexed headshots.
Step 03 · MatchID review
Confirm against the roster.
After processing, open a photo in review. The app shows the full image with a box around each detected face and the name it matched. The filter tabs up top group faces by status, so you can land on the ones that need a human look and skip what MatchID already nailed.
MatchID review. The detected face is boxed and matched to #14 Camille Dubois from the indexed VB 2025-26 roster on the right.
Hover the matched thumbnail to approve, reject, or remove the tag. Need to fix a miss? Pick the right athlete from the roster panel by name, Person ID, or jersey number. Because the match runs against the headshots you indexed, the suggestion is always one of your own athletes. Each confirmation also sharpens the local index for that player, so the next gallery matches a little better.
The tabs across the top, All, Review, Auto, and No Tags, let you spot-check the app's confident calls or focus only on faces that still need a name. Turn on Show jersey matches to surface names the app read from the jersey number when a face wasn't legible.
When the photo's clean, the names are written into IPTC and EXIF on export. Deliver by FTP or ZIP, and the names travel with the file into Photo Mechanic, Lightroom, or wherever it lands next.