diff --git a/TODO b/TODO
index 62ade08..d148c7c 100644
--- a/TODO
+++ b/TODO
@@ -1,17 +1,13 @@
 ## GENERAL
-    * Face matching:
-        - upgrade to face distance per face per file [DONE]
-        - face locations:
-            START FORM SCRATCH for prod so all images have face_locn data
-        - need to reconsider whether current distance algorithm gives best match - can I do better?
+    * face locations --  START FORM SCRATCH for prod so all images have face_locn data
 
     * per file you could select an unknown face and add it as a ref img to an existing person, or make a new person and attach?
     * from menu, we could try to get smart/fancy... say find face with largest size, check it vs. other faces, if it matches more than say 10? we offer it up as a required ref img, then cut that face (with margin) out and use it is a new ref image / person
         - read that guys face matching / clustering / nearest neighbour examples, for a whole new AI capability
+            https://www.pyimagesearch.com/2018/07/09/face-clustering-with-python/
 
     * fix up logging in general
     * comment your code
-        * js files
         * html files?
 ## DB	
     * Dir can have date in the DB, so we can do Oldest/Newest dirs in Folder view
diff --git a/pa_job_manager.py b/pa_job_manager.py
index c629fe4..1cba183 100644
--- a/pa_job_manager.py
+++ b/pa_job_manager.py
@@ -1679,20 +1679,57 @@ def MatchRefimgToFace( refimg_id, face_id, face_dist ):
     return
 
 ####################################################################################################################################
-# BestFaceMatch(): take in required threshold for anything to be a match, then go through the 'dist' array - it contains the face
-# distance for each refimg / face.   So this func loops through all of them,
-# return the best match (with lowest distance) by checking each one that is a match based on the threshold
+# Util function to remove the matching face number (which_f) from the dist array (which has dist[who][<face_num>])...
 ####################################################################################################################################
-def BestFaceMatch(dist, fid, threshold):
-    # 1 is not a match (0 is perfect match)
-    lowest=1.0
-    which=None
+def RemoveFaceNumFromDist(dist, which_f ):
     for who in dist:
-        if who in dist and fid in dist[who] and dist[who][fid][0] < lowest and dist[who][fid][0] <= threshold:
-            lowest=dist[who][fid][0]
-            which=who
-    print( f"bfm: return {which}, {lowest} for {fid}" )
-    return which, lowest
+        if which_f in dist[who]:
+            del( dist[who][which_f] )
+            next
+    return
+
+####################################################################################################################################
+# go through dist array and find the best single match (lowest face distance)
+# returning the matching refimg (which_r), and face num (which_f), and dist (which_fd)
+####################################################################################################################################
+def FindBestFaceMatch( dist, threshold ):
+    which_r=None
+    which_f=None
+    which_fd=None
+    lowest=1.0
+    for who in dist:
+        for fid in dist[who]:
+            if dist[who][fid][0] < lowest and dist[who][fid][0] <= threshold:
+                which_r=who
+                which_f=fid
+                which_fd=dist[who][fid][0]
+    return which_r, which_f, which_fd
+            
+####################################################################################################################################
+# Okay, go through dist array and find the best single match via FindBestFaceMatch(),
+# then record that match and remove that refimg/person and that face number
+# from the dist array, and rinse/repeat until we have no more faces to match,
+# or the 'best' match is > threshold, so no more matches...
+####################################################################################################################################
+def ProcessFaceMatches( job, dist, threshold, e, name ):
+    while True:
+        print( f"ProcessFaceMatches() - finding best match left with dist={dist}" )
+        which_r, which_f, which_fd = FindBestFaceMatch( dist, threshold )
+        print( f"seems that best match is r={which_r}, f={which_f}, with fd={which_fd}" )
+        if which_r != None:
+            print( f"okay, which_r is real, so we have a match" )
+            MatchRefimgToFace( which_r, which_f, which_fd )
+            AddLogForJob(job, f'WE MATCHED: {name[which_r]} with file: {e.name} - face distance of {which_fd}')
+            # remove this refimg completely, cant be 2 of this person matched
+            print( f"now remove this refimg from dist" )
+            del( dist[which_r] )
+            # remove this face id completely, this face cant be matched by someone else
+            print( f"now remove this face from dist (if it is connected with anyone else)" )
+            print( f"dist now = {dist}" )
+            RemoveFaceNumFromDist( dist, which_f )
+        else:
+            return
+
 
 ####################################################################################################################################
 # ScanFileForPerson(): for a file, check to see if a person is matched via face_recognition
@@ -1700,13 +1737,8 @@ def BestFaceMatch(dist, fid, threshold):
 # NOTE: can pass force into this, but no f/e to trip this yet
 # if we do not have (any) faces for this file, go get them and their locations and store them in the DB assocaited with this file
 # then for each face (known/matched already or not), create a new array 'dist[refimg][face]' and run face_recognition code
-# to calculate the distance between the refimg and this face, for each refimg and each face
-# then invoke BestFaceMatch above to find the best match, and then store that in the DB for each face for this file
-# TODO: I *think* by going face-1, face-2, etc. we could find a suboptimal solution:
-#       e.g face-1 (Cam 0.55, Mich 0.6)
-#           face-2 (Cam 0.45, Mich 0.54)
-#       Algo would pick Cam for face-1, delete Cam, then Mich for face-2.  
-#       Should have chosen the other way -> might need to research best algo here
+# to calculate the distance between the refimg and this face, for each refimg and each face, then call ProcessFaceMatches() to 
+# go through the best matches (1-by-1) until no more faces match and store matching faces in DB
 ####################################################################################################################################
 def ScanFileForPerson( job, e, force=False ):
     # get default_scan_model from settings (test this)
@@ -1751,15 +1783,8 @@ def ScanFileForPerson( job, e, force=False ):
             refimg_face_data = numpy.frombuffer(r.face, dtype=numpy.float64)
             dist[r.id][face.id] = face_recognition.face_distance(unknown_face_data, [refimg_face_data])
 
-    # if you need to check face distances, uncomment this: print( f"dist={dist}" )
-    # TODO: I think this next line is not needed anymore (does the same as above?)
-    faces = session.execute( f"select f.* from face f join face_file_link ffl on f.id = ffl.face_id where ffl.file_eid = {e.id}" )
-    for face in faces:
-        who, fd = BestFaceMatch(dist, face.id, threshold )
-        if who != None:
-            MatchRefimgToFace( who, face.id, fd )
-            AddLogForJob(job, f'WE MATCHED: {name[who]} with file: {e.name} - face distance of {fd}')
-            del( dist[who] )
+    # record matches in DB...
+    ProcessFaceMatches( job, dist, threshold, e, name )
     return