I found quite a lot of topics concerning duplicate media but none seems to fit my problem. If I missed something please correct me.
When I create a GEDCOM file out of RM (10) with “Multimedia links” checked in the process and create an new database out of this file in MFT or FTM I find a lot of “duplicate” media in it. All of them are of course no real duplicates but one and the same media file is linked several times.
It is linked as many times as it has tags in RootsMagic. Why isn’t there only one linked media file in the new database also connected (aka tagged) to several facts and people as it is in RM itself? Is there an export option I’m missing?
Thank you!
GEDCOM transfers NO media. The importing program just receives the path(s) of media files in the .GED file. So, it’s what the receiving program does with those paths, right?
I know that GEDCOM files only transfer link informations for media and not the media files themselves. I assume RM is responsible for the LINK duplicates in the receiving databases because a GEDCOM export out of FTM with (linked) media files attached to several facts etc. does NOT produce multiple links when imported to another database. And the the number of duplicate links in the receiving database matches the number of tags for that media in RM.
Apparently neither is smart enough to know that a single path to a single media file should be one and the same singular file? Did you uncheck Extra details (RM specific) on export?
Does this mean that all media are exported via separate OBJE lines tied to the associated persons or events, and not as shared OBJE structrures that are exported at GEDCOM level 0?
Checkout my utility TestExternalFiles
It will check for duplicates in several ways. It won’t address any
GEDCOM issues however, but at least you’ll understand the status within RM.
Suggestions to upload tree to ancestry via RM are probably best. A RM gedcom to FTM does not seem to work well. I have tried sending gedcoms from FTM to RM to use some features and then back only to get a mangled database.
FTM sync to ancestry works well and never seems to fail like RM often does for many people.
when you compare file properties they differ significantly both up and down for many parameters. I did a test and made NO changes in RM before the Gedcom export.
For data transfer between RM and FTM this seems to be a good workaround, but this does not help with MacFamilyTree and others.
Anyway, thank you for that hint!
I did a small test here, and that suggests that the duplicates are caused by RM. I tried that with my Ancestry tree, which has only 6 media, of which 5 are census scans shared by multiple people. And in the GEDCOM, I can see 34 exported media files, with only 6 unique names.
This can be done better, because GEDCOM has a feature where persons, sources, and events, can reference shared media objects, which can then be exported once each.
I made a small test file with 5 people. Added a census fact with a source with an image and then a different image on the fact level, and copied it to everyone. There were only 2 media items on the Media page, and 1 source and citation on the Source page. I created a GEDCOM and imported into a new database. I still have only 2 media images on the Media page linked to citations and facts. I have 1 source and 5 citations. After merging the citations into one there is still only 2 media images.
The only way I could get multiple images in the GEDCOM is if the original file had them to begin with. If you have 5 separate people you were copying census facts with images in TreeShare it would add the media images each time. Ancestry doesn’t know you already added the image earlier so it adds a new one each time. The only time it doesn’t do this is if you downloaded the full tree again.
And this suggests that RM knows that there are 6, so there is a possibility to export 6. I see 34 exports now, because 1 item is attached to a person, and the other 5 are used by 33 citations, which I can’t merge. But these would also be impossible to merge, if each citation had its own line number in the scanned document, or some other unique feature.
I see duplicates, because the media exports are embedded in the citations, and not as separate objects in the GEDCOM file, which are then referenced by their object number.
For me, the issue with duplicated media still isn’t solved. Since I seem to have made the exact same experiences as @ennoborg I’m interested whether some solutions have been developed or discovered.
To make this perfectly clear: the duplicated media links in the afterwards imported RootsMagic-gedcom file are not originating from any “multiplied” TreeShare downloads. The respective files are just some scans (jpg) on my hard disk, linked ONCE to RM and then being tagged to multiple instances in the app → Only one image appears in the Media section of RM. The amount of duplicates appearing after import fits to the amount of links attached to them in RootsMagic. Furthermore this behavior does not only occur in one but two other programs (FTM and MFT) so I’d strongly suspect RM to be the bad guy.
In the Family Historian 7.05 update many moons ago:
When importing a GEDCOM file, several Media records could sometimes be generated unnecessarily for a single imported media image. This happened particularly with imports from RootsMagic. Fixed.
Looks like a RM construct that needed special handling.
It is indeed, and there is a simple solution, which is exporting shared media objects as separate objects at the GEDCOM level. That is a legitimate construct in GEDCOM 5.5.1, which we already use in Gramps.