* Switching census record filing from Head-based to Sheet-based

[Gary_G]

I'm at a decision-point in redoing my method 1 data-entry. Filing census data by the head-of-household is causing a few problems and I can see more on the horizon. I'd like to switch to filing the census returns and similar "group-events" under a census directory. However; I am not sure how to keep tract of the person(s) to whom the census sheet or sheets pertain while I am doing so. After I've switched, I'll have to rely on FH7 to keep track of this.

I currently have multiple copies of the census; one for each Head-of-house or for each person (as needed). The associated filenames show the target Head-of-house or person. I'd use the A.S. image rename-during-import feature, but I also have research files and images associated with the main image. I need to rename those to maintain the association and move them to the same target directory as the census image, so they don't "get lost".

Has anyone tackled this issue before and found a workflow that works?

[Mark1834]

Not sure I can help with transitioning, but my census source is the household entry. I identify it by location (place/address) and a suitable reference, so a typical England source would be something like “1861 England Census (High Street, Anytown, Anycounty), RG9/1234 folio 56 page 78 (#9)”.

I make no specific reference to any individual in the census apart from in the transcript. In particular, I find “Head of Household” adds little if any value for me, as more often than not, it’s not the blood relative anyway.

My Media record is the census page, so multiple sources can link to the same media. I originally had two separate media records for an entry spanning two pages, but I now create a new composite image of both pages, as I find that more useful for presentation.

[Gary_G]

Mark;

Thanks for the input.

I guess the big issue for me is more that all of the naming systems work, once one has the data in FH7 and fully cited. However; I'm currently worried about losing track of the purpose of various images while I'm in the actual process of changing filenames. Once everything is linked up, I can find records by name, by address or by census reference.

Because A.S. renaming can only affect items that are actually added as images in the FH7 media library, I'm starting to think I'll pretty-much have to rename a set of files by hand and then immediately manually enter the key images into FH7 or I'll risk losing track of some files. I've toyed with using some form of research notes feature to keep track of the images that aren't actually imported, but that concept is still in its infancy.

[tatewise]

In my experience when two key households appear on one census page the key people have the same surname.
So I have the Census page file named as:
Media\GROCensusRecords\1881 RG11 1226 064 08 OLDEN, James & William.jpg
The Media Record Title is the same as the filename.
That Media Record is linked to two Census Source records; one for James and another for William.

I don't focus on the Head of Household but use the key person in the household who is my relative.
They may be the Head but also might be a Lodger or an inmate in an institution.

[ColeValleyGirl]

In my experience when two key households appear on one census page the key people have the same surname.
So I have the Census page file named as:
Media\GROCensusRecords\1881 RG11 1226 064 08 OLDEN, James & William.jpg
The Media Record Title is the same as the filename.
That Media Record is linked to two Census Source records; one for James and another for William.

I don't focus on the Head of Household but use the key person in the household who is my relative.
They may be the Head but also might be a Lodger or an inmate in an institution.

Not my experience, Mike. And even when they have the same surname, they're not always related (the joys of Wales).

[Gary_G]

Helen and Mike;

I've had the much the same experience as Helen, but with Scotland.

I just realized it may help if I posted what I currently have as a filesystem, so that one can see why I need a system that can keep track of more than one page per census, regardless of whether there are multiple people-of-interest on a single image. Keeping the associated census images has saved me more than once, when I need auxiliary info. The 1911 census just happens to be one in which two sisters are on the sae page, but that is by no means the rule.

Example of types of files stored for one census

Screenshot 2024-03-31 at 9.12.20 AM.png (113.84 KiB) Viewed 612 times

[Gary_G]

I wonder...

Can the FH7 Project subdirectories contain a hierarchal structure for files?

If so; I could set up a Census folder that is subdivided into directories uniquely named for the census year and region. In those, I could have a directory for my page-images for that census. That way I could keep related information grouped.

I could then include a text-file "manifest" that identifies the relevance of the files, so I can refer to it. If I import the manifest and the relevant image(s) from the directory, I should be able to call up what I need to complete the related FH7 census entry.

I suppose it's worth a try on a test set...

[ColeValleyGirl]

I wonder...

Can the FH7 Project subdirectories contain a hierarchal structure for files?

Yes, you can have any structure you want under the Media sub-folder.

[tatewise]

Yes, as illustrated by my posted Media file path & name:
Media\GROCensusRecords\1881 RG11 1226 064 08 OLDEN, James & William.jpg

You can have any valid Windows folder structure as long as it does not break the 255 character length for the entire path from drive letter to file extension, i.e. C:\ to .jpg

[Jean001]

No help with the process, but...

I store media by 'type': Births, Deaths, Marriages, Baptisms, and Burials in one folder/directory, Census returns in another, plus a variety of other folders as required. I don't have many overseas (non-UK) records, if I did I would create a structure starting with country.

Filenames start with the name of the key person, followed by date, type and reference.
The Media Record Title matches the filename.

After 'processing' the Source Record I add the Media Record ID and the Source Record ID to the Media Record Title, using a Plugin to update the original filename. (Years ago before FH, I had a problem with lost links and had a long job matching every source with its media.)

I keep my image files in folders not in the FH Project folder (other software is also linked to the files).

[Gary_G]

Yes, as illustrated by my posted Media file path & name:
Media\GROCensusRecords\1881 RG11 1226 064 08 OLDEN, James & William.jpg

You can have any valid Windows folder structure as long as it does not break the 255 character length for the entire path from drive letter to file extension, i.e. C:\ to .jpg

Mike;
The limitation is even worse for a virtual or network filesystem. Windows sees a path length that is longer than the file+path length relative to the virtual drive C root. I had this issue with RM9 for years.

It appears that, on a virtual machine, one gets fairly close to the maximum file+path length limit without even trying. this is largely due to the path "overhead" of Windows plus that of the Mac on a virtual machine. I did some calculations and while both configurations will work, the safety margin wrt. file+path length is far better if files are linked in from the Mac file system. This is because one avoids any overhead associated with where FH7, by default, wants to place files in the virtual windows filesystem.

[Gary_G]

I really would like to get away from putting the name of a person in a census record title. This is because there may be several from one or more families present on the same image. So; I'm still playing with a test project to see if I can capture the key people that were the reason for having downloaded the image(s).

Because of FH7 importing certain Windows-only metadata from an image (eg. comment), a Mac user can't do much more than set the title of the file prior to import and expect to see it in the results. This means that my best hope is to import my data and then go into the media record and add a note to document the person or persons of interest. Once that is done, I can do a text search to find them during data-entry. It's awkward, but it should work.

My only other option is to go into the Windows side of Parallels and play Windows user to add a comment. I'm not yet sure which will work better.

[KFN]

Long time ago I stopped naming my images based on the contents of the image. In my database all images have a randomized unique name.

I let the database object record “title” and “note” tags list the content of the record. In cases where the content is not obvious by looking at the image (census pages have readable names, vs images of people do not), I use descriptions and/or tags inside the .jpg image to list the content.

In census cases where 2 or more households reside at the same location, the title of the object record can indicate the location, and the object record note tag can outline the members of each household.

In my primary software I can use markup to create a grid inside the NOTE to better organize the various households and names!

[t4ms]

I'm not sure I'm adding anything to this but here's some examples of how I name my census images (I work with folios not people):

RG 12_166, Folio 107; 1891 England Census, Islington, London, England
This is the same as the title, exchanging / for _:
{Dept_Series_Piece_Book}, Folio {Folio}; {Census_ID}, {Parish}
I use this for 1841-1901

RG 14_9721 Sch 95; 1911 Census, Walthamstow, Essex, England
{Dept_Series} Sch {Schedule_number}; {Census_ID}, {Parish}

RG 101_1062A, Sch 42, 1939 Register, Chigwell, Essex, England, 19 Avondale Drive
{Dept_Series_Piece}, Sch {Schedule_number}, 1939 Register, {Place}, {Street_address}

Tamara

[Jean001]

Gary,

If I understand correctly, your images are not yet included in your FH project.

I appreciate that you are going to drop the name from the start of your filenames, and perhaps rewrite the filenames.

In the first instance would it help to copy the original filename to the Media Record Note? I did that in the early stages of using FH as my existing filenames contained information that helped me identify the contents.

I started off by adding my existing media files one by one and making the appropriate changes and links but soon decided to bring all the available image/media files into FH. I could see everything in the one place and use FH's tools to make changes.

I have added many images/media files to FH that weren't linked to Individuals or Source Records until later. For a specific batch I would create a Source Record and link all the relevant Media Records to that, adding a pertinent note to the Source Record. In that manner, I could easily pick out the desired batch from the many Media Records available.

At times I've changed the folder structure concerning my image/media files, pointed FH in the right direction and it relinked all the Media Records.

One example of my updating work from within FH:
Moving to FH caused me to overhaul my image/media filenames. Many of the files were not linked to my previous genealogy software so in FH I added them as Media Records (no links to Source Records at that stage) and 'rewrote' the Source Record Title. As I worked on this process over time I also added a 'marker' to the end of each Media Source Record to mark those 'completed'. Then using the Plugin (Rename Media to Match Title Field), I updated (as a batch process) the image filenames to match the Media Record Title. So, I could tell at a glance whether a filename had been updated or not. When the work was done I used Mike Tate's Search and Replace Plugin on the Source Media Titles to remove the marker, ran the Rename Media Plugin once again and the job was finished. (This was in my early days of using FH so had no experience with Named Lists, which I would now use.)

If only I had the skill to write my own Plugins...

I'm not saying for you to copy what I have done but to illustrate some ways that can help if working from 'within' FH.

[Gary_G]

Tamara;
As I'm trying to drop the names in my media titles, it's not actually a question of how I name my files, but more of getting enough info into the media record that I know the person to whom the file pertains.

[Gary_G]

Jean;

You're correct. I have TONS of files that I want to get into my media directory, but am trying to figure out how to do so and not lose track of the person(s) to whom they pertain. My old naming convention had the person's name in the filename and the target convention is just based on the name of the image.

I thought of the same idea of embedding some key info in the media window notes. However; as I'm working on a Mac, putting in the Title, Comment, Date and Keywords, is a BIG challenge. This is because FH7 depends on the Windows custom tags; XPTitle, XPComment and XPKeywords to be able to automatically pic up the key info during import. Thankfully it does use one standard tag; Create Date. I have used a tool called EXIFTool to inject such tags into files by using a spreadsheet as an input source for the keywords and values for each file. But that is only one way. I can also just import the files and use the media screen to manually add the same data.

I'm not anxious to dump everything I have for each person into the FH7 media window, because too many files would clutter the media window and I wouldn't be able to find the ones on which I need to work. However; I do plan to put all my background files in the project media directory and link the essential ones from there.

I think if I get the Persons-of-interest assigned to the Notes for the imported key images and the Title is set up to just document the image itself, then I should be able to figure things out when I come to doing the citation.

I will be going over your post to see what I can adapt to my needs. Thank you for showing me your workflow.

[Jean001]

Gary,

I do understand because I too had many images with embedded notes which FH could not read...

I have thousands of Media Records but rarely use the Media Window.

I select Media Records from the Records Window or via a Source Record.

I should remind you that I make all entries 'manually' so your system of working will be different.
I do not keep my images/scanned documents in the FH project folder.

Without being aware of the detail of your system, I might work something along the lines of:

Create an Individual Record.
From somewhere on the Individual Property Box (maybe Note, Custom ID, or a Custom Fact), create a Source Record.
In the Media Tab of the Source Record, add the required Media (multiple files can be added as a batch).

Repeat as required.

The Source Record will keep the connection between a person and his/her media whilst you are making your revisions. When all is done you can delete the Source Record.

EDIT
One of these temporary Source Records per person (perhaps, depending on your existing system), but more than one Source Record can be linked to a Media Record if needed.

[Gary_G]

Jean;

Thanks. I appreciate the suggestions. I plan to try refining a workflow on a test database, before trying it on my "live" data.

[t4ms]

Tamara;
As I'm trying to drop the names in my media titles, it's not actually a question of how I name my files, but more of getting enough info into the media record that I know the person to whom the file pertains.

Ah. I'm getting lost in all the information. Have you thought about a note for the media and linking all individuals to it? Apologies if this has already been suggested.

[Gary_G]

Tamara;

I have to import the images from an externally linked directory to an "in-project" one, rename the files to keep the path-length+filename length to a minimum, as well as keep track of who is contained in each image. I decided to add an image note listing each person-of-interest for each image transferred to the "in-project" directory. I can then filter on those when doing my citations for an individual. If I find more people in a given image, I'll update the image note.

So far, the process seems to work...

[Little.auk]

In my experience when two key households appear on one census page the key people have the same surname.
So I have the Census page file named as:
Media\GROCensusRecords\1881 RG11 1226 064 08 OLDEN, James & William.jpg
The Media Record Title is the same as the filename.
That Media Record is linked to two Census Source records; one for James and another for William.

I don't focus on the Head of Household but use the key person in the household who is my relative.
They may be the Head but also might be a Lodger or an inmate in an institution.

I use a similar approach, and format to Mike - which follows the hierarchical method of drilling down in detail. For Example : Media\Census Records\1871\Census 1871 (IM) RG 10 4374 64 14 84 ROLLIN William [F0008-00].

I use (IM), and other two letter "flags", to identify the type of media - in this case an image of the census page. The name appears last, purely as a reminder of who is referred to and the square brackets contain the Family Custom ID --- William Rollin is a very common name in my tree, so the Family ID (for head of household), or the Individual ID if not, are useful reminders which one it is!

I have one or two instances where two "Families of Interest" appear on the same census page including one with different surnames - in that case I would enter, for example, "ROLLIN William [F0008-00] & CRAWSHAW Sam [F0023-01]"

If my image has come from a website like FMP, I also save an image of their transcript (TR). I think this is important as it is this transcript data that is indexed on the site. My main research surname is ROLLIN, but I have found census mis-transcriptions as bad as BOLLIN.

FMP are very good at reviewing and correcting errors reported to them, usually within 72 hours. but other sites are not always so timely. So, having a record of the search criteria that actually found the record is important.

[Gary_G]

"Little_auk";

Thanks for the suggestions. I'm currently copying my existing filenames into a structure that hopefully reflects their final names. This will highlight images that are duplicates and give me a chance to consolidate them and eliminate the duplication. I'll do further renaming and organization prior to entering the data.

You are correct about FindMyPast. I've found them to offer excellent service and a far more logical archival structure than their competitors. Consequently; they are my main source of English and Welsh data. I have found ScotlandsPeople to also have excellent service and to also be well organized. My only suggestion for them is to do as Genealogy Quebec does and add a standardized name to their index. This helps in situations in which names have variants.

[Gary_G]

Within my project media directory, I've created some of the typically suggested subdirectories. However; as I add images to them, I'm seeing that the way they categorize things may not be all that workable for me. My goal is to file in a way that avoids duplicating images in my overall media directory. This seems to mean that I must avoid filing images by individual, but instead after their source. If anyone is using such a system, please read on...

For instance; a single parish register can contain births, marriages and deaths. For these, it appears that I may be better off to file images based on the church/parish and document from which they came. I suspect for imaged books and rigidly structured documents like census returns, this type of filing may be the way to go.

Unfortunately; the forgoing logic doesn't work for images gleaned from some online collections or compendiums of loose records, which can be particularly poor in identifying the exact source of their images. They tend to aggregate things and say something like "Kent Baptisms". So; I'm not sure what to do with these. Perhaps the name of the collection and then a group within that collection would work best.

Artifacts that one has on file are another issue entirely. They often don't fall into any obvious category. So; I'm thinking that they should likely be filed by the individual from whom they were obtained (not to whom they pertain). At least that way one might be able to infer the correct directory in which to look.

While I hope to file images in categories, I will likely still default to using just split sources. This is because I haven't yet found a robust rule to define when it is strategically better to use lumped ones. So far; it seems that; when a record transcription references more than one ancestor, using split sources would avoid issues resulting from editing the transcription.

If you're using a similar system, I'd really like to hear about how you've tackled collection and artifact filing in your FH7 media directory. I'd also like to hear about for which of those categories you use lump or split sources.

[KFN]

This may be too far out there to consider.

The Windows hierarchical folder system is not real good at supporting robust multi-dimensional data organization! However, you can use shortcuts to better support some organizational concepts.

For example: Suppose you would like to have two folder organizational structures like,
1) Family -> Individual (where the individual folder contains all images from sources and photos.)
2) Country -> Source Type (where the source type folder contains images from that source)

You could load all images into a single folder (aka a repository) called Genealogy Images. Then create a "Shortcut" in the appropriate organizational structure that points to the image in the repository directory.

Genealogy Images {image RJ1234.jpg} (contains the actual image of a birth certificate or church record)
Johnson Family->Ralph 1875-1821 {Shortcut to Genealogy Images for image RJ1234.jpg}
UK->Birth Records {Shortcut to Genealogy Images for image RJ1234.jpg}

A simple program could be written to move files to the "Genealogy Images" fold and create the appropriate shortcuts in the "organizational structures"!