Q: Does anyone know how to bulk enter information into the media Caption field?
I’m sure there’s probably a way with SQL, but I’m not skilled enough to do that myself. I checked @RichardOtter 's tools and that capability isn’t available.
Reason: In the past I wasn’t in the habit of adding a caption when adding media. Now I find that I’d like to sort media by Caption and the result, sans captions on most, isn’t helpful with those blanks.
Optimally I’d like to figure out how to scan through media, identify those that are blank, parse the filename of the item (not full path), and inject that filename into the Caption field. (My media naming and storage structure are rigorous enough that just the filename-as-caption is enough.)
I think it depends on what you want to see. My experience with RM tells me to always give these things need some careful thought first before making wholesale changes, by whatever means.
I like to name media files with the person’s name first followed by the event/fact, for example : Joe Bloggs - Census 1861. This means all Joe Bloggs media files are kept together alphabetically in my Windows media folder. But when it comes to captions it always seems more sensible to use the opposite : Census 1861 - Joe Bloggs. That way, on Joe Bloggs’s edit page, his media gallery shows a meaningful description for each media item.
It works but it is a bit long winded. Am I doing this in a sensible way, or am I just making a lot of work for myself?
Yes, I too would like to cut down on the typing, although I can’t really see how it could be easily done. I may be wrong of course.
@steve.turner What you’ve written is true, about being careful.
Actually, a bulk update utility would be most useful if there were some kind of switches available to determine what was updated.
In my case, my ultimate goal is to update all the newspaper clippings I’ve collected and have a caption that is of the form “Newspaper - {filename}”.
I’ve already updated the various other media, such as census (“Census - {filename}”).
Eventually I’ll have all the media meta-categorized so they all group together. Birds of a feather and all that. And because the filenames all use dates at or near the front, everything in a category will chronologically sort too. (Handy since right now the “Date” field isn’t sortable.)
I have done it (Sqlite) – I will PM you. I will send a couple later that you should be able to model after – make sure you have at least one backup before run – do you at least know how to execute a sqlite script before I send.
I was given the nickname “DATA JUNKIE” By one colleague and and at two former places someone referred to me as having and “easy button” but Wizardly is a new badge of honor lol.
There are three main things I’d really like to see in media management:
First off, there is metadata in pictures that could be used for captions, so you could use the wide variety of tools for media management to set the metadata in bulk. Then RM should use the Caption metadata to set the caption in the database.. Likewise, if you set the caption in RM it should set both the database and the metadata in the picture.
Next, there should be a “rename” facility that changes the names in the database AND in the file system.
Lastly, there should be a duplication check that consolidates duplicate media database entries (and links) into a single database record.
#1 - would be cool but I wonder if people would agree on what metadata to stuff into the caption field. Is metadata structured enough that the info to extract could be well described? It well could be, I don’t mess with this very much.
#2 - This can be done but it’s multi-step and a bit awkward… click on the document icon in the Edit Media filename field and then right click on the file to change the name in both OS and RM.
If you mean something like the Windows File Properties, then I wholeheartedly agree. I already use file properties to help me organize my genealogy files and discussed it here:
But @kevinm has a good point about getting people to agree on what file property field(s) to use.
I use ADD EXISTING or Drag N Drop method – in theory you should NEVER create duplicate media. I do not recall the last time other than when I moved media to different folders and used the fix media tool – which caused issues.
If you use ADD MEDIA that can / will add duplicate media – there have been questions if this was intention design to allow for multiple captions and descriptions (unknown if it was intentional “feature”
as for RM making use of the meta data – that would be a nice enhancement.
You can not do that via RM – the main file properties are available through excel query (not sure about the metadata) so that might be option use Sqlite or Python
The issue is when using TreeShare is you have 5 people on a census you’re going to get 5 copies of the census page with various nnnnnn (nnnnnn) names to keep them unique. If you fix them so they all point to the same image you have 5 database entries with different links that point the same image. I want one database entry with 5 links.
Likewise, you’re not going to get metadata via Excel, you need to have a tool that looks at it. There are defined fields, one is Caption. What I’m talking about isn’t rocket science, if the Caption exists in metadata it’s pulled into the database either on import or when thumbnails are generated. If you update the caption in RM it’s also updated in metadata. Other tools already exist for batch entering metadata into images, I’m not asking for that, I’m simple asking for RM to use, and maintain the data that can already exist.
It’s not a Windows file properties. Image metadata is platform independent and has existed for years. There is a Caption property, RM can use it. No huge complication. We’re not talking rocket science. Nearly every image management tool supports updating the caption property.
Below is what you can do with Excel PQ – but the EXIF data is not easily accessible:
When you use Excel Power Query to connect to a folder (using the Folder or SharePoint Folder connector), it extracts a robust set of metadata for every file in that directory.
By default, Power Query pulls a standard set of top-level columns, but it also hides a deep treasure trove of OS-specific properties inside a record column called Attributes.
Here is a breakdown of exactly what Power Query can extract.
1. Top-Level File Properties
When you first load a folder, Power Query automatically generates a table with these standard columns:
Content: The binary data of the file itself (used to combine or parse the file).
Name: The full filename, including the extension (e.g., sales_data.csv).
Extension: Just the file extension (e.g., .xlsx, .csv, .txt).
Date accessed: The timestamp when the file was last opened or read.
Date modified: The timestamp when the file was last written to or changed.
Date created: The timestamp when the file was originally created.
Folder Path: The complete directory path leading to the file.
2. The Attributes Record Column
Next to the standard columns, you will see an Attributes column containing nested Record values. Clicking the Expand icon (two opposing arrows) in the column header allows you to extract specific OS-level file attributes:
Standard System Attributes
Size: The file size in bytes.
Hidden: A boolean (true/false) indicating if the file is hidden in the OS.
System: A boolean indicating if it is a critical Windows system file.
Directory: A boolean indicating if the row is actually a subfolder (usually false if filtering for files).
Archive: A boolean used by Windows backup utilities to track if a file has changed since the last backup.
ReadOnly: A boolean indicating if the file is write-protected.
Advanced/Extended Attributes
Depending on your operating system and network setup, expanding the Attributes record or navigating deeper can sometimes expose a sub-record often labeled Content or Extended, which can yield:
Device: Boolean indicating a physical device file.
Normal: Boolean indicating no other attributes are set.
Temporary: Boolean indicating a temporary file (like the hidden owner files Excel creates when a workbook is open, e.g., ~$workbook.xlsx).
Encrypted: Boolean indicating if the file is encrypted by the filesystem.
SparseFile: Boolean for large files containing blocks of large zeros.
ReparsePoint: Boolean indicating a symbolic link or shortcut block.
Compressed: Boolean indicating if NTFS compression is applied to the file.
this is why using TreeShare with media is problematic. Also why one of many reasons why I finally moved to downloading, renaming with a naming conventions, organizing in a folder/sub-folder stucture then linking. I can have ONE media file linked to 5 or more people or linked to two different 5 facts and/or citations and so on.
Now it could be argue that RM can & should have some sort of function to create caption (Maybe similar to its sentence tools.
Right, but those are all Windows file system attributes, not image Metadata contained in EXIF. Image metadata is totally separate than file system attributes, and is a standard supported by nearly all image editors and all manner of command line tools. Caption is a standard field, up to 2000 characters, in the IPTC section of metadata. See more at: IPTC Information Interchange Model - Wikipedia
correct - but you can build caption with NON EXIF info and actually I have done that before.
Just wanted to clarify so no one was mislead that about the other info NOT being quite easily accessible via Excel PQ (Power Query) when it is easy for anyone who has used Excel PQ before.
