Workflow to embed photo caption

[dpop100]

I am scanning my parents' slide collection from the 1950's. My mother has documented each slide with a lot of detail and I would like to save that directly to the scanned image file. I understand the amount of effort involved but I want to be able to automate the process as much as possible and insure that I use an image standard that will hold up over time. My ideal solution would be to save the filenames with description, location and date taken to a spreadsheet or CSV file and then run a  script to apply to each image as embedded data and as a Watermark along the bottom of the photo. I'm sure I am not the first person attempting this process and I'm pretty sure imatch can do it. I would appreciate any thoughts.
I have attached a sample of the slide and the results I would like.

David Poplawski

[Mario]

1. IMatch can import data from CSV files - which allows you to manage the data in Excel, export to CSV and then import the data into proper metadata fields like title, description, headline etc.

2. The metadata you edit in IMatch (via the Metadata Panel & Keyword Panel or via import from your CSV file) is embedded to the image file itself, as IPTC/XMP data and hence
a) can never be lost and
b) is available for all modern applications, from Windows Explorer to Photoshop. Also for web sites, when you upload your files at some point.

3. It makes not much sense to modify your original files/scans to add a caption border or similar. I would not do that.

a) Instead, keep your scans unmodified, with the metadata properly embedded by IMatch as XMP.
b) If you want, you can export your images via the Batch Processor, adding a canvas to make room 'below' the photo and there you use IMatch variables to render the metadata you have created in step 1. You can do that on demand.

If you have never used the Batch Processor, watch the IMatch Overview video here:

https://www.youtube.com/watch?v=q7x4oelSvYw

It shows how the Batch Processor works and an output very similar to what you have in mind at around 2:15.

The general idea is to embed data like titles, dates and descriptions in XMP data inside your images first.
Then you can use that for sorting, searching and for export, e.g. via the Batch Processor.

[mastodon]

dpop100, another question: Where do you store the slide roll number and the slide ID number. Or what kind of identification do you use to connect the slide and the jpg file?

[jch2103]

I am scanning my parents' slide collection from the 1950's.
...
I'm sure I am not the first person attempting this process and I'm pretty sure imatch can do it. I would appreciate any thoughts.

No, you're not the first person to do this, and yes, IMatch is highly suited to doing it. IMatch gives you lots of options, so you're doing the right thing by sorting this out before you get too far in the process.

To add to excellent suggestions by Mario and Mastodon:

- Process. You may find that you can skip the process of putting metadata in a spreadsheet and then importing the data (unless you already have such a spreadsheet filled in). IMatch gives you considerable flexibility in adding metadata to images, including to batches of images at once, which could simplify your procedures.

- File names. You may want to consider using a date-based file name that includes slide number and/or side sequence number (E.g., '1955_08 559.jpg, etc.), along with a date-based folder system. Also keep in mind the excellent Renamer tool in case you change your mind (!) on naming, which I have done on more than one occasion...

- Dates. You can take advantage of the IMatch Timeline by adding the original date to your metadata. Be aware that metadata date names are somewhat similar and confusing; see https://www.photools.com/community/index.php?topic=1129.0 for some suggestions about choosing metadata date tags. One issue about digitized images is that you usually only know an approximate date (e.g., August 1955). Some users use notation like '1955-08-01 00:00:00' to indicate approximate dates in August, for example. For additional insights and options for approximate dates see https://www.photools.com/community/index.php?topic=7483.

- Additional metadata. In addition to title, description, headline, etc., you might or might not want to consider things like film type (your sample appears to be Kodachrome slide film, for example) and camera make/model. For film type, I finally decided on using hierarchical keywords (e.g., Film/Kodachrome, etc.). This kind of data is well suited to batch entry. Depending on what's in the images, it's possible you'd want to use face recognition; there have been numerous discussions about this topic on the Forum. This is something that could best be done after the rest of your process is complete.

- Location. It looks like the slides you're working with are well-labeled as to location (you're quite fortunate in this!). I'd suggesting using the Composite location tags for this (e.g., Country, State, City, Location). That would enable you to quickly find all slides taken in Garmisch, Germany as an example.

- Keywords. Lots of options here, based on what your needs are.

- Output. As Mario suggests, your most flexible approach would be to add metadata first. Once that's done, you have a huge amount of flexibility for how you want to create other media/forms (e.g., prints, online, etc., etc.).

Good luck! Let us know if you have other questions, etc.

[dpop100]

Thank you all for you replies!
Mario - I have used IMatch forever but I still only use 1% of what it can do. I never even ventured into the Import/Export option. Wow. I am learning all the tweaks to place the text in the right place and size on the canvas, but this does exactly what I am looking to do.
I will need to digest the IMatch process to import the data into proper metadata.

mastodon - I usually name my image files using the format YYYYMMDD-Subject-####. Fortunately my mother, in addition to the details, has labeled all of the files sequentially. You can see that in red pencil on the original slide as number 559. I have not gone though all the slides but I am hoping they remain sequentially numbered and unique!

jch2103 - Thank you for all your suggestions, specifically clarifying the various date/time options. I have never used the Commands/Image/Modify EXIF Date and Time, but that will be useful to stamp batches of similar images. And I will definitely use the Composite location tags.
I looked at adding the titles & descriptions using the Ultimate Description Writer. Not sure if there are other/better methods in IMatch, but I think I would feel more efficient documenting all the information from each slide to a spreadsheet, then process in IMatch

[ubacher]

Another view to ponder: You will want to keep the info for each slide but do you really need it inside the metadata of the stored image?
What about just photographing each slide to show the description and then give this image the same name as the scan but with a postfix ("_info")
The images will be side by side in the browser (assuming standard sorting) and thus the info close by. You can stack each image with the info image
so that the info only displays if the stack is expanded.

Date and Location will probably be the metadata you will want to add into each file.

If, at a later time you do need to add the description into the metadata for selected images it will be easier to do so.
Anyone out there who has a good idea how to do this: Display the info image large (Quickview?) yet add metadata into another(adjacent) image?

[Mario]

Anyone out there who has a good idea how to do this: Display the info image large (Quickview?) yet add metadata into another(adjacent) image?

That's easy and a standard feature of the Metadata Panel.
Select the original first, then _info. Now you see _info in the Quick View Panel in large.

The data you enter in the MD panel will be written to both files.
If _info already has the metadata, just click on the pen icons in front of the metadata tags you want to write to both files.

See the Metadata Panel tutorial in the IMatch Learning Center for more info: https://www.photools.com/imatch-learning-center/

[jch2103]

Ubacher ninja'd me. I was going to add that adding a scan of the entire front (and possibly back) of the slide might be helpful, whether or not you add metadata from notes to the image. As already noted, IMatch makes it easy to manage sets of images together.

In my case, I have a number of Cabinet Card format family portraits that are ~130 years old or so. I've scanned the entire front and back of these as there's interesting information there (e.g., the name and location of the photo studio is often on the back in addition to any handwritten notes). Unfortunately, few of the family 35mm slides in my collection include useful notes...

[dpop100]

Some great thoughts there. I like the idea of using the same name as the scan but with a "_info" for each slide face. Having the slide face with the actual handwritten note has a lot of charm which I hadn't thought about. I did have some foresight when I started scanning to scan the face of each slide group of about 20 slides (see the attachment). My purpose of scanning these images was to have a legacy for my children and their children. Having this data stored in IMatch is fine for me, but maybe not of so much interest to future generations.
The depth of what can be done in IMatch continues to amaze me! I only wish my mind could absorb it.  Thanks to you all.

[Mario]

The general idea of recording your data in proper metadata inside the images (with IMatch) is to conserve it for the future.

Having a scan of your slides with all handwritten comments is a great way to persist and document the information.
Copying some/all of this data into the metadata in your images makes the information available for searching, sorting, categorization (data-driven categories!) and for producing 'outputs' with tools like the Batch Processor or Design & Print.

If you upload your files now or at some later time, the embedded metadata will allow the service to organize your images.

[dpop100]

Thanks for all your help and direction on this effort. I am finalizing the type of data I want to import into XMP. I was intrigued by the links that jch2103 mentioned and would like to utilize a Circa date that was mentioned here https://www.photools.com/community/index.php?topic=7483.msg52194#msg52194  and http://u88.n24.queensu.ca/exiftool/forum/index.php/topic,7956.msg40580.html#msg40580
It seems to be {File.MD.XMP::iptcExt\CircaDateCreated\CircaDateCreated\0 } but I do not see it when searching the Metadata Tags. Can you tell me how to add the value when importing using the CSV tool?

Also it was mentioned to use Composite location tags (e.g., Country, State, City, Location)
https://www.photools.com/community/index.php?topic=2402.msg15473#msg15473

Would that be {File.MD.Composite\City\City\0} and {File.MD.Composite\Country\Country\0} under the "Standard" Tag selector?

Thanks
David

[Mario]

I can find the XMP::iptcExt\ArtworkOrObjectAOCircaDateCreated\ArtworkCircaDateCreated in the tag selector.
Use the standard tags where possible. this includes the standard tag named city and country.

ArtworkCircaDateCreated is a rarely used tag and I don't know if it is a good idea to use it.

All date-based features in IMatch (from sorting to searching to timeline) are based on the created/digitized standard tags. I highly recommend using them, even if the exact date is not known You can always use one of the other XMP tags or an Attribute / collection to indicate that the date in these fields is not precise. For most purposes using a quarter-based (1/1, 1/4, 1/7, ...) date schema with a note that the actual date is unknown is more than sufficient.

[dpop100]

Mario,
Thanks for the tips. In the past I have used YYYY0101 to indicate sometime that year, but I like the idea of breaking it down into quarters/seasons where possible.
I will definitely use the date fields you recommend. Can I assume from your answer that if it does not already appear as an item in the Tag selector, that I cannot add custom values.

[Mario]

When I do a

exiftool -list -XMP-iptcext:All

to list all tags in IPTCExt the Artwork circa date is included. This means it is also included in IMatch and accessible via the tag selector dialog.
Just search for circa to find it (don't use a group or use the iptcext group).

[dpop100]

Got it. Thanks!