I have it open right now. Sometimes it’ll forget who’s who; telling it to store name tags with each photo helps with this. That’s in Tools 
Options Name Tags.
I don’t know how it happened but somehow my photo libraries went from Scan Always to Scan Once. Many of my name tags got lost when I reenabled Scan Always and I’ve been going back through, making sure everyone is identified.
I just love the program. So easy to use! I was disappointed when they stopped supporting it.
True story: my mom fell and broke her arm. My sister flew up and cared for her. She took a photo of mom at the sink, washing dishes, and sent it out to let us know mom was doing ok. When I opened that, Picassa looked at the metadata from my sister’s cell, consulted Google Earth (?), brought up a map, and showed exactly where the picture was taken.
I wondered if any women had posted photos to dating sites and had pervs track them down.
I would hope that a dating site would strip out metadata on upload. But that’s a good point.
It sounds like you may not want to keep the original photos once they’re scanned, but please don’t throw them away. Check with family members – you may have a distant relative who’s a genealogy buff & would love to have & keep those old photos. If you aren’t in touch with family, Ancestry.com (many libraries offer free access) is a good place to look for researchers who might like to have them – just search on a family name to find trees with those names. You could also sell them on Ebay, but then you will likely have someone trying to return them to you – there are several groups dedicated to reuniting old photos with current family & they buy a lot of stuff on Ebay. And a local genealogy or history group would be happy to take them.
Digital copies could be posted on a family tree on FamilySearch.org (free). It’s likely someone has your ancestors in their tree & you could just add the photos to the tree, making them accessible to anyone interested in that ancestor,.
An absolute minimalist approach would be to have a document - label photos as you scan them, sort of “Roll5Frame3” or some such system (Or dates too…). In the document, add the comments from each photo. If necessary, a short description of the key elements of the photo - “Mom (Mrs. Jane Lobo…) washing dishes with injured arm 01/01/2019”. Even better with spreadsheet, you could sort dates.
the convenience of this is the one searchable document rather than metadata that could be lost.
Many cameras (and phones) will geotag the photo with GPS coordinates. It’s elementary to disable this feature, but some may be insufficiently familiar with camera options. I turned this off on my camera because GPS can be an additional battery drain.
That’s what I do. I have scanned thousands of family slides and photos. I put them in a directory named for the year of the photo, if available. If I don’t know the exact year, they go into a folder for the approximate decade.
In the root folder is a document where each photo is entered by name, along with any other information about the photo. You can then record what was on the back, but also what a relative told you about the picture, your own thoughts about it, information you gleaned from later research, etc. Even restoration notes if you get there. Having all that in one searchable document is handy - and the easiest way to record that info.
I would NOT modify the picture to include the text at the bottom. For one thing it’s a lot of work, and for another metadata just doesn’t belong inside the image as far as I’m concerned.
Other details:
Store the raw image from the scanner, and do any cleanup on a copy. When I do this kind of work the scanned photos go into a single folder called ‘raw’ or ‘unprocessed’ under the root folder of the project and the sorted, restored photos get the treatment described above.
So, I wind up with a directory like this:
[root]
…[Old Family Photos]
…[Unprocessed]
…[1940s approx]
…[1941]
…[1942]
…[1950s Approx]
…[1956]
…[1958]
…FamilyPhotos.txt
Don’t use JPG for this. JPEG is a lossy format, and if you open and edit the image multiple times it can really degrade. I prefer PNG as a good compromise between quality and file size.
Don’t store the metadata in a commercial program like Picasa - it locks you into using Picasa. Keep all your archive data application agnostic.
There might be a lot of people interested in those photos for various reasons, so I would keep the originals. Forget about the family members for a minute - old street scenes from a small town or city may be of interest to their respective historical societies, for example. I have donated a number of otherwise uninteresting pictures to my city’s history archive because they contained pictures of now-demolished buildings or pictures of school assemblies from the past, etc. When going through the photos, keep an eye out for interesting things in the background.
If you are going to restore any of them you should open a restoration thread. Some of us could help out for fun or just give some advice to fix some common issues. If you open an imgur or similar account you could post examples of damaged images for advice.
I use JPG, but then - treat the original - whether JPG from the digital camera or scanner - as a “negative”. take original, modify it however you want, how often you want, but don’t overwrite the original. Then if you decide to do more or different changes, you can start with the original again.
If you can accurately date the pictures, dates are the best differentiator. However, I ended up with a collection from multiple sources, my stepmother’s family and my dad’s stuff going back to his childhood. So simpler to have folders sort of like like -
Dad -childhood and family
-university
-England
-Canada
Stepmom
-stepsister’s family
etc.
My digital Camera photos go back to 2000 and are organized by year and sometimes event. (Vacation, Christmas, etc.)
Unfortunately, there are a huge number of photos of people I have no idea who, probably friends and co-workers from the 1950’s and 1960’s or earlier.
Thank you all.
I guess this has been asked and answered, but rather than use a photo editor to type in the information you want preserved, why not just write it on a piece of paper and put the piece of paper on the scanner along with the photo? That seems a lot easier than using an image editor to open each photo, add white space to the bottom, and type in the info. Anyway, glad it’s working.
I came to suggest this!
I digitised my gran’s photos recently. It was a joy to clean them up and colourise them. Before my gran passed away I got her to describe some of the colours of her clothes in the photos. There was no writing to preserve but if there was I would have added the text in a black bar immediately underneath the image.
Hope you did a better job than AI can do:
Remember when you’d send a roll of 126 or 127 film in for processing, and those square pics would come back—with a date printed in the white border? Of course, that was not be when it was snapped but rather, when it was developed. If you shot a roll of film in a short period of time, it was probably fairly accurate. But how many times did we hear, “Wow, I wonder what’s on this roll that we just finished!”? It was kind of exciting to imagine what might come back.
I think there was a thread in here about “What if I find a roll of film I found in the garage and take it for developing and it turns out to be kiddie porn etc.?” Not only can we not pinpoint when the picture was taken, but we can’t say who clicked the shutter release. When we scanned the pictures I mentioned, we found one of a house that didn’t quite look like one we lived in 30 years earlier and left. On consultation it turned out to be a place my folks had considered but decided against.
Otherwise it was usually obvious to us who the people were etc. But if you approach this as a document of historic value, something future generations may consult, I’d be careful about altering the source material. Don’t superimpose text over images, etc.
You could add a brief description of the photo in the file name - e.g. “1966 Mom and Dad at seashore.jpg”
Thanks! Yes I’d seen that Twitter thread before. AI is no fun. It’s enjoyable and therapeutic to colourise manually plus you find yourself researching the objects to try to get it right. However there will often be a large amount of guesswork involved. Will stop talking about this now lest I hijack the thread.
I tried to scan my dad’s Slides. My scanner had a special attachment.
I was extremely disappointed. The scans were dark and the colors were off. I spent several hours trying to fix a handful of scans.
Each slide required documenting the people’s names and location.
My dad had 20 carousels of 35mm slides.
There was no way I could spend hundreds of hours on that project.
I have a Canoscan 8800 and I found it scanned slides (and negatives) very well. I had seen B&W pictures of me and my brother (ages 2 and 3) sitting at a lookout point looking at El Capitan in Yosemite from the late 1950’s. However, until I got the slide, I did not realize it was a colour original.
Also many of the envelopes of pictures included negatives, so scanning the negatives was more productive than scanning snapshot-sized pictures, especially if the paper had curled or cracked. the only thing it does not do well is professional quality pebble-grain finish, like my uncle’s wedding party in 1950.
One thing the scanner was good at - I could scan very marginal poorly exposed negatives and use Photoshop to make them reasonably presentable… Although my niece has threatened me with dire consequences if I pass on the picture of her and my nephews in the bathtub when she was 4, or the one of her at 14 beside my motorbike in a swimsuit wearing my leather jacket and helmet. 
My dad took a lot of pictures of scenery during their later vacations. My wife and I spent a few hours in a hotel room one night holding up slides from a few giant boxes and discarding ones with nature scenery - well over half the collection. I also digitized a large collection of B&W negatives.
My nephew’s wife was being the “archivist” for my stepmother’s side of the family, so I could simply take the entire collection, write it to DVD and give her a copy. A lot of the people I could not identify, one or other of the grandchildren could.
I actually tried the AI colourization for a few photos in my dad’s collection. There were two, of my uncle about 16 in a military uniform and one sitting on my dad’s motorcycle - AI did an impressive job of colourizing them with very appropriate colours, and I sent them to my cousin for my uncle’s funeral and memorial service. For some others - where a face is not prominent in the picture, AI was completely lost.
It’s amazing what can be done, for example, with a flatbed scanner and something as basic as, “Microsoft Picture Manager”. Old pictures have to be pretty bad not to look markedly improved after scanning and basic editing.
Several years ago, I had a friend with about 150 old pictures that she wanted to save via digitization. I told her I’d give it a try. She was very impressed with the results, and I have no special skills in that area whatsoever. If I can do it, believe me, most anyone who is willing to work at it for a while can do it.
Middle of last year I spent some time going through my slides. Some from my Boy Scout trip to Yellowstone Nation Park in 1958.
If it didn’t show someone I still knew or exceptional scenery it got tossed in the trash. I scanned the rest in my flatbed scanner and made two DVDs.
After I’m gone, no one else will care.