I’m currebtly working on a project where I have to scan a lot of analog photographs from my familyarchive. I actually don’t need all the photographs to be scanned but I want to start it the right way so I can transfer it into a well sorted and intelligent structured digital archive lateron when I have the time to do it. Maybe someone of you has done this before and I’m open for your recommendations. The archive includes maily photo albums and some lose photographs. So here are some of my questions:
How to name the files
how to ensure that I know where to find the analog counterpart
What folder structure to use for photo albums and the pages of the albums
How to number serially the images on the pages?
where to store the names of the people shown in the photograph or additional informaions like year, happening, location,…
what software can you recommend in this process?
what is a good resolution to scan a huge archive and what file format to use? For the project I want to scan a lot of images in a fair enough quality with the right mix of quality and filesize. Afterwards i will decide which images to scan in a high reaolution. But the first scan should be good enough to store it in an archive so I could use it in the future for small prints or something like this.
maybe you know some websites with good informations for archiving or similar projects. It would be also great if you could post them here.
So that’s it. I’m really looking forward to hear from you. Any additional advice you have is also very welcome.
I would try to find/establish a system in the analog albums and then adapt that for the naming scheme (this would be step 3 in the above mentioned archival tips).
I like to name my files in a system like “year month day_keyword”, eg. “220314_Pi_Day”, so I know when I took them and whats roughly to expect.
You might have your analog albums sorted by year, so I would go by year, then enumerate the albums (unless they have a specific name), enumerate the pages and enumerate the pictures from left to right, top to bottom. To specify the enumeration, you could add a letter, eg. a for “album”, p for “page”, i for “image”, so it becomes something like “1975_A01_P023_I013_Christmas” or some variation of that - if you can associate the albums/pictures with a specific date, it might be a good idea to add that after the year:
“1975_03_14_A01…”
This way, you have the advantage of being able to sort your folder structure in a timeline that represents the physical objects.
Moving away from the folder structure to the individual pictures:
Most picture editing and displaying software allows you to store and display metadata right away (take Digikam, for example), however I do not know if you can add metadata in your scanning software.
In this metadata you can also add tags, ratings, colors codes and such; however I am not sure how good those things are interchangeable in between software apps - but that would be the place for names, locations, years, topics…
thanks a lot for you replies. I’ll develop a structure in the coming days with your suggestions and will get back to this and how it worked out.
I’ll use vuescan and can add metadata through it but don’t know yet how handy it is. I’ll give it a try.
But I still need some help with the format. I think it’ll mainly depend on scanningspeed of my scanner. I hate sitting in front of it and waiting too long for the image. For me every image can be about 30mb. So a fairly good quality should be possible. Not sure yet if I want to use tif or jpg. Propably jpg.
VueScan does not know about metadata. You need to use another tool like EXIFtool for that purpose.
JPG is fine if you do not want to further edit the images. Otherwise use TIF and at the end convert to JPG.
I recommend using EXIFtoolGUI for that purpose. But I would not use any of the commercial software to create a data base for archiving the images. You are then bound to their proprietary format, which might no longer be supported in the future (like it happened with Google’s Picasa). I only use ASCII-files and TIF/JPG images which I assume will last “forever”.
I use EXIFtoolGUI and geosetter to write the metadata. Then the metadata are exported as txt-files with EXIFtoolGUI. These txt-files are imported into an EXCEL spreadsheet each line holds all the information about an image with a link to display this image right out of EXCEL. I could use the filter function in EXCEL to select and search images or copy selected images to another directory for export e.g. on an USB-stick for friends and relatives.
Once the metadata including creation time are set for all images you could use EXIFtoolGUI to rename the file names to include date/time in the file name as suggested above like YYYY-MM-DD_HH_MM_filename, which is very convenient.
If it comes to the worst, I can always convert the spreadsheet into an ASCII-file.
I’ve been doing this very type of project. If you can organize by date that will be great. It is much harder if you have undated photos/ I have some that are over 100 yeas old… dates are hard to come by for these.
I started to organize by family. Separating by my grandparent’s and parent’s sides. and then grouping within them by their siblings, and so on. This should make working on the smaller grouping easier, rather than slog through uncategorized images. Still have many to cut out as they were in strips of negatives.
Use the highest scan resolution that is practical, this will vary from system to system as is dependent on the memory and processing power of your computer. Some photos that i’ve scanned are small maybe less that 4 sqr. inches. I personally scan them at high dpi to get as much detail as possible from the small form.
I have the original scans on an external drive and will work with a local copy of a small group at a time, saving as most likely as a jpg.
Will be working with a certain group of images that i need to finish, to have available for a family member’s 100th bday, and have her id some of the people in the photographs.
That’s exactly one problem I have. The archive is very mixed and the information I have from photographs differs a lot. So sometimes I know a name or two from persons. Sometimes there is a date or a location written at the back of the photograph. The albums have themes but no names. I think I’ll sort them roughly by date and family part and then I’ll try to number them. Dunno yet.
I’ve also some landscapes which are printed in 1 sqr. inch. Actually quite funny to print these so small. Propably a money issue.
Here is where a good database program will help you (digiKam, for example). You can base the folder structure and naming convention in something that you can define for all images (original album and picture position?), and then fill the metadata with the extra (partial) information you have. What the centralized database bring you then is the possibility of searching and filtering irrespective of the folder structure.
I would also think if an “Unknown” explicit value for certain metadata fields is needed or not, depending on the database and how do you want to search.
Yes, you’re right. I used digikam but after switching a lot of laptops and external drives it got too confzsing for me to handle the database. I wanted to give it another try when I buy a desktop pc with enough diskspace and everything keeps in place. I know it’s possible with external drives but I should read thr manual of digikam before. I guess it’s the next big project after thisone to a propper database in digikam for all my images.
why don’t you put it on youtube? so others can watch it as well. If you’re particular about sharing it with the unknown public, just set it to unlisted so it won’t be found by searching but only those with the url will be able to access it.
I put the digiKam tutorial up on Youtube. They compress it to the point where the text is almost unreadable, but it should work. This is my first youtube post so let me know if it works okay. Thanks.
Hi @troodon
Thank you a lot for the video. Made me remember that digikam is the tool i need for my images. So I think I’ll dig into it when I’ve some time.
The text in the video is readable very good. All in all it’s a very good video, calm, comprehensive, good video an audio quality. Very helpful. So thanks alot.
One question that I still have with digikam. Do you use the option to write the information in xmp files or is everything you showed in the video stored in the database of digikam. I remember that working with digikam and darktable in the same xmp-file led to some problems.
Another thing @troodon : ) It seems you’ve some knowledge about digitizing images. I would be very curious what thoughts you had before starting to scan regarding quality, dpi, file format,…
No I just write the metadata into the jpegs or tiff files, not xmp sidecar files. Raw files from the camera, such as Nikon NEF files, can’t be written into, so some users convert their raw files into DNG format, which can be written into.
Quality etc: I digitize all negatives and transparencies in raw and process in RawTherapee. If I expect to do more manipulation on colour or tonal range I save as 16-bit tiff, otherwise just jpeg. Most of the time I use a 36 mp Nikon D800, but sometimes a 24 mp D7100 for 35mm.
When I need large prints from 120 and large-format film, I’ll shoot several overlapping shots and stitch them together using Hugin in mosaic mode.
DNG as a container format allows setting custom EXIF data for individual frames. The RAW scan files are never changed again.
Do not archive to JPEG or any other “lossy” format alone. I did that 15 years ago, when scanning some of these photos. They are completely useless for any processing or even publishing now.
VueScan features change, the processing methods may improve, a better (lossless) image format may become popular, etc. Keeping the RAW scan data allows for easy re-processing in the coming years.
Just take an archivist mind-set, stay as close to the original source, and that’s a film roll.
