Hey, I'm building that!

https://photostructure.com/faq/why-photostructure/

The novel random "taste" UI makes navigating even very large libraries fun and serendipitous.

There are a ton of configurable image and video filters that can prevent the exif-less screenshots and other random nose from getting imported in the first place. Hop into the discord if you need any help with setup, I'm online.

(Also, know that I've open sourced a bunch of the core functionality, but this is commercial software, albeit with a very generous "free" tier).

Personally, with things as serious as photo libraries, proprietary software is a no-go for me. I'd rather use something with community support like Lychee than pray that e.g. photostructure will be available years later.

As an open source developer for more than 20 years, I agree--using proprietary software that "locks up" your work leaves you somewhat of a hostage to that app.

That's why all of PhotoStructure's metadata changes are stored using EXIF or XMP standards either within the file, or alongside in a sidecar (if you prefer). File organization is also completely customizable, and designed to work in concert with other DAMs, if desired. Details are here: https://photostructure.com/faq/system-of-record/

Also: if the business ends, the code becomes open source. Details here: https://photostructure.com/faq/why-photostructure/#if-photos...

This looks really nice and I think I'll try it soon in my home server. Out of interest, do you have some blog posts on the "how it's made" side of things? I see some toml files for configuration so that would point to core being written in Rust, what else? Thanks!

The frontend is in Vue. Both the FE and BE are in TypeScript.

Parallelism is provided by https://github.com/photostructure/batch-cluster.js/

Metadata reads and writes are via https://github.com/photostructure/exiftool-vendored.js/

Persistence is via SQLite: https://github.com/WiseLibs/better-sqlite3

Image transforms are via Sharp: https://sharp.pixelplumbing.com/

My more nerdier blog posts are tagged here: https://photostructure.com/tags/coding/

Thank you very much for all the links and detailed answer, I wish you all the best with this product and business.

This sounds like something I always wanted. But I can't really figure out if I understand correctly from your website. So could you please clarify: when I edit the keyword of a photo in your software, will this keyword be written back into the EXIF data of the original file? If so, you just got a new paying user.

That's a fair compromise I think. Photostructure does look quite nice, so I'll give it a go.

Haha, I already built my own solution. :-)

What’s it do differently for you?

IDK it's just tailored for my own needs. There's nothing fancy there, standard features: importing from all our family devices, optimizing for browsers, showing date/folder trees, face recognition, showing on map, geotagging, etc.

no demo?

Have a look at https://photoprism.app/ then.

I think their actual multi-user story is somewhat lacking, but if you want a shared picture dump with some features that help sorting through a bunch of random images it should work well.

I've tried most of these tools at home and I tried to get on with Photoprism, but once I got to the point where I'd increased the VM RAM size all the way to 32GB and it was still eating it all, I gave up; the experience wasnt quite right particularly for multi-user; what's the use in a family photo library if my wife can't access it too?

Right now I'm working on a desktop tool (Linux-first because that's what I run) to take a bunch of folders as input, find duplicates and let you clean up the duplicates or "merge to destination" because this has been one thing I've struggled to find something nice to use for.

It's early days and I'm only so far at the point of hashing images and detecting and counting duplicates, but as an experienced backend software engineer the UI tools are a real learning curve.

I'm also trying to make the UX clean and simple with next to no UX nouse.

I'm using Wails + Svelte; I've not worked with Node, Svelte or Wails before so I may or may not end up switching techs but the backend is in Go.

It's not one of these photo gallery/management tools but I've been finding that really, all I honestly need for my family photos that are reaching terabytes in size is: that I'm not waating space and I have a good archive of everything in one place.

I'm curious if this is a tool others might be interested in.

Weird. I'm interested to know your setup. Did you use MariaDB or the built in SQLite?

I have photoprism running on docker, which is running as a VM on a proxmox server that is a 13 year old Dell Optiplex 9020. Only 6Gb ram dedicated to the VM (2GB initial, ballon to 6GB), and photoprism uses 2 workers. I have a large 50k set of photo's from starting from when my daughter was born till now (17). I have never sorted through these photo's other than to dump them from my phone every 6 months (camera before smartphones become common). Initial import took less than 45 minutes and that was before tweaking settings. Full manual index with all options, still takes less than 45. Initial face recognition was about 33 people, once I tweaked settings, it's recognized 97 people with little input from me other than tagging the name correctly for a few photo's per person.

The only difference in my docker file vs the official are optimized settings, and I'm using a separate VM for the database server as I find multi tendency and backups easier with a VM. (VM only has 2GB ram allocated to it)

I'm extremely happy with it, as I always put off organizing my photo collect due to how massive it is. It's made the task so much easier.

I am however always looking for other tools. I'm interested in give yours a try, just wanted see what could be the massive difference in experiences.

I don't remember specifics, though I can find out; I have it seemed in Kubernetes with hand written manifests. I had to tie Photoprism to a specific node which had had more resources than it's peers; I would have to refer to the manifests to recall which DB I used. All storage was over NFS (meaning the storage speed was limited by my 1Gb network backbone), which probably created some of the resource issues.

Edit: as for my tool, it won't replace Photoprism, its primary goal is to deduplicate your photo library.

Not sure about the memory requirements. My container currenty uses 128MB RES memory. But I also only have a couple thousand images. Not sure how strictly in memory some DB/structures are. I assume importing an existing library takes a lot, as it does the inference/training for face detection and auto-tagging. I've seen that it uses a lot of CPU at least, haven't looked into the memory.

Multi-user is lacking to missing. I'm not 100% sure which state photoprism is in, as I was shopping for a self host service. OTOH I don't care about that for family. I'm not hosting anything for untrusted people, or to default backup everything, but as common library. So I'm fine if everyone sees everything and expect people to play nice.

But I do agree that multi user (and maybe integrated auto-sync for phones) are major missing features. I'm just in the situation that it doesn't really matter to me currently.

That's fair enough. Same situation for me really I don't need the auth functionality.

The way I had Photoprism set up is that mine and my wife's phones would sync images nightly to the server, the photos would be rsyncd from the public facing sync server to an internal one (not in the DMZ) then I'd have the server send me a link to the import page on Photoprism because it lacks the feature to autosync when new files are added; I think having to open the page while on LAN (link was internal only) and click a button was the straw that broke the camels back here for me.

I think I also had something like 10-12 cores dedicated to it, which alongside the RAM use that constantly crept up, and the manual intervention, was just too much for a photo management tool that wasn't really even managing the photos in a way that was useful to us.

Are you planning to make it available in some form?

It'll be open source for sure.

I had considered selling it for a nominal fee too but that comes with additional commitments from me, that I'm not sure I want to take on.

Maybe I'll just add Github Sponsors, KoFi or such and see if there's interest, when it's ready and released.

I haven't tried PhotoPrism, but I'll give it a whirl. In my case, I'm looking for the best way to enable multiple family members to upload, tag, and effectively browse tens of thousands of photographs as we all continue digitizing family albums.

I'm currently using HomeGallery[0] behind Authelia[1] for authentication to view so many images effectively. For uploading, I'd been using Nextcloud, but it began to lag noticeably after a few thousand photos. I switched to FileRun[2] with symlink'd photo directories and a user for each family.

With HomeGallery, I get the desired performance on mobile devices with de-duplication and tagging. My instance detects objects fine, but I owe it troubleshooting time to figure out face recognition. The "similar images" feature can be fun with so many photos. A nice tagging modal on keybind per image would be a nice-to-have.

Using FileRun for uploads works fine, but I also needed a continuous cron job for docker exec to generate any missing thumbnails.

[0] https://github.com/xemle/home-gallery (or https://home-gallery.org/)

[1] https://github.com/authelia/authelia (or https://www.authelia.com/)

[2] https://filerun.com/

I tried it, but was surprised by the way it imports pictures.

It does it in few steps, like reading metadata, creating thumbnails, transcoding for browsers, face detection, nudity detection (lol they really do have it) etc etc. The thing is _all_ the steps are performed for each picture/video in your library. Picture 5 won't be even visible in your library until pictures 1,2,3 and 4 don't have all the thumbnails, optimized versions, labels from TF model, detected nudity and many other things.

My 10yo two-core celeron nuc begged for help, I couldn't see it suffer.

I tried to contact devs on github, the response was "go import all your 40K pictures manually, subfolder by subfolder". I checked, it's now marked as "accepted answer", haha.

If this existed, but also contained AI object recognition to automatically tag and describe photos, I would pay actually quite a lot of money for it.

Google photos but self-hosted?

Photoprism sorta bills itself this way. I've found the AI features surprisingly useful, as well as the geotagging.

I'm also looking for "Google Photos but self-hosted" with the main feature being automatic object detection.

I just tried the Photoprism demo (https://demo.photoprism.app/browse), searched for "person", "grass", "wave", "head", "eye", "neon" and a bunch of other terms, nothing seemed to find any pictures although I took the concepts from existing photos I found.

Back button is also broken on the demo, which makes it seem like the most basic UX is not there yet in the application.

The quest continues...

Photoprism advertises itself as only having facial recognition, is that not correct?

It recognises objects in the photo as well. Here are some random labels from my library:

- Altar

- Bookcase

- Festival

- Keyboard

- Streetcar

Some are very wrong (A photo of a lake as "gallery"), some are wrong but tricky (red geode with white veins as "meat") and most are spot on.

Edit: In fact, I'm not sure that I have the facial recognition configured... perhaps its in a newer release

Yes please.

I'm pretty happy with Synology Photos on my NAS. It's got face detection, interprets the exif and gives you geo maps, has albums and sharing. Got it set up for my entire family.

It doesn’t back up photos. It exports them. Not the same. Take a slow motion video. The version on your NAS will be 30 FPS, slowed down in the middle. For a backup I expect it to be the 240 FPS version.

Does it have mobile upload?

Sure!

immich isn’t quite there yet but it’s trajectory looks pretty good to fill this role https://www.immich.app

This is on my list to keep an eye on. I like that it's just a simple front end and not trying to be everything like Photo prism. I use Photo prism but I don't like how heavy it is. I want to use something like immich but it's not quite ready for production usage for me.

I'm making something similar for my own use for now.

Create a trip, upload a 1000 of pictures, delete all but 100-200, see them on the map or on the timeline. Add comments to individual pics or to location or to dates.

And yes, the whole family can upload pics and they can be batch assigned a date or location.

Too early to show anything yet.

iOS 16.1 adds a family camera roll.

It's not a shared album with lower res dupes, it's a real camera roll implemented as iCloud photo stream that multiple people can be permissioned to, with photos from your regular camera going in either manually, or automatically based on geofencing or proximity to other roll members. It can also retroactively suggest merging photos from shared events etc.

Your main camera roll then can show only private photos, shared photos, or both.

Apple One for Family includes 2TB storage and account owner can now optionally merge in their own 2TB for a total of 4TB for family photos.

In other news, iOS 16.0 now makes it easier for photo notes to end up in Notes not photos, also easier to snap send and delete, snap images of text to actual text, etc, many easier ways to avoid gunking up the photo roll.

photoprism is pretty good. Open source, face recognition, and automatic classification of pictures based on their content and location.

https://photoprism.app/