Can i use dark-table without importing/catalogue? Just with file browser and side-car?

Quick question. Darktable looks awesome, but I hate importing and using a catalog.

I prefer to work in the manner of RawTherapee or Adobe bridge with Camera raw. Just like a file browser that leaves side-car files. No importing or program based catalog.

Can I work like this with darktable?

3 Likes

I’m not sure I fully understand what you are thinking but my approach may be similar. I do all my photo transfer, set up into folders etc. outside of dt. I use “import folder” and only load in the photos in that folder, then I tag/star/colour code and process them, export where I want them, cull out any raws I don’t want to keep as raw files. Then “remove” the folder.

I know many other import and leave all photos within dt but when I first started with the program it tended to crash more with many photos imported, so I just got used to doing it on a folder by folder basis.

Yeah, that’s kind of what I’m not into. Managing folders/catalogues/ film rolls or whatever…

Just need a file explorer…Rawtherapee gets it right in this regard.

Damn…wanted to use dark table, but catalogue is a deal breaker for me.

3 Likes

Just start with

darktable --library :memory:

After start darktable creates a new database in memory which is being deleted when darktable is being closed.
I still think this should be an option in the preferences:
https://github.com/darktable-org/darktable/issues/2386

1 Like

You don’t have to use the lighttable mode, as far as I know. You can just use any file manager and open the image with darktable (darktable /path/to/image), and it’ll be opened in darkroom mode.

1 Like

I’m using Windows 10 and I can right click on a photo, from say a folder on my desktop, and choose “open with darktable”

1 Like

Can you tell me more about this?

How do I apply this : ’ darktable --library :memory: ’ ?

Good to know! Yes one thing that is a shame is that you have to ‘import’ from one folder at a time. Rawtherapee has built in file browser so you can flip around, I like that approach.

Each software has its own concept. darktable includes dam features others doesn’t. So you can choose the tool that fits best to your needs. I won’t expect darktable to switch to a filebased concept since a lot of effort was spent to improve dam features :wink:

1 Like

Agreed…just my stupid preference, nothing more. Some like chocolate, others vanilla.

I think it is possible to compile darktable without lighttable somehow. Always wanted to ask this.

interesting question. in principle yes. also i think darkroom mode was there before there was lighttable mode. nowadays this isn’t such a good idea any more. just tried to compile without lt, which works fine. it even starts, but then crashes as soon as you do anything really.

also, if you’re not using the view, why would you be bothered by an extra library, it’s 342K on disk. the --library :memory: approach is probably more what you’d want.

These are starting parameters for darktable. On Linux just type it in the console, on Windows you can create a shortcut to darktable.exe and add --library :memory: to it (if I remember correctly)
https://www.darktable.org/usermanual/en/overview_chapter.html#darktable_commandline_parameters

1 Like

Thank you ill take a look at this…I saw a mention of compiling without light table. Not commenting on the possibility but just noticing some essential functionality is built into lightable such as ‘styles’ (processing profiles) ect.

Also, I think my initial assumptions about Darktables functioning were wrong. I had thought that my raw images were imported to a darktable database, and therefore duplicated and stored there. But it does seem to access the files directly in their folders and using a sidecar file. I open my files using the folder button rather than the import button. Maybe that is the difference?

Am i understanding this correctly?

Also, thanks for all the replys!

When you import an image into darktable (default settings) it reads out metadata from the file and

  • writes metadata into the database (on your harddrive)
  • writes metadata into a sidecar file (.XMP) next to the image

Basically the database enables you to sort and filter all your images. The XMP can be used to synchronise with other programs.

Both places contain your edits, too. Some users want to keep everything in a database and disable sidecar files in the options. Others use darktable just for image editing and don’t need a database. They can start darktable with the parameters written above.

1 Like

Yes, darktable never stores the image file itself in the database.

darktable keeps following data in his databases (data.db and library.db)

  1. image metadata and path of the file on disk
  2. development (history stack)
  3. tags, ratings, color labels
  4. presets from modules
  5. styles
  6. … for sure there is much more which I am not aware of.

The sidecar file on disk, which is created on import, contains 2 and 3 as well. If the database get lost, you will be able to recover this data from the xmp files. So I think it is a good idea to keep xmp writing on.

The difference between the import options seems to be that in case of an import from a device, images are copied from the device to a folder on the disk. In the other cases - import image or folder - the image itself is not copied or moved.

Unfortunately, the manual (darktable import section) does not state this general behavior clearly enough.

Personally I really enjoy the lighttable DAM features. darktable v3 got a new mode culling, which is cool for initial sorting. tagging is fast and easy. And there are much more features for a fast workflow.

2 Likes

I think it’s more the other way round, about being sure that everything is stored in xmp and I can safely ignore or delete the database, and opening the file again will get me back to where I stopped editing last time. Time correction was not written to xmp for a long time, only recently it was added IIRC. But there was something else important that is db only which I do not remember.

And 1 not? What about changed metadata such as title, copyright, …?

Btw, how are duplicates managed if there are only xmps but no database entries?

1 Like

1 refers to data stored in the original file and get extracted on import.

The list is for sure not precise and complete and from a users perspective. But I hope that at least the understanding of what goes into the databases is a bit clearer.

I forgot to mention previews: Previews are organized on disk as files in a cache folder and also not in the databases.

E.g. all your module presets? You can not export them. I think that was one reason why the database was splitted into data.db and library.db. data.db keeps your presets, list of tags, …

Perhaps slightly off-topic but I think I remember a case where I updated an XMP file outside of darktable, and darktable overwrote it with its old data from the database. darktable may have been open while the XMP file was modified, but it’s still an unfortunate thing to happen. Is there any way to make sure that darktable not only writes to the XMP file but also reads back from it when appropriate? Does it automatically do it on startup, or must the folder be imported again?

Enable “look for updated xmp files on startup” in core options. Not sure if that will do anything if you change xmp files while darktable is open, though.