DAM help: filtering method(s)

[gabrielh]

Hello all,

I have read the DAMbook (second edition), and it really helped me a lot. But I have a problem which I present to you here:

As suggested in the DAMbook, I name my original files as "G_090101_1234.dng", for example, thus the letter G, followed by the date, followed by a four-digit number. When making a derivative, I add the postfix "_Deriv" to the file, thus naming the derivative "G_090101_1234_Deriv.tiff", for example. Both derivative and original have the exact same metadata (rating/caption/headline/situation/keywords/...). Now, of course, I don't process my originals immediately after a shoot, so sometimes it takes a couples of months to make a derivative. So here is my problem:

When I would like to filter a bunch of pictures, I'd like to be able to get the derivative files, if they exist, and the originals, if they don't have derivatives (yet). So, I'd like to be able to use an "if"-clause, which, if I understand it correctly, Lightroom or Bridge aren't able to do. I'd actually want to filter out any original file, if it has a derivative file, otherwise, not---as otherwise, if I filter only derivatives, I won't see any original not processed (yet).

So, my question is: how would you recommend me adapting this DAM method? I have thought about several things (like simply deleting the originals---but keep them on a separate drive, or add a keyword to every original and once a derivative is formed, give this this keyword while removing it from the original), but these look to complicated... I hope I can get help from here  !

Many thanks to any answer to this post, I greatly appreciate it!

Gabriel

[danaltick]

Gabriel,

The only way I can think of doing that without a specialized script at the moment would be to create a collection or catalog set that contains all your originals that don't have derivatives and keep that up to date as you import new images and/or create new derivatives.  Then just run a filter for your derivatives and add your collection to the filter when you need to see everything.  Of course this method is not much different than the keyword method you proposed, but unfortunately I don't know of any simpler way outside of using a script.  Hope that helps.

Dan

[gabrielh]

Thanks, Dan, for your reply! I really appreciate!

So, how could one make such a script? I'm using Mac OS X, could Automator or Script Editor do the job?

And... Have you also found this a problem when searching? Or do you organize your Originals/Derivatives in another way which doesn't have this issue...?

Thanks a lot for your reply!

Gabriel

[johnbeardy]

For Lightroom it's one type of script, written in an obscure program called Lua, while for Bridge it's necessary to write the script using JavaScript.

I have toyed with the idea of writing something similar, a "sync metadata from files with extension xyz to identically-named files with extension abc" and its code could be adapted. But it's not something I think I'll do in the near future.

Whenever I need to do something like that, I do something like this. Select one of the two groups, and give them a coloured label. Then get both groups of images into a single collection, sort it by filename, and display it in grid view with 4 across. I can then scan down and quickly spot a pattern.

John

[danaltick]

Gabriel,

I would have the foggiest idea how to make script.  John is the scripting expert.

I haven't really run into your problem because I'm usually only interested in derivatives and originals for a particular job or project I'm working on rather than on a global basis in my catalog with large amounts of files.

Dan

[gabrielh]

Thanks Dan and John for your replies!

As I'm not a programmer, I don't have any idea either on how to make the script...

But this has me led to wonder: am I the only one which encounters this problem?

I understand, as Dan noted, that this problem doesn't arrise when one is interested in the derivatives only or originals only. But what about those pictures which never will be processed, because they are "good enough"? Should these files just be saved with the "_Deriv"-postfix, just so a search on derivatives will be possible? Or is this some "overkill" on one way or another? 

Just asking away, wanted to have your opinion about the subject!

Every comment is very appreciated!

Gabriel

[danaltick]

Gabriel,

You should be rating your images as a part of your workflow.  Your higher rated images will rise to the surface when you filter by rating.  That is how I find my best images.  As raw editors continue to improve, I find myself making fewer and fewer derivatives.

Dan

[gabrielh]

Thanks, Dan, for your comment! I indeed use ratings in my workflow too, but every original and derivative get the exact same metadata, thus also the same rating. Should I do this differently? If I understand the DAMbook well, I think this is how Peter Krough does it too...?

Thanks!
Gabriel

[danaltick]

Gabriel,

That is what I do and I believe Peter as well.  If you try to narrow your search to rated images, some genre (e.g. architecture, people, nature, etc.), timeframe, etc. you will find that your search results will be a much smaller collection to visually decipher.  That's what I usually do.  Search filtering can be very powerful; learn to use it well.  If you still find you need to separate out those originals that have no derivatives, then I would keep a catalog set for that as mentioned above and just maintain it as you ingest new images and create new derivatives.  Hope that helps.

Dan

[markpirozzi]

When I would like to filter a bunch of pictures, I'd like to be able to get the derivative files, if they exist, and the originals, if they don't have derivatives (yet).

Gabriel,

The way I interpret your criteria is that you never want both the "original" and it's derivative in this search, only the derivative if one exists.

Im curious about the scenario in which you need to do this.  Do you only do a handful of images at a time or a large number?  (I assume it's a large number since a few would be easy to deal with manually.)  Why do you usually need to to make this search or what is the end product so to speak?  Do you ever have more than one derivative?

Mark

[danaltick]

Gabriel,

It just occurred to me what I do when I make derivatives.  I learned this from the book.  When I make derivatives I group them in a catalog set under the project I'm working on along with the originals used to make those derivatives.  Peter calls the set "Make Master Files"  in that set he creates several subsests, "Originals", Master Files", and "Deliveries".  If you do this for each project/job you have then you can perform a global search for all originals not under a "Make Master Files" set.  This will give you all originals that have no derivatives.  Hope that helps.

Dan

[gabrielh]

Dan,

Thanks for your helpful comment; this is indeed a good point---I must admit I read these parts of the DAMbook, but for my own workflow, I don't use this system---maybe one day I will, but for the moment I don't need to, I think...

Thanks again for your kind replies, it's nice to think about DAM! 

Gabriel

Gabriel,

It just occurred to me what I do when I make derivatives.  I learned this from the book.  When I make derivatives I group them in a catalog set under the project I'm working on along with the originals used to make those derivatives.  Peter calls the set "Make Master Files"  in that set he creates several subsests, "Originals", Master Files", and "Deliveries".  If you do this for each project/job you have then you can perform a global search for all originals not under a "Make Master Files" set.  This will give you all originals that have no derivatives.  Hope that helps.

Dan

[gabrielh]

Mark,

Thanks for your good questions, I'll try to answer them the best I can.

For the moment, as I'm new to DAM and the DAMbook, I have decided to keep originals and derivatives in the same folder---one folder per event organized by date on the hard drive. As I use Time Machine as backup, I don't need to remember which version of a certain derivative is up to date, it does it for me...

So, actually, I'd like to have an end product like this: for a specific event, when I want to browse through the images, I'd want to see only the derivatives if they exist, and otherwise the originals. When I currently see an event, I have to browse through the originals (which have derivatives), and when I want to show the shoot to somebody else, this is quite annoying...

Thanks for all comments, I really appreciate any input!

Gabriel

When I would like to filter a bunch of pictures, I'd like to be able to get the derivative files, if they exist, and the originals, if they don't have derivatives (yet).

Gabriel,

The way I interpret your criteria is that you never want both the "original" and it's derivative in this search, only the derivative if one exists.

Im curious about the scenario in which you need to do this.  Do you only do a handful of images at a time or a large number?  (I assume it's a large number since a few would be easy to deal with manually.)  Why do you usually need to to make this search or what is the end product so to speak?  Do you ever have more than one derivative?

Mark

[markpirozzi]

Gabriel,

Currently, the only way to do what want is to do it manually.  The main problem with using collections (or other manual methods) is you have to remember to add derivatives to the collection and remove the originals.

I use a series collections as a checklist (idea from John) to make sure tasks are done to files before "archiving".  I move a group a photos from one collection to the next as tasks are completed.  As an example of a task is to make sure derivates have the same metadata as the originals, if desired.  It's more efficient for me to move a large number of files through the latter stages of the pipeline, so the delay from taking photos to archiving them is fairly long.  Another example of a task is to make sure the files' star rating is written in the form of a keyword.

Mark

[peterkrogh]

Gabriel,
As others have pointed out, this is not easily possible to track automatically.  There certainly is no reason why this could not be possible. (hide all files where a derivative exists for an original).

At the moment, you'll want to use some kind of tags to indicate "this is a candidate for a derivative" and "derivative has been made".
Peter