3. step: filenaming-system (Renamer)

[sinus]

I works since 2001 with IMatch, before with some other DAMs like Portfolio.
Independent from the DAM I am still today sure (well, for me), that a good filenaming-strategy is the basic of a good DAM-system.

I tried several systems, but I like still the system with long filenames, containing several things, what are useful to know, even if I see only a file on a stick, without any DAM. In this case (and many other cases like backup and so on) I know quite good, what this file is, simply from the filename.

The only drawback is, that a long filename for some programs is a bit "clumpsy", because they do not  expect such long filenames. But IM5 is here quite good, specialy for example the layout for the file window.

And yes, of course I do not deliver a file with such a long filename to my clients (nor private), because for them this helps not that much. For giving shorter filenames I can use also some tools in IM5, like the renamer.

And finally, I want remark, if you have a consistent system, IMatch (also 3.6) makes it very easy to change a filename-system slightly, like I did also 2-3 times in the past 12 years.

My current filename-system is as follows:

20040303-1455-011283-s-kun-brunner-machines.nef
19990422-1600-034000-c-pri-designstuhl-yellowish.jpg
20130315-0724-025256-s-kun-monika-at-beach.jpg
yyyymmdd-hhmm-unique-*-***-text.ext

In front Year-month-day without any separators (separators are also possible of course like yyyy-mm-dd

Followed by a - and then hour and minutes.

Then a - and the follows a unique number. This number is created by a IMatch-variable and does count upwards.
The first * stays for "from where comes the image" (like from a scanner or digital cam...)

The second *** stays for "for whom is this image" (like for a client or private or ...)

Finally after the - comes a "free text", what I find very useful. It describes in SHORT words, what the image is, like names or describings like "dog at sea" or "vacations-venezia" and so on.
In earlier years I shortend such text, so that each file had the same lenght. Nowadays I leave it open, so that each file can have a different length.

IM5 give us with its power a lot of possibilities to format such long filenames.
We can also user parts of the name for data-driven things. In my case I have categories with data-driven cats for different clients (the *** in my files) or for other things.

The most important thing in my opinion is, that the filenames does have a kind of a system and hence are consistent.

Maybe this helps some users or gives some ideas.
I try to add more steps from my workflow here, from time to time.

Of course, other ideas are always welcome, surely also for other users.

[Richard]

My system is a whole lot shorter, eight characters. The first five are are the actual names and six - eight are used for versions, etc. Using 0-9 will allow 10,000 files. Hardly enough for many. If you include 26 alpha characters in upper case, that takes it to 60,466,176. That may cover many users but for those who need more, 26 lower case letters can be included. That will yield 916,132,832 distinct names while leaving three places open - yet staying with the 8.3 convention.

Many users include dates in their file names but most of my images could only have estimated dates. Something like: 17440101 indicates a precision that just does not exist. Therefore I prefer to put my estimated dates in Attributes along with an explanation.

As for including names of persons in a file name, it would be too complex for a group photo with 18 persons even when I know all the names. Instead I rely on Categories. Each family member has their own category. The labels for those categories begin with a complex number which causes the categories to sort into a family tree order. Thus when I select a thumbnail I can see which lineage categories the file has been assigned to and that gives me names as well as lineage information.

[JohnZeman]

Mine is also short. 
All I want and need are for the photos to be naturally sorted in the order they were taken so I name mine in the YEAR_MONTHDAY_HOURMINUTESECOND_SUBSECOND format so a typical file name would be 2014_0122_122433_000.jpg

[sinus]

As for including names of persons in a file name, it would be too complex for a group photo with 18 persons even when I know all the names. Instead I rely on Categories. Each family member has their own category. The labels for those categories begin with a complex number which causes the categories to sort into a family tree order. Thus when I select a thumbnail I can see which lineage categories the file has been assigned to and that gives me names as well as lineage information.

Thanks, Richard, for your interesting information. Of course, if I have a group, I do simply write something like "...-group-everton-" or "... gruppe-verwaltungsrat".

All names I do also enter into properties (in IM5 Attributes). These names goes also into the keys.
For private names I use also a category-system with names and numbers.

[sinus]

Mine is also short. 
All I want and need are for the photos to be naturally sorted in the order they were taken so I name mine in the YEAR_MONTHDAY_HOURMINUTESECOND_SUBSECOND format so a typical file name would be 2014_0122_122433_000.jpg

Hi John,
seems to be a good system, and of course automatically. Yes, the sortorder is also important for me, it works also for me, like your system. But, ok, mine is a bit longer

[dcb]

I just go a straight 6 digits i.e., ######.ext and for versions ######-##.ext. That gives me a million files with a hundred versions of each. All the other information such as dates is in the metadata. I have too many files with unknown dates that break under a date based filing system.

[sinus]

I just go a straight 6 digits i.e., ######.ext and for versions ######-##.ext. That gives me a million files with a hundred versions of each. All the other information such as dates is in the metadata. I have too many files with unknown dates that break under a date based filing system.

Well, I think, as long as your filesystem works for you, and you find your files, all is ok!
Although if you have a million files, then I would think about expanding your 6 digits to 7 digits 

[Photon]

For me most important is to keep the original unique numbers of the camera when I receive images from other people.
My naming convention is "yyyymmdd-HHMM_xx####_zzz.ext".
The unique number #### (e.g. 5761) is derived from the original file name, e.g. IMGP5761.JPG or DSC05761.JPG or P1050761.JPG, ...
This number can be quite helpful, when searching for specific files. Four digits derived from original file name numbers are by far sufficient.
With "xx" letters I am coding the aspect ratio and orientation of images and with "zzz" letters edited jpg versions, but this is quite special.

For my scanned images the unique number is generated with a MSDOS randomize command.
The short unique number is also quite helpful for friends/customers, when they need specific images or printouts.

[DigPeter]

I like to have some indication of where and when the photo was taken.  So a typical file name looks like:  20140221_Tenerife1234. where the last four digits are the last four from the camera.  I am unlikely to take 10000 photos in a day.

[stahl]

I just go a straight 6 digits i.e., ######.ext and for versions ######-##.ext. That gives me a million files with a hundred versions of each. All the other information such as dates is in the metadata. I have too many files with unknown dates that break under a date based filing system.

Hello dcb,
I am wondering how you keep track of the numbers that are still available so you can avoid duplicates? Do you count up, using a user defined variable as sinus described, too? And do you use a script?