[ This article was first published in the May, 2007, issue of
Larry’s Final Cut Pro Newsletter. Click here to subscribe.
Updated in 2007, 2008, and 2009 with comments from readers. ]
Figuring out how to get your files properly named and organized is one of the big challenges in editing, because there’s no sense in putting a file into your system if you can’t find it later. So, last month, I asked readers to contribute their file naming and organizational ideas. And I was amazed with what you sent in.
Take a look at what follows – this will cover everything from simple projects to over 500 broadcast shows a year. (Plus, if you read carefully, there’s also a free organizational goodie in here for you as well!)
Larry Jordan’s System
Here’s the article that describes the procedure that I teach in my classes:
Mark Raudonis – Bunim/Murray Productions
This is an incredibly UNDERVALUED issue. If you can’t find a file, you can’t work… period.
We have one simple rule: Never label anything FINAL! That’s the kiss of death. Call it final and you can bet your life that it won’t be! Then what do you do? Final, final? Stick to a “version number”. That way, when they decide to change the “final” it’s only an increment issue. For example: EXT1cut7 NOT”: FINALEXTCUT” .
For tape names we use the classic, “date, camera, load” system: 715A03= July 15, A camera, third load. Then, since we’re working on at least six shows in post, we append a “show ID” like “RW19” or SL5 etc. EX: 715A03RW19.
Along with “filenaming” conventions, you may want to discuss “bin naming” strategies as well. For example, we insist that all editors create a bin called “1.Current cuts”. The “number 1” insures that the bin rises to the top of list and makes it easy for a coworker to find the “current cuts”. We also require that there ONLY be current cuts in that bin. Everything else goes to a second bin labeled “2.OldCuts”.
Organization is the key to creativity!
John Ramsden – BBC
The naming convention we use for BBC London news is short, simple and generally people stick with it. The server is a Leitch 1TB, but other computers are involved meaning that there are a limited number of characters and no spaces (like old DOS names)
First simple rule is:
Each story is named in the running order, and the first part of its media name is the first 5 letters of that running order name or ‘slug’ (for rushes or edits)
The date is important as material over 2 days old is wiped, so the date is included. (For material to be kept longer a later date can be used or for material that is never to be deleted, KEEP is added.)
For rushes a short description can be added.
For edits the reporters initials are added.
For edits the bulletin name is added.
For edits the type of story is added.
The end result for rushes is:
The end result for edits is:
[SLUG]_[DAY]_[MONTH]_[YEAR]_[TYPE OF STORY]_[REPORTER]_[PROGRAMME TIME]
Max 5 letters with leading zeros package or OOV first letter of first name & 628, 658, 728, 755, 828, 858, 1330, first 2 letters of surname 1520, 1830, 2230, SAT, SUN
The types of story are abbreviated to:
Example: TUBE_18_01_02_P_AWI_1830 (a tube package on 18th Jan 2002 by Andrew Winstanley for the 1830 bulletin
Rushes are saved as the original tapes, edits are automatically archived to tape overnight.
Other information that could be included but is covered by another part of our transmission system is the duration.
Hope this provides a starting position for anybody writing a naming convention.
Aha. File naming. I don’t declare myself to be an expert in any area (it seems just too egotistical), but this is something that I have thought a lot about over the years. When it comes to my physical possessions, clothes etc I’m really quite messy, but on my computer I tend to be a bit more anal and I just can’t stand badly made filenames.
In addition, I was involved in IT for many years (albeit in a semi-professional way – I didn’t have any formal qualifications but was the guy everyone came to with their computing problems) and in one of my roles I redesigned the computer filing system in an organisation. They had a Windows system and their computer files were a total and utter mess. I’m sure many people relate to this – they had some awful filenames: files called “Document.doc” or “May.xls”. Their structure, well, really didn’t exist in any logical format.
When I take on a task like this, i tend to go a little over the top on detail. I ended up listing the names of about 1500 files and trying to form a file naming convention which could deal with any of these files. I redid their whole directory structure, tidied all documents up and issued all staff with a file naming convention that I designed. The point is, I thought about it a lot (maybe a bit too much!) and spent a lot of time analysing what the files were storing and how best to describe this in a name. I tried to trade off the length of a filename (and therefore the hassle to type and read) against the amount of information it carried. The convention I came up with for this particular organisation was as follows:
DocumentType DocumentDetails <VersionNumber> Date.extension
This describes the type of document. It could be a Letter, a Logo, an Ident or a Budget for example. It’s important that this doesn’t just give the file type – we can see this from the extension and therefore the icon assigned. Saying it’s a Word document or a text file in the filename is a waste of characters and doesn’t give any new information.
This holds more details about the document. I realised that trying to prescribe what goes in here for any file at all is difficult due to the diversity of documents needed to be named. So I gave a list of the kind of extra details it could hold based on the document type. For example, with a letter, it could hold the company name and a condensed version of the subject line. If the company name was possible to abbreviate, my recommendation was that this shorter version was used – Letter to BBC about New Series 11-03-05.doc rather than using “British Broadcasting Association”. With the BBC this is obvious common sense, but with some other companies it might not be. I also stated that this section could not be any longer than about 5 words (preferably 3 or 4). It should describe the file very briefly and be immediately obvious.
This was optional. Some files are in constant revision, whereas for others a revision is meaningless. I recommended a simple V1 for the first version then V2, V3 etc. If it was a file that was being revised very regularly (for example software) this could be made into a 2 or 3 figure number to indicate minor revisions – i.e. V134 would be 1.3.4 (although this does mean no number could be more than 9). This was because at the time Windows couldn’t have more than one full stop in the filename or it would read it as an extension. Mac OS X and modern systems now can.
This could be a number of well defined formats:
Format 1: Date: XX-XX-XX i.e. 09-04-01 or 11-12-02
Format 2: Month and Year: Xxx XXXX i.e. May 2004 or Jan 1998. The Month was abbreviated to 3 characters to save space
Format 3: Quarter: Xxx-Xxx XXXX: i.e. Jan-Mar 2003
Format 4: Year: XXXX i.e. 2003
Format 5: Financial or Academic Year: XXXX-XXXX i.e. 2000-2001 (this is a new one I’ve just added)
The important thing is that it’s all consistent.
There were a few other guidelines I had to make the names readable:
Incidentally, no one really bothered to follow this scheme properly, which somewhat depressed me at the time, but which I could sympathise with. It did help people to form slightly better filenames though. It was interesting to see how people started to take more care over forming filenames. I also made a whole directory structure to accompany these naming specs.
Nowadays things, especially if you’re using Mac OS X 4.8, have significantly changed since I formed this scheme:
With my system at home, I don’t really use as strict a naming convention as above – it’s only me using the computer, but I still type filenames in Title Case and include version numbers and dates. When I see others with sloppy file naming habits, it drives me crazy!
I suppose the principles I still work to are:
When I’m editing, I’m using a directory structure based on the one Larry advises:
Software companies (specifically Microsoft) have been saying for more than a decade that they are going to overhaul computer filing systems. The way forward is with having a true, database driven filing structure so it doesn’t matter what directories you store things in, and there are no real “directories” as such – just a load of interlinked data that can be searched in a fraction of the time it takes Soptlight. It’s long overdue and we’ve had the technology to do it since the late 80’s. Personally I was getting really excited as Microsoft said that this was going to be with us in Vista some years ago (WinFX) and got some early beta versions of it designed, but has delayed it yet again. Now I’m just impatient. Here’s hoping that it’s one of the surprise features in Leopard. Hmmm. I think not, somehow.
Right, I think that’s all I have to say on this topic.
Neal Glover – Victoria, Australia
When capturing to FCP after a shoot, I use a clip naming convention with two groups of 3 digits and an X, like this: 001x001x
The first group refers to the shot number, and the second refers to the take number. The X’s not only separate the numbers visually, but also when shooting I occasionally end up with pick-ups and other shots (added because they seemed like a good idea at the time!) which can be referred to as a, b, c and so on. It keeps all the clips in order when they are listed in the Browser, and I can easily refer a clip back to the storyboard, shooting schedule or shot log. Works for me and the students I teach at a technical college!
Anders Teigen – Norway
I am an editor in Norway, and my main body of work consists of documentaries – often direct cinema or other material intensive forms. I’ve often got several thousand clips to handle, and a good file naming convention is essential. This is the basics of my system, and has worked for me since FCP 2.0
There are two major concerns which I address in my system:
I always log the clip before capturing. I make one bin for each reel.
From the start I use three digits, only numbers. 001, 002, 003 and so forth. Same number of digits on all reels makes all sorting work in both FCP, in the finder, and in other aplications. EDL export is for instance dependent on a simple reel name regime.
Again I write the reel number, check the checkbox. This way all clips will have names starting with the reel number, which is step one in making a clip name unique, and making it possible to locate it with just a glance at the timeline (I use names, not tumbnails in my timeline. For speed and tidyness)
One or two words pointing to the scene, i.e. location or subject. This is the second step in making the clip name unike. Check the checkbox.
Two digits should do it. Starting on 01. Check the checkbox.
Optional for multicam. Using letters, A, B, C, etc. Numbers would be confusing because of shot/take. Use different reel numbers for each camera, write in lognotes which reel(s) “belongs” to the scene. Check checkbox if used.
The clip name/file name will look like this: 004_The stables_07
A short name is important, too long names will be shortened for the file name. Sadly I do not know how many characters are the limit.
I use log notes for everything else I possibly would want to search for later, Starting with scene description and characters, continuing describing image and dialog content. It is important to be consistent in the choice of words, and spelling must be correct. When I do a search, I want every relevant clip to show up. Misspell a name in a clip, and it will not show up. I should add that I have until now relied on transcripts on paper with timecode pointers. I would like to try a logging program some day soon. The FCP browser doesn’t break the lines, and makes it difficult to read longer notes.
If there are TC breaks, it could be hard finding the correct piece of media if it comes to redigitizing. I handle the segments of the tape divided by the breaks as independent reels – using letters: Say there are two TC breaks on one reel, that makes three segments. 004A means that I will find a clip BEFORE the first TC break on reel 004. 004C tells me to go looking for the video after the second TC break.
Ed McNichol – University of Washington
Background: My organization provides the media creation services for a large University. We produce content for the 17 schools and colleges that make up the University of Washington. On our main Seattle campus, we have more than 70,000 people. The programs we create are distributed on UWTV and the ResearchChannel, both carried statewide on cable and nationally on the DIsh network (11 million homes). We also produce a wide range of media that is distributed via DVD and the Internet. In addition, we support the development of Internet2, the next generation network, by transmitting high-bit-rate video information (multiple streams of uncompressed HD).
We operate as a self-sustaining entity, and charge for our services. We work on 300-600 projects a year, and each project can involve up to 20 or 30 hour-long masters. We have 40-60 active projects at any point in time. To scale our resources to accommodate both busy and slower periods, we rely on a large part-time hourly (freelance) work force.
My department focuses on the editing of this content. We have choose Apple’s Xsan as our Storage Area Network solution. We run with three fully outfitted Final Cut Studio suites. Due to our heavy workload, we must operate at peak efficiency. This involves having numerous people work on each project. Because of this, it’s essential that we all use a common terminology. It’s also important that we use a standardized system that can scale from simple to ultra-complex projects.
Creating the System: To create a scalable, uniform system, I first forensically went through completed projects to observe how various team members were naming items. I then crafted the first version of our system and worked to get feedback long before I deployed the first iteration.
Basic Premise: I created a system that is available IF people need to use it on their project, BUT doesn’t mandate an overly complicated system for simpler projects. It also provides a uniform naming convention for items related to a specific project. My theory is that, for frequent tasks, we should take the time to get it right once, and then replicate that success on future tasks.
The System: We have created a folder structure based on our 4 digit project numbers. Everything runs from a single template folder. At the start of a project, the editor or artist duplicates this folder and renames the copy with their project information.
The folder contains a set of subfolders and other assets that we use on most projects. At it’s most basic level, an editor can just use the main folder as a giant bucket to hold all of the assets for their project. (We’ve standardized the location of FCP assets on a separate location on the Xsan.) But as the project becomes more complicated, they have subfolders available for use to help them organize. Our top level folders are:
We’ve added an underscore “_” before these folder names so they get sorted apart from user generated files and folders.
So, as a project is underway, the editor or artist can either put the files they create into the main folder itself or the appropriate subfolder. Or they can wait until they have time, and then organize their content.
Eight of the thirteen subfolders contain 5 identical subfolders. These are;
The intent here is that, once again, these sub-subfolders can be used IF NECESSARY. But it’s perfectly acceptable to have all of your stills, for instance in the Stills folder. But as your program grows, you can use the subfolders to quickly organize the project.
You’ll notice that these subfolders have a word in parenthesis after their name. This is the shorthand we use in file names to indicate file contents.
For instance, an editor may create a Photoshop folder for a lower third. They can store that in the main project folder, or keep it in the Lower Thirds subfolder. Either way, we encourage them to add the appropriate tag to the name. In this case, that would be “EDIT”, since it is an editable file. Now, if they were to rasterize the text for some reason, they should store that version of the file with the “FLAT” tag, since it can’t be changed.
If the client were to send in their own files, we always keep a copy of that marked RAW, so we can always get back to their native content. If we have a change to an existing program, we’ll move the old copies to the OLD folder.
Some of the main folders are unique to that particular workflow, like DVD.
In addition to the folder structure, you’ll notice that there is also a default FCP project provided in the template. We rename this (to match the folder name). It contains standardized bins that match the naming convention in our folder structure.
Schema: OK, so that’s a lot of writing to get to file naming conventions, but it’s important to know that file names don’t exist in a vacuum. Here’s the standardized names we use at the University of Washington:
To describe the contents of these files, we use these terms;
So, I would summarize by saying that folders play a more important role in our system than file names. I hope this information is helpful to your readers and my colleagues.
I’ve attached a copy of our folder structure, but have deleted many of the larger files that we use, such as DVD label templates. I taught a class at NAB on “SAN Strategies and Workflow for Editorial,” and covered this material in that class.
Click here to download the template (120KB).
Don’t know if this helps, but here’s what I do. First, create a folder on my media drive with the name of the project. All my files go in here (except captures which go into the Capture Scratch disk). For my footage, when I was using tape, I came up with this and it worked great. Say I’m doing a live concert video with three cameras. Each camera had a designated number. My main cam would be Camera 1. So I’d label the first tape, C1T1, for Cam1Tape1. Let’s say you’re loading your third tape of the event into camera 3, that tape gets labeled C3T3. All tapes also have the name of the event and date on them.
When I captured, since some tapes had glitches and required more than one pass to get the whole thing captured, I’d label my captures similarly. For example, the second tape from camera 1, the first capture attempt would be C1T2C1. That means, left to right, Cam1Tape2Capture1. If I hit a TC drop out, the capture I had to start at that point would become C1T2C2, Cam1Tape2Capture2.
To tag them by project, I’d use the Project name in the Reel name field when capturing.
In my Browser I’d have one Bin called Clips. Inside that would be three Bins called Cam1, Cam2, and Cam3. Inside those would be the captures for each camera.
This is a very simple, basic labeling scheme. It works for me, personally. If I need to cut a tape up into individual scenes, I use the DV Start/Stop Detect according to my tutorial in the May ’06 Cut Lines column in Event DV magazine:
I’ve freelanced for a marketing firm that has several clients in the gaming industry, all with similar sounding and looking video (there’s a formula to designing a successful casino). They assign each client a discreet 4-digit number, and each project for a specific client a 3-digit number. Then, each clip name is preceded by the hyphenated number groups:
1234-123 CU hands shake dice.
This worked well on a couple of levels. One, the staff editor never separated his clips into folders on his array (they started in-house editing on a VideoCube system). So his clip list was enormous, but sift-able (maybe not a real word) by clip name and, therefore client and job. it was easy for me as a freelancer to find clips and import them into various timelines and bins. Second, if the producer requested the use of a clip from another job, he just referred to the client-job number. Also, after awhile, they just started using shot numbers referenced by the scripts to shorten up the names.
It worked for them since they “re-purposed” generic clips between jobs. They then used the same naming system for media backup to DLT. At last count they had over 500 DLT tapes filled with footage.
My brain works chronologically! I remember the order that I shot things and apply that to my naming configuration after capturing… so I can view everything chronologically.
I use the 6 digit date format to start off the naming of my tape. if I shot today… my first tape would be named 042007 Tape #1, Then 042007 Tape #2 and so forth. That is the actual name that goes on the tape… and I will add a location and description if I need to.
While capturing, I always find it a big pain in the ass on Final Cut Pro’s naming convention. Reel / Take / Description and whatever else. So what I do is capture the entire tape and rename the individual files using Adobe Bridge
My naming convention is always the same: Date Location Description
Occasionally I will add a sequence number between the date and location to keep things coming up in order. There is nothing that kills me more then me not being able to sequentially view my clips.
I do the same thing for my photos and still camera videos…. an example of the naming convention is like this:
This is only a general description and something that I have been slowly coming up with year after year… and, finally, I am more than satisfied!
More important than a system for organizing things is to make sure that the way you name files and even clips and sequences in fcp will not cause problems down the road. Clip and Sequence names are often used by fcp to name exported files and if you use illegal characters, this may come back
and bite you. Also, if you have any possibility of sending files to a Windows system this just makes it more important to be careful. Also, keep file, clip and sequence names relatively short.
Again fcp may add things to your file names (compressor also does this) and if you’re not careful, your filenames may become too long for some applications (I’ve seen bitvice truncate file names in a way that can become very confusing). Also, although I’m not sure about OS X, Windows needs to keep the total file name (INCLUDING the path) within a certain length. If a file gets placed too deep within the disk structure, it may become corrupted or difficult to access.
I work in a broadcast environment for GOD TV.
We don’t have a standard across the company yet, but will shortly be introducing a Digital Library System so this will become vital, so I’m looking forward to seeing your results!
Currently I would name a typical few clips shot in Africa of a children’s orphanage:
But what may be of interest is the way we name music. We import a track from CD via iTunes converting it to aiff onto the media drive. Then go into the drive, rename the track, and move it to the right folder (Fast, Interview Bed, African, etc).
So a track from Bruton’s CD called Widescreen Drama would be named as follows:
So the tracks called “Star Birth”, the catalogue number of the CD is BR441 and the track number is 2.
This enables us to track back when needed to the filing cabinet where the cd is stored under BR (Bruton) in numeric order, and find name of composer etc for the music cue sheets.
Tom is the senior editor for the JMS Group Ltd n Norfolk, UK, writes:
I was reading Anna B’s question on ‘Where to store files’ and thought I would share my filing system. Being extremely paranoid about keeping my data both organized and safe I’ve always operated the procedure of a single directory for each job – even though some may consider that placing all my eggs in one basket it is a basket I can mirror each evening to an external drive, or copy to my thumb drive to take home with me – all using a single drag and drop.
To begin with I create a folder on my RAID with the client and job name. I then create an FCP project, save it in the folder I just created, then I switch the scratch discs to also reference this folder (the only snag, but one I have got used to, is making sure I remember to change the scratch disc preferences each time I move between projects). Now that I have a client/job directory containing an FCP project, and the automatically created Capture Scratch and Render Files/Audio Render Files directories I go on to add a few extra folders of my own. These are: ‘AE’ for After Effects work relating to the edit – within this folder I have sub-directories named ‘Sources’ (very handy place for housing client-supplied artwork and logos etc) ‘Workspaces’ and ‘Renders’. I create a folder named ‘DVD’ in which I can place the .m2v files and DVD Studio projects for the inevitable final client copies of the job. Lastly I create an ‘Imported Audio’ folder – where I can place the music tracks, voiceover and final mix which have been produced elsewhere and sent to me as files. I am now working for the most part tapeless so don’t capture anything directly into the capture scratch directory – but I manually place my XDCAM footage into this folder, with the added benefit of being able to split it up into scenes/shoot days using the Finder/Quicklook – and then just use ‘Import Folder’ in FCP to import the pre-arranged footage bins.
I use an identical file system for every job – so I can go back to work I was doing three or four years ago and not have to refresh my mind where anything is. It makes archiving easier too – knowing 100% that everything relating to a project is in the one directory eliminates the possibility of ‘could not locate media’ warnings later on.
Hope that’s of help to a few people!
Larry replies: Thanks, Tom, for writing this up. Organization is SO critical to successful editing, I’m happy to share your system.
I have a real problem recommending changing scratch disks because if you forget even once, FCP starts storing files in places you don’t expect. FCP is just not designed to move scratch disks between projects. I’m hoping this gets fixed in a future version of FCP.
That being said, I agree with the rest of your ideas where you store all project files in a single folder, using subfolders to keep things organized.
Larry replies: Thanks to everyone for their suggestions. Feel free to send me additional comments and I’ll update this from time to time.
NEW & Updated!
Edit smarter with Larry’s latest training, all available in our store.