Final Cut Pro Workflow Ideas
A couple of colleagues recently discussed with me problems using Final Cut Pro. One is a feature editor. The other does documentaries. I’ve most recently used FCP on an episodic series. Three different types of productions, but I think there are some common solutions. Solutions we’re planning on using on the second season of Warehouse 13 for the SyFy channel.
Project Organization:
Instead of one project for each episode, a common organization on Avid MC, we will have several separate projects for each episode: a master editing project, dailies project, backup project. FCP works poorly when projects get too large. On the other hand, it is easy to have several projects open at once.
And instead of adding music and sound effects into a master project, we will have a master sound effects project, and composer / score music project. We may also create a needle drop separate project. And we will have a project for stock footage. Each editor or assistant can access (read) any of these projects, just can’t write to them without drive writing access.
Storage Organization:
We will have several terabytes of storage that the four editors and three assistants will all need access to. Each of us need read access. But we only need limited write access, basically just to the project we are editing. The networking storage we’re using is not Avid’s Unity, which allows for changing partitions at any time. We will have to partition all the storage before we start work.
Our plan is to format the shared storage so there multiple partitions. As we will have 13 episodes, each episode will have its own partition. Additionally, each assistant will have a partition, as well as partitions for sound effects, music, stock footage and visual effects. About 20 partitions. The assistant or editor will gain write access only to that partition as needed. This scheme will depend on how well the storage software works.
One suggestion from Apple techs is to limit the use of subfolders, as they make drive access slower for FCP. We will likely keep all sound effects in one large folder, have all the composer’s music in one folder, and all needle drop in one folder. This may mean renaming some files to have a complete reference to their origin.
One advantage to having each episode have its own partition is rendering. In this scheme, each episode renders to its own partition, which should reduce render files from becoming unlinked. So anyone that needs to work on that project will know where to find all render files.
Work in Standard Definition:
We considered Pro-res, but the render times (a substantial problem in SD) are unworkable in HD. It is likely we will get Quicktime files for dailies – last season we worked with DVCPro-50. And mostly ACE editors are creating off-line edits. On-line editing involves very different considerations and tasks.
Titles:
If you make a title with the Boris title tool rather than the standard FCP tool, it does not need rendering (unless a drop shadow or other effect is added). Anything that reduces rendering is good.
Boris Continuum
Finally, we’ve requested Boris Continuum Complete to be added to each system, and to our on-line editor’s system. Using BCC on my current (Avid MC) project has been a revelation. It has a function called Pixel Chooser, which is available in most of its plug-ins. Pixel Chooser is very much like Animatte in Media Composer. Creating simple or complex mattes for effects is now very simple (Four and Eight Point Garbage mattes are terrible to work with). And Boris has greatly increased the usability of its plug-ins, to where I’d rather have it over Sapphire.
Project Organization:
Instead of one project for each episode, a common organization on Avid MC, we will have several separate projects for each episode: a master editing project, dailies project, backup project. FCP works poorly when projects get too large. On the other hand, it is easy to have several projects open at once.
And instead of adding music and sound effects into a master project, we will have a master sound effects project, and composer / score music project. We may also create a needle drop separate project. And we will have a project for stock footage. Each editor or assistant can access (read) any of these projects, just can’t write to them without drive writing access.
Storage Organization:
We will have several terabytes of storage that the four editors and three assistants will all need access to. Each of us need read access. But we only need limited write access, basically just to the project we are editing. The networking storage we’re using is not Avid’s Unity, which allows for changing partitions at any time. We will have to partition all the storage before we start work.
Our plan is to format the shared storage so there multiple partitions. As we will have 13 episodes, each episode will have its own partition. Additionally, each assistant will have a partition, as well as partitions for sound effects, music, stock footage and visual effects. About 20 partitions. The assistant or editor will gain write access only to that partition as needed. This scheme will depend on how well the storage software works.
One suggestion from Apple techs is to limit the use of subfolders, as they make drive access slower for FCP. We will likely keep all sound effects in one large folder, have all the composer’s music in one folder, and all needle drop in one folder. This may mean renaming some files to have a complete reference to their origin.
One advantage to having each episode have its own partition is rendering. In this scheme, each episode renders to its own partition, which should reduce render files from becoming unlinked. So anyone that needs to work on that project will know where to find all render files.
Work in Standard Definition:
We considered Pro-res, but the render times (a substantial problem in SD) are unworkable in HD. It is likely we will get Quicktime files for dailies – last season we worked with DVCPro-50. And mostly ACE editors are creating off-line edits. On-line editing involves very different considerations and tasks.
Titles:
If you make a title with the Boris title tool rather than the standard FCP tool, it does not need rendering (unless a drop shadow or other effect is added). Anything that reduces rendering is good.
Boris Continuum
Finally, we’ve requested Boris Continuum Complete to be added to each system, and to our on-line editor’s system. Using BCC on my current (Avid MC) project has been a revelation. It has a function called Pixel Chooser, which is available in most of its plug-ins. Pixel Chooser is very much like Animatte in Media Composer. Creating simple or complex mattes for effects is now very simple (Four and Eight Point Garbage mattes are terrible to work with). And Boris has greatly increased the usability of its plug-ins, to where I’d rather have it over Sapphire.
8 Comments:
When using FCP, working in multiple projects on shows like this essential. One thing to keep in mind however is that you'll be unable to use a find in bin command for any clip that lives in an outside project. Additionally, the master clip/affiliate clip relationship in FCP can get broken when moving clips and sequences from one project to another. It's a good idea to move sequences, clips and multi-clips in individual groupings. Move all of your clips, then all multi-clips and finally all sequences from one project to another rather than all at once. This is especially important when cutting anamorphic SD dallies and using multi-clips since the formatting of the multi-clips can often become corrupt otherwise.
I noticed a few comments that don't seem to match my FCP experience.
AFAIK, your editors and assistants cannot share projects even in a read-only mode. They can share the media created from those projects, but must each work with duplicate copies of the project file.
If more than one person makes changes to these projects, you will have to reconcile which is the correct version.
Second, Boris Title 3D does require rendering. All effects require rendering in FCP before any output to tape for review. Even the grey-green bar segments are not full-quality until rendered; however, they may play fine in realtime and could be output "live" via an I/O card for review copies, but NOT masters.
- Oliver
Thank you for sharing this workflow. Having assisted on a couple of FCP features myself, I found that it's not the amount of clips that increases the project size, but the amount (and length) of sequences in the project. Having all dailies, sfx, mx in separate projects creates also the problem that the match frame/find bin functions don't work (unless each clip is copied into the main editing project before being cut into the current cut).
I'm going to do a show cutting on FCP with 2 editors and 3 assistants (one assistant will be the HD digitizing and coordinating person): shoot 35mm, telecine MOS camera rolls to HDCAM, digitize in cutting room (using JH-3 deck) to ProRes HQ, sync dailies in cutting room, cut using NO shared storage. Each editor and assistant will have a FirmTek SeriTek/5PM hot-swappable drive array with the top 4 drives configured in 2 RAID 0 sets. The bottom slot will contain a single 1 TB drive with all the stock, music, and sound effects on it. The assistant running the digitizing station will copy each day's dailies onto a G-Raid portable drive which will be given to the appropriate assistant to sync and organize. At lunch time, those dailies will be copied to the editor's station. This system will use a lot of SATA drives, which will keep the dailies in three places. SATA drives are cheap. Shared storage is VERY expensive.
13 volumes sounds like a logistic PITA. I can see where you are coming from with the render files. We don't have these issues (much) working with XSAN since it's a file-based permission scheme. I can only guess this is a workaround for a volume-based scheme. You'll want to also keep in mind that Apple suggests having one gigabyte of RAM for each volume your server maintains. If you're following their advice already, you kind of have to go whole hog. That means your server will need 16GB of RAM for this to work. I would highly suggest you talk with your SAN (or otherwise shared storage) integration company rather than with Apple about these matters. Editshare, I'm fairly confident, has better ways of dealing with this sort of workflow.
THe only thing I may add is that we use a networked drive to store our projects. This way any editor can open up and write to project. IF they make changes, they usually date it or add initials to it to alert others.
I am also trying out software called FCP versioner..which can keep track of changes to a project by xml, making revisiting older versions much easier in terms of finding the correct one.
As a Final Cut Studio Master Trainer and technical editor for the book Final Cut Pro for Avid Editors, I've specialized in training and consulting for facilities switching from Avid to Final Cut Pro. As such, I've been dealing with workflow complaints from Avid editors since Final Cut Pro came out, and recently developed a solution to address those complaints, one that will allow Final Cut Editors to follow a workflow that is almost identical to Project Sharing on Avid Unity. Unlike EditShare, it requires no proprietary hardware or software, and works on systems like Unity and Xsan, which allow file-level permissions.
(A demo of the solution can be seen at http://www.editdog.net/Final_Cut_Synergy/Demo.html)
It gets around the major pitfalls of collaborative workflow in Final Cut Pro, such as users saving over each other's work, tracking project versions, bloated project size and more. It's not the only FCS collaborative workflow out there, but it is the only one that mimics precisely the experience of sharing projects on Avid Unity.
Telecine to HD tape? Guys, this is 2010.
How about "telecine" to prores quicktime files with synced sound, direct import into FCP.
You just eliminated the need for 1 1/2 Assistants, and will have much better overall workflow.
Cost? Figure out the cost of the Assistants, the HD Deck Rental, and all the tape you're going to use. I think you'll quickly realize that this is a much cheaper way to go.
Tape is dead.
Post a Comment
<< Home