Shooting with hybrid DSLRs has many advantages, but without a proven workflow one can quickly become lost beneath an avalanche of data. Worse still, data itself could become lost. Here we'll explore some workflow basics as practiced by the pros.
HDSLR cameras are taking the digital video production world by storm. Cameras like the Canon EOS 5D Mark II video DSLR have large sensors producing gorgeous high definition images with incredibly shallow depths of field and interchangeable lenses. It's easy to see why the Canon EOS 5D Mark II video camera and others like it have become so popular in such a short period of time. Of course they do have their challenges. Their physical form is not particularly conducive to hand-held work so they must be attached to a tripod or some other stabilizing device. Their audio capabilities as a professional video camera are, well, awful. And many of them leave much to be desired in terms of LCD screen resolution, focusing aids, exposure assistance and more.
Fortunately, for each of these shortcomings there is an answer, even several. And the pros of working with these cameras are great enough that it is often worth the added expense required (for the professional at least) to purchase the extra gear needed to overcome the cons.
Whether you're a weekend shooter or a professional DSLR user, fully equipped and heavily funded with the highest end professional video cameras and gear, or you've just pulled your sub-$1000 camera and stock lens out of the box, one thing you simply must do, in order to save yourself a world of potential grief on your next digital video production, is follow a solid, proven, workflow. Here are a few shooting with DSLR tips that might help you accelerate and refine that workflow.
In its simplest sense, the term workflow refers to a repeatable sequence of steps performed to get from state A to state B. For our purposes we'll use workflow to describe each of the steps taken from the moment we press record on the camera until we close and lock the door on the "archive vault." A workflow is designed to assure optimal efficiency throughout the production and post-production process and to guarantee, to the highest degree possible, the security of the data acquired.
A workflow should be repeatable to ensure that it will become automatic over time for all persons involved. It should also be scalable so it can be adapted to the size of the operation. A simple, one person, single camera shoot and a fully crewed, multi-camera feature production are vastly different in terms of scale but have two things very much in common: the need to be efficient, and, most importantly, data security.
Time lost due to misplaced, mishandled or lost footage results in a serious reduction in efficiency. But imagine completing a week-long, multi-camera shoot, only to discover in the edit bay that the media card containing the best takes from the best camera angle was inadvertently recorded over because it wasn't handled properly. Perhaps worse, what if this were to happen on a single-camera shoot with no coverage whatsoever from other angles? If you're out shooting some kids and their skateboard acrobatics at the local skate park it might not be such a big loss but what if it were your daughter's graduation - or wedding? If a commercial project, how do you explain to your client that you lost the key footage and will have to reshoot? The dog ate my media card probably won't fly any better today than the excuse did in school. The message here is that by following a solid workflow, and following it fanatically, you should never have to face any of these devastatingly nightmarish situations.
Cradle to Grave
Once the record button is pressed, birth is given to an HDSLR digital asset. In its rawest form, it must be protected from corruption and loss. The workflow process is designed to ensure its safe journey throughout gestation, delivery, and finally, taking its place in digital asset heaven - the archive room. It is transferred from camera to card then from card to computer. Here it is cloned; multiple copies to ensure its survival should disaster strike. Once the clones are sent on their respective journeys it's off to Post, where it will be cut, rolled, time remapped, color corrected, blended with others and compressed. In this final state, more clones will be produced with some being delivered to anxiously awaiting clients and others sent to join their raw, primal brothers in the archive room.
Dangers lurk in every corner and care must be taken to prevent the loss of data. One of the first steps to take is in the area of proper media management. This becomes especially critical on multi-day, multi-camera shoots. A single media card is one thing but trying to manage a dozen or more media cards is another disaster-waiting-to-happen altogether. Someone must be responsible for the proper handling of all media. Not several people - one. Having just one individual in the role of digital media manager will avoid confusion. If you're a one-person show, then it's pretty evident who that person is going to be.
One key task is to develop a naming system that will enable you to keep your media cards and their digital assets straight and intact all the way through the process. Start by labeling your cameras externally with tape: A, B, C, etc. Next, the camera is labeled internally. Place a card in Camera A and take a still shot. Remove the card from the camera and connect it to the computer via a card reader. Find the image you just captured in the card's DCIM sub-folder and rename it using the numbers 0999. Now, the next images captured on this camera will be in the 1000 series. Repeating the process for other cameras make Camera B the 2000 series, Camera C the 3000 series and so on. When developing a naming scheme, be sure to check your camera's manual for its accepted file format. Set file numbering to continuous to ensure that file numbers continue in sequence when changing cards rather than starting over with each new card.
When shooting, as cards are filled, remove them from their respective cameras and label them with tape; i.e. the first card out of Camera A is labeled A1; the third card out of Camera C is labeled C3, etc. Once filled and labeled, the cards are placed in a protective case, flipped over, backside up, making it obvious that these have yet to be downloaded onto the computer and copied to backup drives. The case goes to the digital media manager who knows at a glance, which cards need to be downloaded, and which camera they came from. The media manager loads the cards into the reader and downloads the images to the hard drive(s). Before sending the card back, the media manager renames the DCIM subfolder as DONE so the camera will read it as empty. The tape label is removed and the card is placed back into the protective case, right side up, to show that it is ready to be recorded onto once again, and sent back to the field. When this card is placed in the camera for formatting it will show that there is data on the card. When play is pressed, however, the camera will indicate that no image is present, confirming that the card has indeed been downloaded by the media manager and is okay to format and re-use. Consider protecting your data by making several copies on different hard drives. These drives can then be distributed to the director and editorial. One hard drive should be kept as a backup and the one should go into a vault for safe-keeping.
Chase Jarvis is a well-known, professional DSLR user. He shares his complete workflow on his website, www.chasejarvis.com. Immediately after recording, cards are given to the media manager (also known as the Digital Asset Manager) for ingestion. With the card in the card reader, data is copied simultaneously to two daisy chained hard drives. In the field, all files are then run through Compressor and output as ProRes 422 files for an even starting point back in the studio.
Each day after shooting the crew returns to base camp where data from the field drive is copied to a 2TB RAID 1 drive. This creates two 1TB drives that are exact copies of one another. Each drive then goes to a different location for safety. If something happens to one, the other is readily available to continue working. Next, initial editing begins with the shots fresh in mind; the best shots are noted now so they can be worked on later.
Back in the studio all data is copied to a server (and backup drive) which runs a 16 drive storage array, set up in such a way that should any single drive fail, the remaining drives contain all the information necessary to restore all the data onto the replacement drive. Chase makes a critical distinction between raw data and worked data. Raw data is untouched; unchanged from its original form and is considered sacred. Any data whose pixels have been manipulated in any way however, is saved separately on the server as Live Work. When raw data comes in it is saved on the server. Then, at the workstation, it is manipulated using Photoshop, Final Cut, Compressor or other software, then saved to the "Live Work" folder on the server before final delivery to the client. Both raw data and live data are backed up and stored offsite in a secure location. Additionally, each workstation is also backed up hourly to an external hard drive.
Another important consideration in the plan to protect your data is the media itself. Paul Hildebrandt, of Eventide Visuals in Northern California, highly recommends using fast media cards: 133X or better for Compact Flash memory cards and Class 10 or higher for SD cards. He suggests that you "never go below these speeds" to minimize the possibility of potential data loss. Paul also recommends using card capacities of not less than 16GB as anything less is not really practical for field use. If your situation allows you time to frequently switch out smaller cards then they'll work fine but as a wedding and event videographer he points out how critical it is "to get through a ceremony without switching cards."
It's well known that the audio capabilities of HDSLR cameras are severely lacking. Fortunately there are a number of satisfactory solutions out there. If your camera is equipped with a 1/8-inch microphone jack you could simply plug in an external microphone with the proper connector. If you want to use a professional mic with XLR connectors you'll have to use one of several adapters that are available.
If your camera lacks a microphone jack altogether you can still capture great audio with a quality recorder such as Zoom's H1, H2 or H4 or the recorders by Tascam, Sony or Yamaha, among others. Recording with these devices requires another step in the post production process: synching the audio with the video. You can use the tried-and-true method of using a slate, clapboard or simply clapping your hands loudly in front of the camera but then you're faced with the tedious task of synching audio and video manually. Alternatively, many video producers are turning to PluralEyes software from Singular Software to synch up their HDSLR footage with separately recorded audio. For an in-depth look at these solutions check out the article entitled How to Get the Best Audio From HDSLR Cameras.
For the professional, time - and data - is money. Lost data means lost income. For the non-professional, memories fade, and captured images, once lost, can never be replaced. Employing a consistent, secure workflow, from capture to archive, can help prevent these, and other mishaps, from ever happening.
Sidebar: RAID Types: Pros and Cons
RAID stands for Redundant Array of Independent Disks and refers to a data storage system utilizing multiple hard drives. There are a number of RAID configurations but three in particular are of use to the video producer.
RAID 1 is a mirrored array meaning that it writes to two discs at the same time, creating two drives that are identical to one another. This arrangement requires only two hard drives and provides complete redundancy; if one drive fails, the other drive continues to house all the data. On the downside, it is slow and results in a loss of half your drive capacity.
RAID 0 is a striped array. In this configuration data is written to both drives simultaneously but in only one location. This results in very fast write speeds, twice the speed of a single drive, with no loss of drive capacity, but with no redundancy. If one drive dies everything is lost. This arrangement is useful as scratch discs, such as when editing video, where speed is a plus but the data is temporary and a loss isn't critical.
RAID 5 requires a minimum of three hard drives but in practice can be very large with moderate drive speed results. This arrangement provides complete redundancy while only giving up about 25 percent of total drive capacity. RAID 5 setups require a hardware controller and are expensive but can be built to provide a large, fully redundant storage solution.
Contributing editor Mark Holder is a video producer and trainer.