How Do Media Storage Tiers Work in Post Production?
With video production projects ranging in size and scale, you can’t address media storage needs with a one-size-fits-all approach. Let’s say you’re an editor at a boutique post house or a producer on a corporate video production team, and you’re planning a big new project—one featuring dozens of interviews, hours of illustrative footage, and some fancy motion graphics to boot. It’s also going to shoot on expensive, high-end cameras delivering massive files. You’re on a tight deadline, so you’ll need multiple editors working on it consistently.
Do you rent or buy storage? Which media storage tiers do you need to make it all happen, and how much of each tier do you need? Getting through a big project like this means artfully leveraging different media storage tiers throughout the various stages of post production.
Three Tiers of Media Storage
There are three main tiers of media storage:
- Online. This is fast, high-performance media storage for real-time workflows. It tends to be more expensive and therefore usually has limited capacity. (Think fast SSD storage.)
- Nearline. Nearline media storage is slower but has a larger capacity—media is still readily accessible, but the storage doesn’t deliver the same level of performance (nor the same level of expense) as online media storage.
- Archive. Archives mean inexpensive, large-capacity storage that is typically inaccessible without some level of human interaction (such as loading an LTO tape). This is also often called offline storage.
You’ll also see hybrid cloud storage setups, which store some footage in the cloud rather than on-prem. (And in today’s remote world, more houses are turning to the cloud for the accessibility it offers.) The cloud can play a role in any media storage tier, but it is most commonly used for nearline and archival storage, since online workflows require quite high bandwidth to meet performance requirements.
Tiered Storage and Editorial Workflows
Some editorial workflows are online at all times. This is especially true of projects working with smaller amounts of media and tighter turnarounds—all high-resolution media stays constantly available on a fast online storage tier until the end of the project, when the completed edit will move to archive storage to make room for the next project.
But on a big project with many hours of footage, budget limitations or editorial requirements can eliminate that possibility. This is usually when a more traditional proxy workflow comes into play. Here, the online storage tier hosts the creation and storage of lower-resolution proxy files, while the original camera media with high resolutions (and large file sizes) are stored on a nearline tier for later use.
When the final edit is locked, the timeline can then be “onlined,” wherein only media that made it into the final cut moves from the nearline tier to the online media tier. This replaces proxy media with high-resolution original camera files, which then can be used for final color correction, motion graphics, and delivery. In those later stages, more powerful performance helps keep a real-time workflow ticking along. Once everything is delivered, the entire project can then be safely stored on the archive media tier.
Planning for each tier requires balancing three vital capabilities: performance, scalability, and reliability.
Performance—How Fast Do You Need to Go?
Your media storage should deliver the level of performance required to play back the files you’re working with, in real time, to however many editors need access at the same time. Although capacity might spring to mind first—after all, you need enough space to fit everything—there’s no point in having terabytes to spare if it’s all too slow to play back the media you have at the right speed.
To start: how many editors will need to access the storage at the same time? And how many streams (or layers or feeds) of video will each editor need to play back simultaneously? Here’s where a bit of math helps.
Editors x Video Streams x Bit Rate of each stream = required throughput speed (Performance), or E x VS x BR = P
For example, two editors cutting multicamera shows with, say, four cameras each will need to play back eight streams of video at the same time from shared media storage. Each of those video streams will be stored in a specific codec (flavor of video, e.g., ProRes, DNxHD, H.264 etc.) and use a specific bit rate (bits per second). Generally, the higher the bit rate, the better quality the video and the bigger the file size.
To layer onto the example, the bit rate for 1920 x 1080 HD ProRes video at 29.97 frames per second is 147 Mb/s (megabits per second), so:
2 editors x 4 streams X 147 Mb/s ProRes = 1,176 Mb/s. Because megabits (Mb) are 1/8th of a megabyte (MB), this math happens to work out to needing to transfer 147MB/s.
Check your own figures, then look at the online media storage in question. Can it deliver—at a minimum—this level of performance? If it can’t, choppy, stuttering playback may threaten to hamper creativity and productivity, both of which most teams can’t afford to lose.
Scalability—How Much Room Do You Need to Grow?
Capacity will be an important consideration across every tier of your storage. Still, higher-performance online storage is going to be more expensive, which can limit exactly how much of it you have to work with.
Also consider that needs change. Will your tiered media storage allow you to expand its capacity?
If you’re making a substantial investment in a shared storage solution, it should be able to expand over the working lifetime of the hardware. Check whether you’ll be able to reconfigure the storage yourself (such as when adding more capacity) and whether any downtime is acceptable during that process. (These types of technological hurdles can make it worth investing in a media storage solution that comes with reliable technical support.)
Reliability—How Safe Is Your Media from Sudden Disaster?
Keeping your media safe has to involve incorporating a level of redundancy into any storage calculation you make.
Storage systems typically accomplish this with RAID (redundant array of individual disks). This means that, spread across the entire array of individual drives that make up a complete storage solution, each drive will keep a partial backup of another drive. This eats up some of the initial total capacity—however, if one drive fails, the other drives can rebuild the contents of that failed drive from their duplicates. Another layer of redundancy is to have a complete off-site backup in case of fire or theft, sometimes being drip-fed to the cloud. These kinds of backup measures protect not only your media but your budgets (and potentially reputation) in the case that disaster strikes.
There’s always a bit of pain involved in large investments, but take comfort in the fact that unpredictable disasters and other complexities can blow out a budget just as easily, and to much greater damage. A small post house choosing to save money by rolling their own tiered media storage is responsible for the time-consuming tasks of wrangling media between systems and maintaining different technology from a variety of suppliers. And if the system fails, they may still lose valuable client media. A well-thought-out plan for shared media storage helps take a lot of that stress out of the equation.
Jonny Elwyn
Jonny Elwyn is a freelance film editor and writer from London and the author of How to Be a Freelance Creative.