Introduction
The post house ordered a new storage server for color grading. Eight 18 TB drives in RAID 5 should give 144 TB of raw capacity, right? Not exactly. RAID 5 with 8 drives yields 126 TB usable because one drive's worth of capacity is consumed by parity data. The facility manager who ordered based on raw capacity was 18 TB short on day one. Factor in the binary-to-decimal conversion (TiB vs TB) and the actual formatted capacity drops further to about 114 TiB.
The RAID configuration calculator shows you exactly how much usable storage you get after redundancy for any combination of RAID level, drive count, and drive size. It compares every RAID level side by side so you can balance capacity, performance, and fault tolerance for your specific workflow.
This is the tool you use before you order drives, not after you discover you're 20% short on a deadline.
What This Tool Calculates
The calculator takes three inputs: RAID level (0, 1, 5, 6, 10, 50, or JBOD), number of drives, and individual drive size in GB, TB, or PB.
It returns the raw capacity, usable capacity after redundancy, redundancy overhead, and storage efficiency percentage. It also shows RAID properties including fault tolerance (how many drives can fail), read and write performance characteristics, and a visual capacity bar. A comparison table calculates every compatible RAID level for your drive configuration side by side. Cost estimates for HDD, SSD, and NVMe at current per-TB rates round out the planning data.
The Formula and How It Works
Each RAID level dedicates a different amount of capacity to redundancy:
RAID 0 (Stripe): Usable = N * Drive Size. No redundancy. All capacity available. One drive failure destroys all data.
RAID 1 (Mirror): Usable = (N / 2) * Drive Size. Every drive is mirrored. 50% capacity used for redundancy.
RAID 5 (Stripe + Parity): Usable = (N - 1) * Drive Size. One drive's worth of distributed parity. Survives one drive failure.
RAID 6 (Double Parity): Usable = (N - 2) * Drive Size. Two drives of parity. Survives two simultaneous failures.
RAID 10 (Mirror + Stripe): Usable = (N / 2) * Drive Size. Mirrored pairs, then striped. Best performance with redundancy.
Worked example: 8 drives at 18 TB each. RAID 5: (8-1) * 18 = 126 TB usable (87.5% efficiency). RAID 6: (8-2) * 18 = 108 TB (75%). RAID 10: (8/2) * 18 = 72 TB (50%). RAID 0: 8 * 18 = 144 TB (100%, no protection).
Important: drive manufacturers use decimal TB (1 TB = 1,000,000,000,000 bytes) while operating systems display binary TiB (1 TiB = 1,099,511,627,776 bytes). An 18 TB drive shows as approximately 16.37 TiB in your OS. The calculator shows both values.
Real-World Examples
Post House Color Grading Server
A color grading facility needed a shared storage server for DaVinci Resolve with at least 100 TB of usable space and single-drive fault tolerance. They compared RAID 5 and RAID 6 with 8 x 20 TB drives. RAID 5 gave 140 TB usable (87.5% efficiency), while RAID 6 gave 120 TB (75%). Both exceeded the 100 TB requirement. They chose RAID 6 for the extra safety margin: during a rebuild after a single drive failure, a second failure in RAID 5 would destroy everything, while RAID 6 survives two failures. The 20 TB overhead was worth the insurance.
DIT Cart Scratch Storage
A DIT building a portable offload cart needed maximum speed for real-time playback of 8K RED footage. Performance trumped redundancy since the cart was a temporary scratch space with backups elsewhere. The calculator showed that RAID 0 with 4 x 4 TB NVMe drives delivered 16 TB at roughly 4x single-drive speed, enough for real-time 8K playback. The DIT used the backup strategy planner separately to ensure the 3-2-1 rule was covered by the shuttle drives and cloud upload.
Archive NAS for Production Company
A production company building a Synology NAS for project archives wanted maximum capacity with basic protection. With 12 x 16 TB drives, RAID 5 gave 176 TB usable while RAID 6 gave 160 TB. They chose RAID 6 because with 16 TB drives, the rebuild time after a failure is 24+ hours, and the probability of a second failure during rebuild is significant with large drives. The calculator showed the 16 TB difference (one extra drive of overhead) was a small price for avoiding catastrophic data loss during rebuild.
RAID Level Comparison (8 x 18 TB Drives)
| RAID Level | Usable Capacity | Efficiency | Fault Tolerance | Best For |
|---|---|---|---|---|
| RAID 0 | 144 TB | 100% | None | Scratch/temp, max speed |
| RAID 1 | 72 TB | 50% | 1 per pair | OS drives, critical boot |
| RAID 5 | 126 TB | 87.5% | 1 drive | General production storage |
| RAID 6 | 108 TB | 75% | 2 drives | Large drives, archive |
| RAID 10 | 72 TB | 50% | 1 per pair | High-perf editing |
| RAID 50 | 108 TB | 75% | 1 per sub-array | Large arrays, balanced |
| JBOD | 144 TB | 100% | Partial loss | Simple concatenation |
Pro Tips and Common Mistakes
Pro Tips
- For drives larger than 8 TB, prefer RAID 6 over RAID 5. Larger drives take longer to rebuild after failure (24-48 hours for a 16-20 TB drive), and the probability of a second drive failing during rebuild increases with drive size and array age. RAID 6 survives two simultaneous failures.
- RAID is not a backup. RAID protects against drive failure, not against accidental deletion, ransomware, file corruption, or theft. You still need a separate backup strategy following the 3-2-1 rule.
- Match the RAID level to the workflow. RAID 0 for DIT scratch (speed, no redundancy needed). RAID 5/6 for shared project storage (balanced). RAID 10 for high-performance editing bays (speed + redundancy). RAID 1 for OS and application drives.
- When budgeting, remember that RAID overhead means you are paying for drives you cannot use for data. A RAID 6 array with 8 x 18 TB drives costs the same as 144 TB of raw capacity but delivers only 108 TB usable. Factor this into your cost-per-usable-TB calculation.
Common Mistakes
- Ordering drives based on raw capacity without accounting for RAID overhead. A RAID 6 array with 8 x 18 TB drives delivers 108 TB, not 144 TB. That 36 TB difference can derail a project if storage needs were calculated from raw numbers.
- Using RAID 5 with very large drives (16 TB+). Rebuild times for large drives can exceed 24 hours, during which a second failure destroys the entire array. RAID 6 adds one extra drive of overhead but survives two failures.
- Confusing TB (terabytes, decimal) with TiB (tebibytes, binary). An 18 TB drive shows as approximately 16.37 TiB in your operating system. A RAID 5 array of 8 x 18 TB drives shows 126 TB on paper but approximately 114.6 TiB in your OS. Always check which unit your planning tools use.
Frequently Asked Questions
Which RAID level is best for video editing?
For active editing with real-time playback requirements, RAID 10 provides the best combination of speed and redundancy. For shared storage servers where capacity matters more than maximum speed, RAID 5 or RAID 6 is more efficient. For temporary scratch storage where speed is paramount and data is backed up elsewhere, RAID 0 provides maximum performance and capacity.
How many drives can fail before I lose data?
RAID 0 and JBOD: zero (any failure causes data loss). RAID 1: one drive per mirror pair. RAID 5: one drive total. RAID 6: two drives total. RAID 10: one drive per mirror pair. The specific drives that can fail matter in RAID 10; losing both drives in the same mirror pair causes data loss even if other pairs are intact.
Should I use hardware or software RAID?
Modern software RAID (like ZFS, Synology SHR, or Apple RAID) is reliable and often more flexible than hardware RAID controllers. Hardware RAID provides a slight performance advantage for write-heavy workloads but ties you to a specific controller. If the controller fails, you need an identical replacement. Software RAID arrays can be moved between systems more easily.
What is the difference between RAID 5 and RAID 6?
RAID 5 uses single parity (one drive's worth of redundancy) and survives one drive failure. RAID 6 uses double parity (two drives' worth) and survives two simultaneous failures. RAID 6 costs one extra drive of capacity but provides critical safety during rebuild operations, especially with large drives where rebuilds take 24+ hours.
Start Calculating
RAID configuration is the foundation of every post-production storage system. Getting the balance between capacity, speed, and redundancy right at the planning stage prevents both storage shortfalls and unnecessary spending on drives that sit idle as parity overhead.
Enter your drive configuration above and compare every RAID level side by side. What RAID level does your facility use for primary project storage, and have you had to recover from a drive failure?
