BRAW Data Rate Guide: MB/s, Mbps and Drive Speed

BRAW data rate is the amount of recorded data produced each second. It determines three things at once: whether recording media can sustain the stream, how quickly storage fills, and how much throughput an offload or editing system needs.

The first task is to read the unit correctly. Camera storage tables often use MB/s, megabytes per second. Network and video-delivery discussions often use Mbps, megabits per second. One byte contains eight bits.

1 MB/s = 8 Mbps

A 200MB/s BRAW stream is 1,600Mbps, or 1.6Gbps before protocol overhead. Confusing the two units creates an eightfold planning error.

What sets the BRAW rate

Four inputs dominate:

• Pixel dimensions: more pixels per frame increase the baseline.
• Frame rate: more frames per second increase data approximately in proportion.
• Compression mode: 3:1 records more than 5:1, 8:1, or 12:1.
• Image complexity in Q modes: Q0–Q5 can vary with detail, noise, and motion.

Camera implementation and sensor window also matter. “6K” may mean several different pixel dimensions. Use the exact recording format, not a marketing label, when choosing media.

Data rate versus file size

Data rate is a speed. File size is the amount accumulated over time.

At 150MB/s:

• One second is about 150MB.
• One minute is about 9GB.
• One hour is about 540GB.

The storage formula uses decimal units:

GB = MB/s × seconds ÷ 1,000

If the rate doubles, both the file size and media throughput requirement double. This is why a single MB/s value connects acquisition and storage planning.

Published rate example

For 6144 × 3456 at 30fps, Blackmagic Design has published rates around 323MB/s for 3:1, 194MB/s for 5:1, 121MB/s for 8:1, and 81MB/s for 12:1 on relevant camera specification pages.

Those become approximately:

These figures are useful only for the specific format and frame rate they describe. At 24fps, a proportional estimate is 80% of the 30fps figure. At 60fps, it is about twice the 30fps figure, subject to supported camera modes and implementation.

Why advertised drive speed is not enough

A storage product may advertise “up to 1,000MB/s” but fail a 300MB/s camera stream. The headline can represent a short sequential burst under ideal conditions. Recording requires sustained performance over the full take.

Performance can fall because of:

• A small fast cache becoming full
• Thermal throttling
• A nearly full device
• A slow flash configuration on a lower-capacity model
• Cable, enclosure, or controller limitations
• Background garbage collection
• Power or compatibility behavior in the camera

The camera manufacturer’s approved-media list tests more than arithmetic. A calculated speed target is useful for screening, but it is not a compatibility certificate.

Add headroom to the peak

BRAW Planner takes the maximum estimated rate, multiplies it by 1.25, and rounds upward to the next 10MB/s. A 194MB/s stream produces a planning target of roughly 250MB/s or more.

Why 25%? It gives a visible engineering margin for normal variability and avoids selecting a device whose rating exactly matches the stream. It does not cover every failure mode. A drive with an “up to 250MB/s” box label is not necessarily safe for a 194MB/s camera stream.

For Constant Quality, use the upper end of the range. The average rate may be much lower, but a difficult frame sequence can push the media harder. Dropped frames occur at the peak, not at the average printed on a post-production report.

Interface bandwidth is a ceiling, not a guarantee

USB, Thunderbolt, CFast, SD, and CFexpress standards describe an interface or performance class. The storage device, controller, camera, cable, and protocol overhead determine the usable sustained result.

For example, an interface with a 5Gbps theoretical ceiling does not deliver 625MB/s of guaranteed application data. Encoding overhead, protocol traffic, controller behavior, and device limits reduce usable throughput. The drive may be the slowest component even when the interface is fast.

Use measured sustained write performance in the real camera whenever possible.

Data rate in offload planning

Recording speed and offload speed are different paths. A 200MB/s camera stream creates 720GB in an hour. To copy that hour of footage in 30 minutes, the end-to-end copy must average at least 400MB/s before verification overhead.

The slowest component controls the copy:

• Source card and reader
• Cable and interface
• Computer port or hub
• Destination drive
• Simultaneous copy strategy
• Checksum and filesystem overhead

Copying to two destinations sequentially doubles the destination work. Copying simultaneously may split source or interface bandwidth. A proper test copies a representative full card and includes verification.

Editing throughput and multiple streams

An editor may play one BRAW stream smoothly but struggle with four-camera multicam. Approximate the storage demand by multiplying per-stream rate by simultaneous streams, then add room for audio, cache, effects, and operating-system activity.

Decode performance can become the limit before storage. Resolution, CPU/GPU support, debayer quality, noise reduction, grades, and effects all matter. Proxies reduce both decode and storage demand, but they add generation time and capacity.

A practical media-speed workflow

1. Find the exact camera format and published MB/s rate.
2. For Q modes, identify a credible peak or range.
3. Add at least 25% throughput headroom.
4. Check the manufacturer’s current approved-media list.
5. Use the exact capacity model—different sizes can use different flash layouts.
6. Fill-test long enough to exceed any fast cache.
7. Repeat at the highest resolution, fps, and scene complexity expected.
8. Check for dropped-frame warnings and verify every recorded clip.
9. Test again after major camera, media, or enclosure firmware changes.

The BRAW data-rate calculator gives a transparent first target. The real camera stress test turns that target into confidence.

Sources and calculation notes

• Blackmagic Pocket Cinema Camera technical specifications: camera-specific formats and storage-rate reference tables.
• Blackmagic RAW overview: manufacturer information about the codec and encoding modes.
• BRAW Planner methodology: the estimation formulas, unit conventions and limitations used on this site.

Workflow suggestions in this guide are planning guidance, not a claim that a particular camera or drive has been tested by the author.