Why Advertised Transfer Speeds Never Match Reality

A drive or interface rated at 400 MB/s rarely delivers that in practice. Real-world transfers lose throughput to filesystem overhead, drive fragmentation, thermal throttling, and the sheer number of small files involved. This tool estimates transfer time using realistic numbers — not the marketing spec — by letting you dial in an inefficiency percentage, per-file overhead, and parallel stream count that match how you actually move data.

Real-World Inefficiency

Interface speeds like USB 3.0 or Thunderbolt describe the theoretical ceiling of the connection, not what you’ll see moving actual footage. Protocol overhead, drive write speed, and controller limitations typically eat 15–30% of the advertised number. The default 20% inefficiency in this tool reflects a reasonable real-world estimate, but drop it lower for fast NVMe-to-NVMe transfers or raise it for spinning disks and heavily fragmented drives.

File Count and Parallel Streams

Transferring one enormous file is very different from transferring a hundred thousand small ones — every file adds a small fixed overhead for opening, writing, and closing the handle, which adds up fast on camera card dumps full of small clips and sidecar files. Entering a file count and per-file overhead accounts for that. Parallel streams model tools that copy multiple files simultaneously, which can meaningfully speed up transfers of many small files even without more raw bandwidth.

This tool is especially useful for:

• DITs and data wranglers planning how long a card dump or backup pass will realistically take.
• Editors and post teams scheduling around large file transfers between drives or over a network.
• Anyone budgeting production time around footage offload before a hard call time.