Data Volume

Data volume refers to the file size and storage requirements for your neural recordings. With modern high-channel-count probes and high sampling rates, data management becomes a significant practical consideration.

Calculating data volume: Data rate (MB/s) = Channels × Sampling rate (Hz) × Bytes per sample (typically 2)

Example: 128 channels × 30,000 Hz × 2 bytes = 7.68 MB/second = 27.6 GB/hour

Typical recording sizes:

• 32 channels, 30 kHz: 1.92 MB/s, ~7 GB/hour
• 64 channels, 30 kHz: 3.84 MB/s, ~14 GB/hour
• 128 channels, 30 kHz: 7.68 MB/s, ~28 GB/hour
• 384 channels, 30 kHz: 23 MB/s, ~83 GB/hour

Storage considerations:

• Multi-day experiments can generate terabytes of data
• Backup strategies are essential (data loss is catastrophic)
• Processing requires substantial RAM (often need to load entire recordings)
• Network transfer times can be prohibitive for large datasets

Compression strategies: Spike sorting reduces data volume 100-1000×, since you're storing spike times and waveforms rather than continuous voltage. Consider whether you need to archive raw data or just processed results.

Infrastructure planning: Budget for RAID arrays, high-capacity storage, fast network connections, and cloud storage for long-term archival before starting large-scale recording campaigns.