NeuroNexus · neuronexus.com · Glossary
Data Volume
Data volume is the size of a recording: channels multiplied by sampling rate multiplied by bytes per sample, which at 16 bits is 2, so 128 channels at 30 kHz produce 7.7 MB per second and about 28 GB per hour.
The rule scales linearly: 32 channels at 30 kHz is 7 GB an hour, 64 is 14, 1,024 SiNAPS channels at 20 kHz is 41 MB per second and about 1.2 TB for an eight-hour session. Multi-day chronic studies run to terabytes.
Volume is decided when the electrode is, not when the disk fills. Site count sets the channel count and the sampling rate is fixed by the signal, so the storage, backup and transfer budget for a study follows from the probe name. Spike sorting compresses a recording a hundred- to a thousand-fold by keeping spike times and waveforms rather than continuous voltage, which is the reason to decide before the study whether raw data is archived or only the sorted result; the high-density planning guide works through the budget. Radiens is local-first for this reason: the data stay on the laboratory's storage, and what moves over the network is views and summaries rather than the recording.
Typical values
| Formula | channels × rate × 2 bytes (16-bit) |
|---|---|
| 32 ch at 30 kHz | about 7 GB per hour |
| 128 ch at 30 kHz | 7.7 MB/s; about 28 GB per hour |
| 1,024 ch at 20 kHz | 41 MB/s; about 1.2 TB per 8 h |
Cite this term
NeuroNexus. "Data Volume." NeuroNexus Glossary, updated 2026-10-12. https://www.neuronexus.com/resources/glossary/data-volume