EMZETT.
Login

Recording

In short: Capturing audio, video or screen content into a file, so it can be played back or processed later.

In more detail: Depending on the source, different hardware or software is needed: microphone and camera for sound/image, screen-capture software for screen content, a capture card for external video signals. The recording is usually saved directly in a compressed format (e.g. MPEG) to save storage space.

In Depth

Quality vs. file size

Every recording involves a trade-off between quality and file size: uncompressed video or audio material is lossless but enormously large (a few minutes of uncompressed 4K video can take up several gigabytes), which is why practically all recording software compresses directly while recording — usually with lossy codecs like MPEG-4/H.264 (widespread, good compatibility) or the more efficient but more computationally demanding H.265/HEVC for video, and AAC/MP3 for audio. How strongly it’s compressed (the bitrate, measured in Mbit/s) directly determines the later image quality and file size — a higher bitrate means fewer visible compression artefacts (blockiness, blur in moving scenes), but proportionally larger files.

Software vs. hardware encoding

For screen recordings, a further distinction is made between software encoding (the CPU handles compression via software, flexible and often somewhat better quality at the same bitrate, but computationally demanding) and hardware encoding (a dedicated, built-in unit in the graphics card or CPU handles it, e.g. NVENC on Nvidia or Quick Sync on Intel chips). Hardware encoding loads the running system considerably less, which is especially relevant when gaming and recording at the same time (streaming), since actual game performance would otherwise noticeably drop if the CPU also had to compress the video in parallel.

Audio recording: sample rate and bit depth

For audio recordings, the sample rate (e.g. 44.1 kHz — CD quality — or 48 kHz, the standard for video/film) and the bit depth (e.g. 16-bit or 24-bit) additionally play a role, together determining how finely the analogue signal of the microphone is broken down into digital values. A higher sample rate captures higher frequencies more accurately, a higher bit depth allows a larger dynamic range (the difference between the quietest and loudest representable sounds) and more room for later volume editing without noise or distortion becoming audible.

Latency in live recordings

With live transmissions (streaming, video conferencing), a further requirement is added: low latency, i.e. as short a delay as possible between recording and playback at the receiver. Highly compressing codecs with high compression efficiency often need more computing time and thus more latency — for interactive applications (video calling), a compromise is therefore often chosen in favour of lower latency rather than maximum compression efficiency.

Container format vs. codec

A common misconception is confusing the container format (e.g. .mp4, .mkv, .mov) with the codec (e.g. H.264, H.265, AV1): the container is just the “packaging” that holds video, audio and metadata tracks together, while the codec determines how the actual image/sound data is compressed. The same .mp4 file can vary considerably in size and compatibility depending on the codec used — older playback devices often only support H.264, while H.265 compresses more efficiently (smaller files at the same quality) but needs more computing power to decode and isn’t supported everywhere. .mkv as a container also allows several audio and subtitle tracks at once, which was historically more limited with .mp4.

Storage location: local vs. cloud/streaming

Recordings can either be stored locally on mass storage (SSD/HDD) or streamed directly, without ever being fully stored locally (livestreaming to a platform). With local high-bitrate recordings, the write speed of the mass storage is an often underestimated bottleneck — with very high-resolution, lightly compressed recordings (e.g. professional video production), even a fast SSD can hit its limits, which is why deliberately faster but more expensive RAID arrays of several SSDs are sometimes used there, to guarantee the necessary sustained write rate.

See also: Capture, Microphone, Camera, Capture Card