Choosing what a camera records in is a storage decision as much as a quality one. The codec sets shooting time, transfer time and how the material behaves in an edit.
Two families with different purposes
Delivery codecs compress heavily by describing how each frame differs from its neighbors. They produce small files and are what viewers ultimately receive.
Editing codecs compress each frame independently and far less aggressively. Files are much larger, but any frame can be decoded without reference to others.
Cameras increasingly offer both, and the choice determines whether the bottleneck falls on card capacity or on the computer doing the editing.
Small files can cost more time overall
A heavily compressed file is quick to copy and cheap to store, but decoding it requires reconstructing frames from their neighbors.
Scrubbing through such footage forces the machine to decode a run of frames for every position, which is why timelines stutter on otherwise capable computers.
Editors respond by generating proxies or transcoding to an editing codec, which reintroduces the storage cost that the recording choice was meant to avoid.
Bit depth and sampling change the arithmetic
Recording more color information increases file size substantially, and the benefit only appears if the material will be graded.
Footage destined for straightforward delivery gains little from the extra data, while anything requiring heavy color work or keying degrades quickly without it.
The decision belongs to the shoot rather than the edit, because information not recorded cannot be added afterwards at any file size.
Cards are rated for sustained writing
Storage media advertise peak speeds, but recording video requires a sustained write rate held for the length of a take.
A card that meets the peak figure and fails the sustained one stops recording mid-take, which is a failure that only appears under real conditions.
This is what the sustained write ratings on memory cards describe, and it is the number that should be matched against the camera's data rate.
Budgeting across the whole chain
Storage cost is not only the cards. It includes the working drives during editing, the archive afterwards, and the second copy of that archive.
A codec that quadruples file size quadruples all of it, which is why productions frequently record a lighter format and reserve the heavy one for material that needs it.
Deciding this before a shoot is straightforward arithmetic; discovering it during one usually means shooting less or deleting something that should have been kept.
The calculation needs only the camera's data rate and the expected hours of recording, both of which are known well before anyone arrives on location.