Struct wfm_draw_t¶
One rendered source instance: its timing AND the values it was actually rendered with. More...
#include <wfm_compose.h>
Public Attributes¶
| Type | Name |
|---|---|
| size_t | delay |
| double | doppler |
| double | doppler_rate |
| double | f_end |
| double | freq |
| size_t | instance |
| double | level |
| size_t | off |
| size_t | on |
| size_t | seg |
| double | snr |
| size_t | src |
| size_t | start |
Detailed Description¶
A wfm_span_t answers when; this answers when and what, for one source of one instance. The distinction is not academic. The SigMF sidecar used to build each annotation from two provenances timing replayed through wfm_compose_spans(), frequency and SNR read straight off the source struct, which for a ranged field still holds lo so every annotation of a --freq 11200:12800 --snr 8:14 scene claimed 11200 Hz and 8 dB beside a sample-accurate start. Measured against the capture itself: up to 1224 Hz and 6.0 dB out (doppler#1086). A 6 dB error is a different operating point, and nothing in the file revealed it.
An un-ranged field reports its scalar, so a consumer never branches on the ranged bitmask.
The spec keeps storing (lo, hi): replay is guaranteed by re-deriving the draw hash, not by recording the draw. "What does this spec permit" and "what did this run do" are different questions and one field cannot answer both which is why this is a separate call rather than a resolved spec.
Public Attributes Documentation¶
variable delay¶
variable doppler¶
variable doppler_rate¶
variable f_end¶
variable freq¶
variable instance¶
variable level¶
variable off¶
variable on¶
variable seg¶
variable snr¶
variable src¶
variable start¶
The documentation for this class was generated from the following file native/inc/wfm/wfm_compose.h