|
| | __init__ (self, wasatch_device) |
| |
| | from_db_to_linear (self, x) |
| |
| | from_linear_to_db (self, x) |
| |
| | get_auto_spectrum (self, request) |
| |
| | get_avg_spectrum (self, int_time, gain_db, num_avg, dummy=True, first=None) |
| |
| | get_laser_warning_delay_sec (self) |
| |
| | get_spectrum (self) |
| |
| | inter_spectrum_delay (self) |
| |
| | measure (self, auto_raman_request) |
| |
| | queue_message (self, msg) |
| | wrappers over stupidly complicated WasatchDevice interface
|
| |
| | set_gain_db (self, db) |
| |
| | set_integration_time_ms (self, ms) |
| |
| | set_laser_enable (self, flag) |
| |
This class encapsulates Dieter's Auto-Raman algorithm, which optimizes
integration time (and gain on XS series spectrometers) to achieve a
target window of counts, then uses the configured measurement time
to maximize scan averaging at with those acquisition parameters.
As calling software will not necessarily expect the configured "default"
integration time and gain to change, the class restores those to previous
levels after a measurement.
@par Design Considerations
Mark made the following changes from Dieter's original algo:
- don't take throwaways during averaged signal or dark collections,
as acquisition parameters aren't changing and sensor should be
stable
- don't include laser warning delay when computing num_avg, since
laser is already enabled and firing
Mark also made the following decisions regarding ENLIGHTEN integration:
- return the optimized integration time and gain back to ENLIGHTEN
in the Reading object so they will be the new GUI settings if the
user simply hits "Play" to resume free-running spectra.
- ENLIGHTEN will override the init_int_time and init_gain defaults
in AutoRamanRequest so that if the user hits "Auto-Raman Measurement"
a second time, the previous settings will be used as the new starting
point, and (ideally) the initial spectrum will determine that no further
optimization is required.
Points to consider:
- consider rolling the "last optimization" measurement directly into the
averaged sample spectra (since it had its own throwaway and presumably
represents a "stable" reading). This could potentially allow for one more
averaged dark, depending on max_ms.
- ENLIGHTEN's laser button should still be able to turn the laser OFF at
any point.
1.11.0