PR #434 proves that the current dRICH test measures the wrong metric and the current dRICH implementation does not work with GDMLs. The test on main introduced in PR #428 shoots photons to the sensors bypassing the dRICH mirrors that the Cherenkov photons need to hit and be reflected onto the sensors.
In PR #434 I launch the photons from open gas behind the sensor shell along the mirror-center -> sensor-patch axis: the photons take one specular bounce off the mirror sphere and refocus onto the same SiPM patch the test in main hits directly. Same geometry, wavelength and seed, only pos/mom change.
Integration.simg4ox.drich then fails: G4 0 hits vs GPU ~5600 .
PR #434 proves that the current dRICH test measures the wrong metric and the current dRICH implementation does not work with GDMLs. The test on main introduced in PR #428 shoots photons to the sensors bypassing the dRICH mirrors that the Cherenkov photons need to hit and be reflected onto the sensors.
In PR #434 I launch the photons from open gas behind the sensor shell along the mirror-center -> sensor-patch axis: the photons take one specular bounce off the mirror sphere and refocus onto the same SiPM patch the test in main hits directly. Same geometry, wavelength and seed, only pos/mom change.
Integration.simg4ox.drich then fails: G4 0 hits vs GPU ~5600 .