Thanks for your info (Alpay Burak & Daniel Larsson), especially that pointing out the problem of coincidence loss in K2 summit at high dose rate.
The K3 has readout rate in 1500 fps, the best dose rate is 7-40 e/ pixel/s. The K2 is 400 fps, so it is better to use the dose rate of K2 is 2-8 e/ pixel/s.
According to the K2 paper (Li et al., 2013), coincidence loss (undercounting of electron events) happens when more than one electron strikes the same pixel within the readout period. For the dose rate, when less than 5 e/ pixel/s, the effective electron counting drops by a factor of ~0.97. At higher dose rates, the true effective counting decreased, ~0.94 at 7.5 e/ pixel/s, ~0.87 at 10 e/ pixel/s, ~0.6 at 20 e/ pixel/s, more than 20 e/pixel/sec) counting is even worse than the integrating linear mode.
So, at the recommended dose rate like 10 e/ pixel/s, the conversion efficiency is only 0.87 counts per true electron.
In consideration to minimize coincidence loss and push the power spectrum resolution limit better, my question is that should we decrease the dose rate from 10 e-/pixel/second (8-9 counts/pixel/sec) to 2- 5 e-/pixel/second (2-4 counts/pixel/sec)?
If so, for the total dose 40 e-/pixel/movie, the exposure time will be increased from 4 s to 8 s or 20s.
Another question is about the resolution gain from the counting mode to the super resolution mode of K2.
Since in K3 is optimized to run in super-resolution mode as same like in the counting mode after it is binned in 2.
But in K2, they are not the same. Like already discussed in the CCPEM list or from some experts, data collection in super-resolution mode theoretically has better DQE, but the power spectrum limit resolution didn’t change significantly in practice.
Now for us, the pixel size at 60K mag is 0.7891A/pixel, so the physical Nyquist limit is still about 1.57 A, less than 2 A, not cause the aliasing problem.
So, it is not recommended to record in the super-resolution mode by K2 for no more resolution gain, however, it uses 4x space and 2x computing time in the movie processing, not worthy it. Did I get this info correctly?
Regards
Lifei
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1
This message was issued to members of www.jiscmail.ac.uk/CCPEM, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
|