The Nebula results page has been updated a number of times:
April 13, 2018
Birdie generation: change non-bary max frequency half-width
from 50K to 20K.
This way a non-bary birdie will fit in our 50K Hz window
for non-bary multiplets.
April 3, 2018
Fix how birdie chirp rates are calculated.
March 15, 2018
Remove signals with excessive power (> 80% of FFT length).
March 8, 2018
Look for combined spike/gaussian and pulse/triplet multiplets,
rather than each type separately.
January 31, 2018
Fix bugs in birdie generation.
Fix bug in chirp pruning.
Small fix to zone finder statistics.
January 1, 2018
Observation selection was using the wrong time interval.
Calculate # of observation periods correctly.
Fixes to multiplet finding; allow several multiplets per window.
New multiplet-finding algorithm
using smarter chirp pruning and observation selection.
October 24, 2017
October 6, 2017
September 30, 2017
New algorithm for period pruning.
Use Eric's new formulas for multiplet exclusion time.
September 17, 2017
Apply spike drifting areas (triangles) to other signal types.
Fix time bugs in birdie generation.
Apply spike zones to other types as well.
August 7, 2017
Increase birdie power range to generate more signals.
Make sure scoring includes all birdie pixels.
July 8, 2017
June 14, 2017
Chirp pruning: discard multiplet signals with
inconsistent chirp rate/freq.