We have checked if there are coincident triggers between DF and other channels using the MF triggers obtained on WSR1 and WSR5 data sets (note that the trigger lists have been previously cleaned using these DataQuality flags given here:

Pr_B2_DC
Pr_B2_3f_d1_ACp
Pr_B5_ACp
Pr_B5_ACq 
Bs_IMC_D1T_DCHF
Sc_IB_SSFS_Corr
Sc_WE_txCorr
Sc_WE_tyCorr
Sc_NE_txCorr
Sc_NE_tyCorr
Sc_IB_MC_ty
Sc_MC_tyCorrR
Gc_Recycling
Gc_Michelson
Em_ACDBDL01
Em_MABDNE01
Em_MABDWE02
Em_SETOBS01

No coincidence have been found with the Em_XX, Bs_IMC_D1T_DCHF and Sc_IB_SSFS_Corr channels (the coincident events are compatible with accidental coincidence). The most promising channels so far are:
Sc_WE_txCorr
Pr_B2_DC
Pr_B2_3f_DC

The following table shows the performance obtained requesting:

• a coincidence between DF and the tested channel within a window of +/- 100 ms.
• a threshold on the tested channel has been applied (the same on all the segments)
• a threshold on the DF channel SNR has been applied: SNR>10
• a test of robustness using all hardware injections (CB + burst) has been done: it consists in generating veto lists using the time window and the threshold given in the follwoing table (no time offset has been added) and then applying the veto on the DF triggers containing the hardware injections. The number of hardware injections which do not pass the veto are then counted.

It's interesting to note that:

• Pr_B2_DC and Pr_B2_3f_d1_ACp triggers seem to occur 30-40 ms bfore the DF triggers, while Sc_WE_txCorr triggers occurr 20ms before (it must be investigated if this is true on all Sc_XX_txCorr ...)
• Among the 3 promising channels, it seems that Pr_B2_DC and Pr_B2_3f_DC have the best use-percentage. Actually on WSR5 seems to be better than Pr_B2_DC. But a complete study remains to be done (no optimization of the threshold has been done)
• Sc_WE_txCorr does not seem to be safe wrt hardware injected events (2 events killed over 112 injections) ... to be confirmed and understood (could be due to accidental ... the threshold is too low for this channel. But anyway that does not smell good)

What remains to be done:

• optimization of the veto parameters (deltaT and threshold)
• understand the origin of the remaining loud events ... find out other good channels!
• complementarity of the XX vetos
• ...

Numbers	peak time difference and SNR	SNR (DF and tested channel)
SNR>15 Pr_B2_DC:347 COINCIDENT:245 accidental: 0.22 ROBUST: 0 killed
SNR>15 Pr_B2_3f_d1_ACp:276 COINCIDENT:124 accidental:.17 ROBUST: 0 killed
SNR>15 Sc_WE_txCorr:1880 COINCIDENT:214 accidental:1.13 ROBUST: 2 killed
SNR>7 Sc_IB_SSFS_Corr:999 COINCIDENT:4 accidental:.63 ROBUST: not tested
SNR>10 Bs_IMC_D1T_DCHF:238 COINCIDENT:124 accidental:1.5 ROBUST: not tested
SNR>7 Em_ACDBDL01:56 COINCIDENT:1 accidental:.04 ROBUST: not tested
SNR>15 Em_MABDNE01:252 COINCIDENT:0 accidental:0 ROBUST: not tested
SNR>10 Em_SETOBS01:20 COINCIDENT: 0 ACCIDENTAL:0 ROBUST:not tested
SNR>15 Sc_IB_MC_ty:99 COINCIDENT: 0 ACCIDENTAL:0.18 ROBUST:not tested
SNR>15 Sc_MC_tyCorrR:295 COINCIDENT: 4 ACCIDENTAL:0.6 ROBUST:not tested
SNR>15 Gc_Recycling:816 COINCIDENT: 481 ACCIDENTAL:1.5 ROBUST:2 killed
SNR>15 Sc_WE_tyCorr:23 COINCIDENT: 13 ACCIDENTAL:0.0 ROBUST:not tested
SNR>15 Sc_NE_txCorr:1671 COINCIDENT: 208 ACCIDENTAL:1.0 ROBUST:not tested
SNR>15 Sc_NE_tyCorr:3 COINCIDENT: 0 ACCIDENTAL:0. ROBUST:not tested