Then S(f) = S(f ) + 20 log ||*- - [H(f) - H(f Q )] 



i.e., Rexburg Local Amplifier Hydrophone 



Calibration Sensitivity Ratio Sensitivity Change 



Now if 9 is the observed signal deflection on the paper (db re 1 mm), R is the 

 attenuator setting in the receiving channel on the Rexburg, A is the local trans- 

 cription attenuator setting and T x (f) is transmitter signal strength (db re 1 |Jbar 

 at 1 yd) the transmission loss is given by: 



Transmission loss = T x (f) - (9 + S(f) + R + A) 



Calibration signals were recorded on the Rexburg at the end of each experi- 

 mental run (except runs 2 and 6) as described above, thus allowing the channel 

 sensitivities, S(f Q ), to be established. 



Table C-3 lists the values of S(f) used. The values of S(f) obtained from Run 5 

 were used in Run 6. For Run 2 the values appropriate to Run 3 were used 

 throughout except on Channel 2 where an obvious change in S(f) for 700 cps oc- 

 curred before time 1819 hours. Presumably this was due to a playback ampli- 

 fier fault since no change was found in V(f) for 700 cps. 



The consistency of the results on Channels 2 and 3 indicated that the values of 

 S(f) for Run 1 on Channel 3 were in error by 10 db suggesting a wrong attenuator 

 setting . Accordingly, these were corrected to the values given in parentheses 

 in the table . 



The values of S(f) for Channel 1 appear to be unacceptable since they contain 

 inexplicable variations of 10 or 20 dbs . The exception occurs after time 1819 

 (Run 5) where an obvious amplifier fault occurred. After this point a new clear 

 experimental calibration obtained at the end of the run was used. For consis- 

 tency it was decided that S(f) = -38/-34/-33 be used on all runs on this channel 

 before 1819 since it was derived from a clear experimental calibration and was 

 reasonably consistent with the other channels. 



The observed noise level (ONL) was used as a check on calibration changes and 

 the general validity of the calculated transmission losses . Since the receiving 

 system is presumed to be sea noise limited the average noise levels can be pre- 

 dicted and compared with the actual noise levels observed on the recordings . 

 The values used for the estimated noise levels (ENL) within the appropriate 

 filter pass bands were as follows: 



700 cps 

 1300 cps 

 3000 cps 



■10 ± 5 db re 1 Mbar 

 ■10 ± 5 db re 1 lobar 

 -15 ± 5 db re 1 i-ibar 



C-7 



Arthur ZD.lUttMnr. 



