R/V Melville Knudsen Echosounder example.

The Knudsen Echosounder may use 4 transmission pulse lengths:

Window         Sampling          Sample
(length)       Frequency         interval
   3ms		25510.2048		39us   (39.2 actually)
   6ms		22222.22222		45us
  12ms		18518.51852		54us
  24ms		14836.79525		67us   (67.4 actually)

In this case the SEG-Y file looks like:
lsd  melville-sample-1.sgy 1 2
      SHOT   TR       RP    TR ID  RANGE DELAY NSAMPS    SI   YR DAY HR MIN SEC
     60037     1       0     0  1      0  6970   1485    67 2000 322  2  45   4
     60038     1       0     0  1      0  6970   1485    67 2000 322  2  45  11

i.e.  a 24ms window using a frequency of 14836.79525 Hz.


    Knudsen offers different digital outputs via different firmwares.
1.  The raw received carrier data.
2.  The match filtered (correlated) carrier signal.  This is what SIO uses.
3.  The "detected" signal (envelope data, also called instantaneous amplitude).
    (Knudsen calls this a square-law envelope detection algorithm which
    uses an 85-tap FIR filter to generate the analytic signal.)
    This is what comes out on the Knudsen display in real-time.  This is
    the Knudsen binary output.

    The SEG-Y file contains the carrier data and thus looks ringy.  (plot)  (script)

    After doing the Hilbert transform and creating the analytic
trace (reals = data trace, imaginary = Hilbert transform), the sioseis 
script creates the filtered (50 x 1000) instanteous amplitude plot.


Date: Wed, 12 Jul 2000 11:27:17 -0400
To: Paul Henkart 
From: Donna Burnell 
Subject: Re: sanple intervals and sample rates
In-Reply-To: <200007031657.JAA24464@sioseis.ucsd.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
Content-Length: 1274
Status: R

Hi  Paul,

Sorry to take so long getting back to you.  Basically the sample interval
should simply be 1000000/sample rate. Thus,  as you noted:

3ms chirp:  fs = 25510.2048:    interval = 39
6ms chirp:  fs = 22222.22222:    interval = 45
12ms chirp:  fs = 18518.51852:    interval = 54
24ms chirp:  fs = 14836.79525:    interval = 67

Our sample rates have not been changed too recently in the sounder firmware
so I do not know why one of the datasets would have had a value of 44,
unless I have a data type handling issue resulting in an invalid truncation
of the data value.  I am in the middle of reviewing and revising the data
file format to make use of the optional trace header fields for additional
sounder information recording.  While doing this, I will review the coding
of the sample interval calculation to ensure it is not being truncated
improperly.

Regards,

Donna Marie
************************************************************************************
Donna Marie Burnell           donna@knudsenengineering.com
Knudsen Engineering Limited        Phone:    (613) 267-1165
10 Industrial Rd.             Fax: (613) 267-7085
Perth, Ontario, Canada, K7H 3P2    
*************************************************************************************