Subject: RE: NTT STEX III vs. STEX II
Date: Fri, 2 Jul 1999 03:37:40 -0500
From: "Otsuka, Yutaka" <Yutaka.Otsuka@mts.com>
To: "Larson, Rod" <Rod.Larson@mts.com>
CC: "Iseki, Nobuo" <Nobuo.Iseki@mts.com>

Hello Rod-san,

Thank you for the quick reply.
But I wondered, they said first iteration and second one were finished
without any error message.
If sample rate is wrong, is it still able to iteration? Anyway we will
check the sample rate and let you know
What happen after all.

Yutaka

        -----Original Message-----
        From:   Rod Larson
        Sent:   Friday, July 02, 1999 3:14 PM
        To:     yutaka.otsuka@mts.com
        Cc:     nobuo.iseki@mts.com
        Subject:        NTT STEX III vs. STEX II

        Yutaka,
        Nice to hear from you.  Working at UCSD San Diego this week.  I
think the
        problem is sample rate.  Remember that NTT can ONLY use 250
samples/second.
         STEX II, I think, uses power of 2 sample rates, like 256
points/sec.  This
        would cause big, big trouble.  The customer MUST re-sample the
data if the
        sample rate is NOT 250 pts/sec.  Because the sample rate is
incorrect, the
        wrong # of total data points would be taken for the FFT process.
This
        process requires an integer number of data blocks.  With sample
rate
        problems, there is considerable aliasing in the data process
that will
        result in the data blowing up quite quickly in the iteration
process.  Let
        me know what the result of checking the data rate is; however,
all of the
        symptoms point to this problem.

        The reason that STEX III at NTT can only use 1 sample rate is
that we only
        have 300 Hz filters on the signals.  There is considerable
aliasing at
        other lower sample rates.

        Rod Larson

