Re: VLF: 8270: now 8269.990Hz

["Markus Vester" <markusvester@aol.com>]

Hi Paul,

 

that shouldn't be so... According to my favourite hex editor, I get the same result as you for Stefan's samples. But my first samples should be

 

698.0      1628.0

2096.01  -4322.74

1390.22  -914.946

-1856.24 1887.09

etc.

 

The sign seems to be preserved in your numbers (mostly?). But the magnitude is apparently being limited to one. Is ebnaut-rx software expecting sample values to be smaller than unity?

 

The reason behind the vastly different magnitudes between Stefan and myself is that we have ticked different formatting options for SpecLab FFT export (float versus scaled-up integer). But this should have no side effects other than a different global scale factor. 

 

Best 73,

Markus

 

From: Paul Nicholson

Sent: Saturday, April 02, 2016 9:09 PM

To: rsgb_lf_group@blacksheep.org

Subject: Re: VLF: 8270: now 8269.990Hz

Markus wrote:

 > I have uploaded the wavs

Thanks I'm looking at those.

Your wav 04010012.wav contains only samples values
of +/- 1,

eg

  sample  ch1  ch2
  0        1    1
  1        1   -1
  3       -1    1
  4       -1    1
  5       -1   -1
  6       -1    1
  ...

Stefan's dk7fc_04010002.wav has samples like

  sample  ch1          ch2
  0    5.2787364e-06  -6.6896901e-06
  1   -3.0420255e-05   4.7603622e-05
  2   -2.1549407e-05  -2.1126121e-05
  3    4.9282797e-05  -9.3932264e-05
  4   -2.4563633e-05  -7.9385005e-05
  5    2.7562026e-05   1.3096724e-05
  6   -0.00012802938   0.00018837629
...

Is that intentional?   I expect it might lose a dB or two.
--
Paul Nicholson
--

wavs_hex.png
Description: PNG image