[aprssig] Observations of Packet Test CD Track 2

[DANA MYERS]

Hello all,
I was "encouraged" to post this here.

Back in the idyllic times of 2016, I noticed that Packet Test CD Track 2
doesn't sound right, so I did some objective measuring. Track 2 is
supposed to be the 'de-emphasized' version of Track 1. I extracted the
same section of inter-packet noise on Track 1 and Track 2, and used
Audacity to produce spectrum graphs of both. (if anyone doesn't
understand this technique, email me off-list and I'll explain).

Track 1 (discriminator output) response is what you'd expect for an NBFM
channel, flat-ish out to a few kHz, then IF BW limited:

https://drive.google.com/open?id=0B6J9M7BemghDMU9QdWJuQk9kNm8

Track 2 (de-emphasized) has an unexpected -15dB notch at ~3100Hz,
but don't left that distract from the real problem - it has a -12dB/octave
response from 300Hz to the onset of the notch:

https://drive.google.com/open?id=0B6J9M7BemghDYlB1QVFZdWRUUms

It's easy to miss that it's -12dB instead of -6dB if you don't look at the
Y-axis scale. Track 2 doesn't represent any kind of real-world scenario,
and should simply be deleted. In discussing this Stephen on the phone,
we lamented that Track 2 was too well-established to correct it now.
Stephen has documented well the use of Adobe Audition (IIRC) that
just didn't do what it said it was doing.

I don't use Track 2 and disregard any test results reported using Track 2,
and would encourage others to do the same.

For comparison sake, here's my own de-emphasized Track 1 produced
using Audacity ( -6dB/octave 300Hz LPF, and a high-pass CTCSS cut
filter because that's what real radios do):

https://drive.google.com/open?id=0B6J9M7BemghDV0JKaTdmaHdRUWs

A real-world example, this is the response of a TM-D700A 1200-baud data
output:

https://drive.google.com/open?id=0B6J9M7BemghDNjM3dXFWS2VzRzA

It has the CTCSS cut HPF at 400Hz.

Here's another real-world example, a proper Batwing Motorola Maxtrac 300:

https://drive.google.com/open?id=1Oks8g-tJABNML3Ky6cB8ZGe1B2DKMm5R

Forgive me for using a wider X-axis scale here, I took this measurement at a
different time.

73,
Dana K6JQ

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.tapr.org/pipermail/aprssig_lists.tapr.org/attachments/20200616/41df5332/attachment.html>