RTTY
[Top] [All Lists]

Re: [RTTY] Waterfall or Spectrum

To: RTTY Reflector <rtty@contesting.com>
Subject: Re: [RTTY] Waterfall or Spectrum
From: Kok Chen <chen@mac.com>
Date: Mon, 02 Mar 2009 11:01:38 -0800
List-post: <rtty@contesting.com">mailto:rtty@contesting.com>
On Mar 2, 2009, at 10:01 AM, Iain MacDonnell - N6ML wrote:

> Hmmm. I suppose one benefit could be that it shows some (short)  
> history.
> If someone tried calling off-frequency, and you didn't get to RIT
> quickly enough (maybe you were working someone on the other radio),  
> you
> can estimate how far off he was by looking at the traces in the
> waterfall.

On top of that, the eye is extremely good at correlating and picking  
up "regular" structures in successive scan lines so you can see  
signals way down into the noise.

I had also mentioned to the original poster by private email that if  
done right, a waterfall can have a much larger "resolution" than a  
spectrum.  A spectrum is typically 100 pixels to 800 pixels tall.   
False color together with 200 levels of intensity scale in a waterfall  
can give you a much larger range of differentiation.  But you do have  
to choose the color transition carefully to take care of the RGB luma  
variation that the eye perceives.  Good waterfall color design is  
obviously not a haphazard picking of colors, and unfortunately, it  
also may not match people with color deficiencies.  I have  
occasionally been tempted to provide a waterfall preference for the op  
to choose waterfall colors that are better for red-green or blue- 
yellow type of color blindness.

If your program provides a way to replay data back, a waterfall is  
indispensable.  With a spectrum, not only wouldn't you know where to  
click for the signal, you also cannot tell the program how much of the  
past to replay back.  With a waterfall type spectrogram display, the  
horizontal axis is the frequency offset and the vertical axis is the  
time offset into the past.

Spectrums are very useful, nevertheless, especially if you want  
quantitative numbers.  If you want to see how wide an RTTY signal is  
spreading, you need a spectrum.  With a waterfall, it is hard to tell  
how many dB a keying sideband is down by, even if you do know the  
color coding that was used.  But for rapid (rabid?) tuning, I always  
prefer the waterfall.

So to anyone implementing software modems -- provide both! :-)

73
Chen, W7AY

P.S. the first place I saw a waterfall use effectively in ham radio  
was in Skip KH6TY's DigiPan program.  That single program converted  
PSK31 from a quaint mode into arguably the most popular digital mode  
today.  If you still have to tune PSK31 with a VFO knob, PSK31 would  
never be were it is today.  Can you even imagine tuning in an MFSK16  
signal without the aid of a waterfall?

(The first place I encountered the use of a spectrogram (waterfall)  
for weak signal detection is in the conference papers of NASA's  
Project Cyclops back in 1971 or thereabouts.  I am almost certain  
though, that the technique was used much earlier for radar and sonar  
work.)

_______________________________________________
RTTY mailing list
RTTY@contesting.com
http://lists.contesting.com/mailman/listinfo/rtty

<Prev in Thread] Current Thread [Next in Thread>