SQR wave ringing problem

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Michael Krusch

Well-known member
Joined
Jun 4, 2004
Messages
444
Location
Berlin, Germany
Funny, bought my Audio Tester on Sunday, I'm one day ahaed of you. :grin:

This isn't ringing, it's only a displaying issue. In CoolEdit (now Adobe Audition) i.e. you can configure this behaviour. Record some square wave and watch it with some different sample editors and you will see some different pictures of the same wave file.
 
There is nothing wrong with the card. In digital after sampling a 10KHz square wave at 44.1 or 48k you will have only the 10KHz sine. The rest of the harmonics that form the square will be gone.
Also the waveforms from your wave editor has nothing to do with the real analog waveform. You need the reconstruction filter in the DA process.

chrissugar
 
Hey Jaako, do not use the computer and the soundcard to solve ringing problems. You need a real analog scope to view the square wave [ you know all the harmonics up to the hundreeds of kilohertz or up to megahertz. A computer card in the AD process will filter out everything that is over 20KHz so you will have no harmonics to analyse. So a high freq square will never look like a square. Try a 100Hz square. It will look much better.

chrissugar
 
Yes.

Audiotester will do all sorts of cool things, but for out-of-band response checks you'll still need your tonegen+scope.

Squarewaves are genericly ****** in digital audio. Nothing upsets these as steep filters.

JAkob E.
 
Sorry for the delay.
Here is the picture:

wave.gif


As you can see, the samples are ideal square wave, but the line suggests some ringing.
 
While I'm not 100% sure - I agree with chris.

There is no such thing as a perfect square wave as it requires infinite harmonics to be created - at best, you have an approximation.

That picture looks exactly like a simulated square wave with upper harmonics missing. I would guess thats around the 13th harmonic or so....

The AD/DA anti-alias filters are obviously going to lop off some of the upper harmonics.

Jaakko - have you tried running the tests at 192k on the lynx.....can AudioTester do that?

I know RMAA can - but I'm not sure if it has square wave generation.

That way you may get a more accurate representation of the square your sending out.

Cheers Tom
 
[quote author="TomWaterman"]...
That picture looks exactly like a simulated square wave with upper harmonics missing. I would guess thats around the 13th harmonic or so....
...
Cheers Tom[/quote]

???
That is a totaly ideal square wave. How else should an ideal square wave look like? You can't have a better wave at this given sample rate.
But when it leaves the digital domain, it will be deformed by the filter of the D/A converter.
 
As Chris says, analog scope and signal generator. Always super impose the pre and post signal. My B&K Presicion generates square waves with overshoot the size of Texas, UNLOADED or with high impedance loads.
I often have to load it down with a 5k resistor to tame it
 
[quote author="Michael Krusch"]???
That is a totaly ideal square wave. How else should an ideal square wave look like? [/quote]

What S/R is that at? What is the fundamental? The totally ideal square wave would have no ripple, instantaneous leading edge etc but as you state not possible at limited sample rate.

Its obviously a square wave with upper harmonics missing - which are not there because of the sample rate of the generation.....I thought that was the captured wave coming back in.

So how does that wave look after going through the DA/AD filters?

Cheers Tom
 
One more thing. People think that because this behaviour [ringy square wave] digital is broken. IT IS NOT. Digital works this way.

Even our ears are like an abrupt low pass filter. As an example try to listen to a 15Khz sine wave and a 15Khz square wave generated with an analog oscillator. The sound should be the same because you do not hear the harmonics contained in the square wave.

chrissugar
 
Back
Top