Converting Triangle Wave to Sine?

Hello, I’m new to this Board and want to know if there is a Nyquist Prompt for Converting Triangle or Sawtooth Waves into Sine Waves.

It shouldn’t be that hard to do. Select a frequency just higher than the fundamental and apply the Low Pass Filter at that frequency. That should strip off all the upper harmonics leaving a single sine wave. It should work on square waves, too. Same reason. Both square waves and triangle waves are sine waves with a lot of upper harmonic stuff added. Strip that off and you’re left with the original sine.

You can use Analyze > Analyze Spectrum to figure out which is the frequency. It will be the lowest lump. Put your cursor on it and the analysis readout on the bottom will tell you the pitch.

You are warned that the resulting sine may be a lot larger than the original and you need to allow room for that.


I don’t know of a Nyquist for that…

It’s fine, I Synthesize Air Raid Sirens and I usually just use Sawtooth & Triangle Waves.

So you want a tone as irritating and attention-getting as possible. That would do it, particularly if some of it ends up in the 3000 Hz range. I was thinking about this. I don’t think you would ever get one tool to do it – it requires multiple steps and processes. Maybe the worst part of the puzzle is figuring the fundamental of the wave, given that it could be anywhere in the audible spectrum. Given the air raid siren, the pitch might be moving as well, although I don’t know that they sound like London during the blitz any more…


Was all this just a commercial for your company?

Commercial? If you mean by advertising or something like that then no. But its an Organization.

Well so your saying that it is possible to convert from triangle wave to sinusoidal wave using the above mentioned procedure, once you have already calculated the fundemental frequency?

Is the fundemental frequency usually the greatest amplitude waveform of a complete waveform across 0 volts from positive to negative?

That is I think what this webpage says:

But what if I want to analyze a stereo audio track?? What if there is more than one instrument playing on the audio track?? When I record audio I deliberaltely only record one instrument at a time to avoid such complications, but what if I have no choice the audio is already recorded with muliple instruments on the same track?

What if the fundemental frequency is varying bc it is not a pure tone, but rather a song? Is there such a thing as a temporal frequency equalizer to apply our tricks that you have mentioned above to convert from triangle waves to sine waves?

Or should I just give up and not convert from triangle waves to sine waves for something that is not a pure tone? What I am trying to do is the track is recorded to have too much distortion such that I want it to have high gain, but the high gain setting on the amp causes when I view the waveforms in NCH Wavepad it shows that the waveforms are triangle waves they are so clipped.

@dietermoreno: your picture above does not show a triangle wave, instead it shows an audio waveform containing several superimposed signals. I do not understand what your questions have to do with “Converting a Triangle Wave to a Sine”. Could you please explain a bit more or open a new thread, because to me it appears as if your questions have nothing to do with triangle waves.

Quick answers:

Most often yes, but not always. You can’t rely on that the greatest amplitude waveform automatically is the fundamental frequency.

Depends on the stereo correlation (how much the signals differ in phase). In case of doubt analyze each channel separately and compare the results.

With superimposed signals (like in the picture above) it’s from very difficult to nearly impossible to compute the fundamental frequencies of all the different original signals.

Then you must chop the audio signal into a series of small pieces (usually a few milliseconds), analyze all pieces, and concatenate the results.

Google for Phase vocoder.

This has nothing to do with “Converting a Triangle Wave to a Sine”.

Signals distorted by clipping can’t be restored because the clipped-away parts are missing and cannot be artificially re-generated because nobody knows what’s exactly missing. There had been many attempts in the past but I don’t know a single one that works with convincing results.

  • edgar

Well I have been in the cheap ware for only $40 NCH WavePad Master’s Edition, I have been able to make a convincing deletion and attenuation of triangle waves and an amplification of sinusoidal waves.

I clicked “Grab noise sample”

The noise sample I took was of a triangle wave.

Then I clicked “Apply spectral subtraction with noise sample”.

Then I went in and boosted the fundemental frequencies of the lead guitar at 20Hz-100Hz and 1000Hz-4500Hz and I cut out its harmonics at 4500Hz-20,000Hz and that was convincing to fatten the sound of the lead guitar and appeared to remove snipets of triangle wave forms.

Well but the method I proposed above I guess it only works for manipulating audio tracks that have triangle waves embedded in sinusoidal waves from background noise and distortion.

So it wouldn’t work for truelly converting a clipped waveform into a sinusoidal waveform bc as you said earlier, many have tried but none have ever succeeded at guessing what waveform should be where the waveform was clipped.

The method I performed above I suppose the source track was of non clipped waves. This was recorded with a FlipCam at a concert where I pressed the shell of the FlipCam right up to the loud speaker while wearing ear plugs.

I’m not quite sure why a FlipCam performed better acoustically coupled to a loud speaker than a dynamic microphone acoustically coupled to a loud speaker.

The waveforms totally clipped to triangle waves that is when I acoustically coupled a dynamic microphone to the speaker cabinet of my guitar amp.

Do you know the reason for why the FlipCam worked better than the dynamic microphone in settings acoustically coupled to a loud speaker?

If you know the frequency of the triangle wave then you can convert it to a sine wave by filtering out all frequencies above the fundamental.
For example, to convert a 440 Hz triangle wave to a 440 Hz Sine wave you can get a close approximation with:

(lowpass8 s 500)

If the triangle wave is aliased, then for a more accurate sine wave you will also need to filter out frequencies below the fundamental, for example:

(mult 2 (highpass8 (lowpass8 s 440) 440))

However, this is of little practical use as it would be much better to just generate a sine tone of the required frequency.

Your posts have nothing to do with the original posters question. If you have a question or something to share, please start a new topic. It is unhelpful to hijack other peoples topics.

Thank you very much!