Appropriate effects for .wav / need help

Hi everybody,

I am an Audacity newbie and therefore I fail to remove the disturbing frequency from the .wav-file uploaded with this post.
I’ve marked position 0.0 to 0.5 to find out the frequency of the disturbing sound and then I used the “High Pass Filter” with >1000 Hz, but I don’t get the expected result.

Can somebody please listen to the file (don’t worry - it’s only 4 seconds long :smiley: ) and tell me how to “fix” that file, i.e. just getting the voice?

Thank you very much!

Regards

Notch filter (available in Audacity 1.3.13)
Run it twice, the first time with frequency set at 1000 Hz, the second time set at 150 Hz.
Higher Q values will remove less of the voice sound but will leave a longer “blip” at the start.

Hi Steve,

thank you very much for the quick reply.
Unfortunately, we are solely allowed to use the following functions (and/or):

  • Plot Spectrum
  • High Pass Filter
  • Low Pass Filter
  • Equalization

These restrictions makes “cleaning” of the file more difficult - at least from my point of view. Any ideas?

I forgot to mention that I am running version 1.3.13

Reagards

Are you asking us to do your homework? :smiley:

You can make a notch filter using the Equalization effect. You may need to check the results against the spectrum plot to get the notches in just the right place, but something like this will work: (see here for an explanation of the settings http://manual.audacityteam.org/man/Equalization )

I would never do that - honestly: let’s call it an act of desperation, because I am sitting here already some hours without results. :wink:

I’ll make myself familiar with the manual, but before I have two question:

  • How did you get to that settings, I mean did you use Analyze → Plot Spectrum?
  • Is there an alternative way to solve that exercise, maybe more easier?

Thank you very much!
Regards

Yes. See the two big spikes - one at 150 Hz and one at 1000 Hz. Increase the “Size” parameter to get more accurate frequency readings.

I could probably dream up some more complicated ways :smiley:
The easiest way is to look at the spectrum and then use the notch filter, but you didn’t allow that answer :wink:
The next easiest way that I can think of is to make a double notch filter using the Equalization effect.

For interest you may also like to try this:

  1. Select the track
  2. Effect menu > Nyquist Prompt
  3. Copy and paste this code into the Nyquist Prompt text box and click OK:
(notch2 (notch2 s 150 20) 1000 2)

Is this another plea for closer integration between Analyze Spectrum and Equalizer? As it stands now, you have to carefully analyze the work and assuming two serious spikes, one at 1KHz and the other at 150Hz, write those two numbers down, open up the equalizer and pull down those two frequencies, or launch the notch filter and do it that way.

I’m the last person to knock paper and pencil, but it would seem there could be a way to say: “See that spike? [click or drag] Equalize it.”


But back to the actual problem.

  • How did you get to that settings, I mean did you use Analyze → Plot Spectrum?
  • Is there an alternative way to solve that exercise, maybe more easier?

Yes and no, in order. Plot Spectrum will show blue waves that will go up and down with the different pitch tones in the sample. Drag-select a portion of the show with interference alone – as much as possible. Analyze that. We already told you where to look for the blue upward spikes in the waves. You can pull the analysis window large and larger to get greater and greater detail, plus you can increase the SIZE number to get even better accuracy.

Write those two numbers down (yes, I know you already know them) and apply them to the notch filter one at a time. Listen carefully before and after you do that and the tones should be gone.

Listening is a very serious part of this and in the sense that photography will tell you how to see, Audacity will tell you how to hear.

Koz

This is an analysis I did of one single piano note. I think the first note from here:

http://www.kozco.com/tech/piano2.wav

This is the analysis.

http://kozco.com/tech/audacity/piano_G1.jpg

The mess to the right of the actual note, G1, is the overtones and harmonics that make a piano a piano. Those spikes are very different between a Steinway and a 7-Eleven or Tesco piano even though they’re playing the exact same note. That tall spike toward the left will be the same on both.

7-Eleven pianos are getting hard to come by.

Koz

I’ve tried that and: Wow! This software is just amazing! :smiley:
But making myself familiar with the Nyquist Prompt would take too much time (especially because I have to write my bachelor thesis at the moment)… I cannot understand why Audacity is for free, I mean this software is like Adobe Photoshop but for audio?

Regards

see GIMP and GIMPshop.

Thanks for the links, but I think my last post was a little bit unclear. From my point of view it’s good for the users that the software is for free, but since I know how many time and effort software engineering costs, it’s a great pity that developers don’t earn money with it. Anyway, I’ll better stop to write off topic related stuff… :blush:

Now we have to use a micro and to record our name (Mono, 8000Hz). Furthermore they want us to add new audio tracks to the voice track and to produce the same interference effects that are in the file I’ve uploaded with my first post by generating new tones.

I’ve recorded my voice but I fail to produce the same interference effects. Any ideas which settings I have to use in the tone generator?

Thank you very much!

Short notice: I’ve managed it. :wink:

Thanks anyway!

Regards