# Phase Cancellation Question (for school)

Discussion in 'Recording' started by mrkite, Nov 4, 2008.

1. ### mrkiteGuest

Hey,

I'm taking an audio class and I'm on a homework question I can't seem to remember the answer (even though it seems to be a pretty easy question).

"If you cause phase cancellation at 1000hz, you experience cancellation at 2000hz, 4000hz, and 8000hz. What is this kind of phase cancellation called?"

Could anyone tell me what this is?

I know the overtones of a sound are multiples of the fundamental frequency (200Hz fundamental frequency, 400Hz, and 600Hz harmonics), which is kind of like the 1000Hz, 2000Hz, 4000Hz, etc.

2. ### moonbabyMmmmmmModerator

Joined:
Feb 23, 2005
Location:
jacksonville,fl
I'm trying to remember this one myself, said I, as I ran MY FINGERS THROUGH MY HAIR (that's a hint)...

3. ### mwacousticGuest

Hmmm... I'd have to BRUSH up on my acoustics theory to answer this one...

4. ### SpaceDistinguished Member

Joined:
Jun 26, 2007
I combed the whole web and didn't find the answer?

edit: changed "at" to "end"

5. ### hackenslashActive Member

Joined:
Jun 9, 2008
Location:
People's Republic Of Mancunia

6. ### CodemonkeyDistinguished Member

Joined:
Dec 11, 2007
Location:
Scotland, UK

Now, I was under the impression that 1000Hz actually had harmonics at 2KHz, 3KHz, 4KHz, - not on a x=2*(x-1) scale.

7. ### naturalActive Member

Joined:
Jul 21, 2006
It's a trick question. Or maybe an incomplete question.

There's not enough information in the question to give an answer.

Are all freq happening simultaneously? Or is it a recording of a 1K test tone, move on to the next tone 2K etc. Plus if the 1K tone is a sine wave there's likely no overtones at the upper freq.

It depends if the Freq at 2k 4k etc are made up from the 1k source. If so then reversing the POLARITY of the source will eliminate all overtones.
If 2k and 4k are separate freq having nothing to do with the 1k, then affecting the 1k has no effect on the others.

That's my best guess so far.
Send my degree in the mail.

8. ### GeckoMusicGuest

x=2*(x-1) simply means that x=2

Do you mean f=fo*(2^n) where n is an integer zero or greater and fo is the base frequency?

That would be 1k 2k 4k 8k etc. Those are the octaves

f=n*fo are the harmonics (1k 2k 3k 4k etc n is positive, fo is the fundamental)

<hint> Note that all the octaves are contained in the harmonics. </hint>

9. ### TVPostSoundGuest

I also combed the web, but couldn't filter anything from it!!!

10. ### BobRogersDistinguished Member

Joined:
Apr 4, 2006
Location:
Blacksburg, VA
For the benefit of Mr. Kite and all of the rest of us, isn't the type of cancellation being hinted at the filtering of all the odd harmonics of a fundamental frequency? As in...

0 = Sin w(t+d) + Sin w(t-d) = 2 Sin wt Cos wd

This implies

wd = (2n-1)Pi/2 for all natural numbers n or

w = (2n-1)Pi/(2d) ... odd multiples of Pi/2d.

Joined:
Sep 26, 2005
This type of phase cancellation is called "COMB FILTERING" based upon its second harmonic. So comb filtering is the answer. Now take a bath and comb filter your hair.

Chrome filtering is created with time delay effects.
Ms. Remy Ann David

12. ### GeckoMusicGuest

Trying to confuse things by moving into the time domain I see. I might need to comb though a text book to figure out what you are saying. You are correct that it is only odd harmonics effected by the comb filter delay. I had it in my brain that it was all harmonics.

13. ### BobRogersDistinguished Member

Joined:
Apr 4, 2006
Location:
Blacksburg, VA
Well, his question asked about octaves of 1kHz. No odd harmonics there.

Joined:
Jul 21, 2006
Oh, I see,
very clever

15. ### GeckoMusicGuest

I think the first 4 did as well as a number of other posters.
hmmm... those are even harmonics of 500Hz. I see your point. Sorry. So it's not a delay induced comb filter...

16. ### GeckoMusicGuest

Oooh! I think I have the more specific answer.
mrkite, are you studying room acoustics and standing waves?

17. ### SpaceDistinguished Member

Joined:
Jun 26, 2007
It was a real clear night here @ Exit4, I just followed the moon baby!

18. ### CodemonkeyDistinguished Member

Joined:
Dec 11, 2007
Location:
Scotland, UK
Gecko: my bad, typo I think. I was trying to imply that I thought the octaves were simply at F = n*x where F = harmonic freq, N = harmonic number and X = base freq.

Ofc that brought 4 posts of :S formulae.

And now we seem to be making randomly emphasised puns at each other.
Great.

19. ### moonbabyMmmmmmModerator

Joined:
Feb 23, 2005
Location:
jacksonville,fl
In Scotland, phase confuses you...

20. ### mrkiteGuest

Thanks a lot for all the great answers, I got my homework answer in just in time.

Nope, just taking a class on how to record things. Nothing fancy just yet