Pushing voice clips to higher or lower frequencies?

A

armadillo

Well-known member
Is there a technique (preferably using FCP or Soundtrack - but open to other techniques) where the frequencies of an audio clip can be pushed down (say to 1/10 or 1/100 of the original frequencies) or minus so many HZ...

Thanks.

This is an independent step of a (likely academic) experiment I wanted to run with some audio cleanup.
 
If you are asking if you can change the pitch of speech, singing, instrumental music, etc., then sure you can. Changing the sample rate will change the speed along with the pitch, but there are software methods of changing length without changing pitch, and conversely, changing pitch without changing length.

However, if you are hoping for some way of implementing your "Creation of missing voice frequencies" you should know that better people (and far richer) than any of us here have been trying to do this for many decades. So far without apparent success. Nothing wrong with dreaming, but don't hold your breath waiting for a practical product. That isn't say that someone won't come up with a way eventually. As we make better/faster/cheaper computer processing, the "dumb, brute-force" method has become practical for many things that we only dreamed of in the past. There are rumors (and some demonstrations) that a company in Europe is close to releasing a product that can reach into the middle of a recorded chord and manipulate single notes. So I wouldn't rule anything out.
 
Thanks for the detailed thoughts.

I guess from a practical perspective, what would the product sound like if you

1. took your clip with missing frequencies
2. copied it and made the copy keep the same length of time but with a different pitch (eg. lower pitch if lower frequencies are missing)
3. overlayed it with the original track

Result: Any improvement?
 
Try it. I do it with sound effects, but never considered that approach on voice. My guess is the resulting odd and even harmonics would make it even less intelligible.

armadillo said:
Thanks for the detailed thoughts.

I guess from a practical perspective, what would the product sound like if you

1. took your clip with missing frequencies
2. copied it and made the copy keep the same length of time but with a different pitch (eg. lower pitch if lower frequencies are missing)
3. overlayed it with the original track

Result: Any improvement?
Click to expand...
 
You can use Amazing Slow Downer to change pitch and speed separately. I'm not sure if you can make changes in specific frequency amounts, however.
 
armadillo said:
Thanks for the detailed thoughts.

I guess from a practical perspective, what would the product sound like if you

1. took your clip with missing frequencies
2. copied it and made the copy keep the same length of time but with a different pitch (eg. lower pitch if lower frequencies are missing)
3. overlayed it with the original track

Result: Any improvement?
Click to expand...

The mathematics involved with doing that are far beyond earth technology at this time. Remember that all that stuff you see on TV and movies is fiction.
 
hey man , wake up and smell the coffee.
technology is doing a lot more today and all that can be packed into an iphone....

amazing slow downer is great but not free, if you run a PC (or emulation of it on a mac) you can use the free one that is called Best Practice
http://www.xs4all.nl/~mp2004/bp/.
quality wise, probably amazing slow downer is better and it runs on Mac.

unfortunately, just mixing 2 tracks will not give the good result.
you will probably to pitch down first, then use a low pass filter to keep only low frequencies and play with the dynamic.(expand or compress)

then you can mix down the 2 tracks.

for audio cleanup, use first izotope RX, the best audio utility to filter audio.
You just need to play real time with the graphic curves until you get the best result.
unfortunately izotope is not this kind of program that just makes the magic for you by just choosing some profiles, so it could take some
time until you find the right way, but it is very powerful.
the secret in working audio, is to find the best level where you can split noise from useful signal. It often require to play with level and dynamic and then
applying graphic equillizer to the right part of the spectrum.
I have done that often, and believe me you can make miracles.
It takes a lot of time so usually the next time you will try to take a better sound at shooting time.

Lot of video people think there is always a "plugin" that can solve their sound problem, like there are so many in video.
unfortunately audio is a lot more complex than picture because what you hear is a dynamical process (mix of waves).
 
Last edited:
Resampling programs that involve time stretching (or shortening) sound so awful to me I can't use them. However, I have seen ppl use them, and there is a market for it.
I guess one way to look at it, is try it out if you really, really need it. And it will sound better on some things than others.
 
in that case he is not speaking about time stretching , but about pitch stretching, that is just the reverse but many people while hearing
voice pitched down, think it is played slower, because they made the vinyl or tape analogy (the slower, the lower)
autotuning is the same principle and can be found on cheap synthetizer keyboard. You sing into the mic. and the synthetizer is just correcting
the pitch so it always sound just in tune. I think , today most of singers are using it soon or later in recording studios.
 
Last edited:
You must log in or register to reply here.
Back
Top