back to list

New here-- questions about real time pitch correction

🔗tonydananza <greg.s.larson@gmail.com>

10/8/2005 3:53:14 PM

I first discovered alternate tuning a few weeks ago and I was very
excited about the idea. I'll be graduating this year in EE from Cal
Poly SLO, and while I haven't been studying music lately, I would like
to build my senior project around music, since it is what has gotten
me into the field. Creating a basic DSP seemed a little bland to me.

I am thinking about making a device that converts an audio input
signal into the just tuning for a key or keys defined on the fly by
the user or automatically if possible. I think I should be able to do
this with an FFT and look for impulses that do not match up with the
correct pitches and multiples for whatever key I'm in, and cancel out
them while creating new ones at the closest just pitch.

A big question I have is: does it make a big enough difference to make
it worth the effort? I haven't seen any products like this, which is
very suprising to me, so I figured there might be a reason why. Thanks
in advance for any thought put into this, sorry I have not found the
answers myself in your archives. -Greg

🔗Magnus Jonsson <magnus@smartelectronix.com>

10/8/2005 7:54:58 PM

Hi Tony,

I think you the most appropriate list for this question is the music-dsp mailing list. They are very knowledgeable. You can join it at:

http://www.musicdsp.org/

Now to address your question:

I would be concerned with transient smearing and/or loss, or metallic sound, which are common artifacts in basic pitch shifting. What you want to do is more advanced than pitch shifting since, if I understand your idea correctly, you may want to shift some signal components up and some down. While I don't want to discourage you, I think you should aware that this can be an easy problem or an extremely hard problem, depending on if you expect the output to sound better than the input and depending on if you want it to work for polyphonic signals or not.

- Magnus Jonsson

On Sat, 8 Oct 2005, tonydananza wrote:

> I first discovered alternate tuning a few weeks ago and I was very
> excited about the idea. I'll be graduating this year in EE from Cal
> Poly SLO, and while I haven't been studying music lately, I would like
> to build my senior project around music, since it is what has gotten
> me into the field. Creating a basic DSP seemed a little bland to me.
>
> I am thinking about making a device that converts an audio input
> signal into the just tuning for a key or keys defined on the fly by
> the user or automatically if possible. I think I should be able to do
> this with an FFT and look for impulses that do not match up with the
> correct pitches and multiples for whatever key I'm in, and cancel out
> them while creating new ones at the closest just pitch.
>
> A big question I have is: does it make a big enough difference to make
> it worth the effort? I haven't seen any products like this, which is
> very suprising to me, so I figured there might be a reason why. Thanks
> in advance for any thought put into this, sorry I have not found the
> answers myself in your archives. -Greg

🔗Carl Lumma <clumma@yahoo.com>

10/10/2005 1:01:03 PM

Hi Greg,

> I first discovered alternate tuning a few weeks ago and I was very
> excited about the idea. I'll be graduating this year in EE from Cal
> Poly SLO, and while I haven't been studying music lately, I would
> like to build my senior project around music, since it is what has
> gotten me into the field. Creating a basic DSP seemed a little
> bland to me.
>
> I am thinking about making a device that converts an audio input
> signal into the just tuning for a key or keys defined on the fly by
> the user or automatically if possible. I think I should be able to
> do this with an FFT and look for impulses that do not match up with
> the correct pitches and multiples for whatever key I'm in, and
> cancel out them while creating new ones at the closest just pitch.

Wow. This is the first proposal I've heard for retuning music
after it's been made. Neat idea. I'd love to hear more about it.

> A big question I have is: does it make a big enough difference to
> make it worth the effort? I haven't seen any products like this,
> which is very suprising to me, so I figured there might be a reason
> why. Thanks

Check out

http://www.adaptune.com

(click on music)

It works on MIDI, before the music gets turned into sound.

-Carl