Reply by Jerry Avins May 16, 20052005-05-16
Manuel M wrote:
> axlq wrote: > >>smuglr <d.mcgilvray@elec.gla.ac.uk> wrote: >> >>>I know the general subject of pitch detection has been flogged >>>to death, but I am looking specifically for pitch detection in >>>a sung melody. >> >>I'm wondering how that will work with some opera singers, who >>have such a heavy vibrato in their voice that there's no >>detectable constant pitch, rather a modulated frequency. >> >>(While I like opera, I also never understood why they need to >>sing that way. Can't they hold a note?) > > > > Even though that last question looks like you don't really expect > an answer, I've found it very interesting. As I'm both a music > aficionado (amateur choir singer) and an engineering > student, I'll try to give a rough answer in the DSP domain: > > Opera singers have to sing without any amplification and be > heard over a full orchestra (often with percussion, brass, > strings and everything). If we consider the orchestra as > undesired interference (a very self-centered singer would), > we find that we have a high-power interference, with peaks > in the frequencies of interest (since the voice must be in > tune with the orchestra). > > In order to have their voice clearly discerned from the > instruments, they make it sound different by rapidly varying > pitch. This creates an effect that allows your ear to separate > that voice from the rest of sounds even though each of its > frequency components overlaps with some harmonics from the > instruments. > > The trick there is that *all* of the frequency components of > the voice change in the same way at the same time. > That is what allows you to discern the voice so clearly, and > it would not be possible if he/she were singing in a constant > pitch, like we choir-singers often try to do.
I have it in a round-about way (through my sister Styra) that Norman Pickering largely agrees with you. I couldn't ask for better authority. It's not just voice you know, but string instruments organs and others that use vibrato, largely for the reasons you cite. Jerry -- Engineering is the art of making what you want from things you can get. &#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;&#4294967295;
Reply by Richard Owlett May 16, 20052005-05-16
andrew queisser wrote:

> Thanks for that interesting explanation [snipped desc of why vibrato]
>- I'd never heard that before but it
> makes perfect sense. I tend to like "cleaner" voices with less vibrato but > notice that the more dramatic roles that go with heavier orchestration go > with more vibrato. > > So we need some kind of correlation algorithm that looks for a modulated > central frequency.
> There are probably specific overtones in the human voice as well. Yes. Google for "formants" One of the first hits is http://hyperphysics.phy-astr.gsu.edu/hbase/music/vowel.html which, with associated links, looks very good. Caution: I'm a DSP and voice issues neophyte. {I've been enjoined from identifying self as DSP newbie ;]
> > Andrew > >
Reply by andrew queisser May 16, 20052005-05-16
Thanks for that interesting explanation - I'd never heard that before but it 
makes perfect sense. I tend to like "cleaner" voices with less vibrato but 
notice that the more dramatic roles that go with heavier orchestration go 
with more vibrato.

So we need some kind of correlation algorithm that looks for a modulated 
central frequency. There are probably specific overtones in the human voice 
as well.

Andrew

"Manuel M" <manuel.matias@gmail.com> wrote in message 
news:1116081530.338388.62050@z14g2000cwz.googlegroups.com...
> axlq wrote: >> smuglr <d.mcgilvray@elec.gla.ac.uk> wrote: >> >I know the general subject of pitch detection has been flogged >> >to death, but I am looking specifically for pitch detection in >> >a sung melody. >> >> I'm wondering how that will work with some opera singers, who >> have such a heavy vibrato in their voice that there's no >> detectable constant pitch, rather a modulated frequency. >> >> (While I like opera, I also never understood why they need to >> sing that way. Can't they hold a note?) > > > Even though that last question looks like you don't really expect > an answer, I've found it very interesting. As I'm both a music > aficionado (amateur choir singer) and an engineering > student, I'll try to give a rough answer in the DSP domain: > > Opera singers have to sing without any amplification and be > heard over a full orchestra (often with percussion, brass, > strings and everything). If we consider the orchestra as > undesired interference (a very self-centered singer would), > we find that we have a high-power interference, with peaks > in the frequencies of interest (since the voice must be in > tune with the orchestra). > > In order to have their voice clearly discerned from the > instruments, they make it sound different by rapidly varying > pitch. This creates an effect that allows your ear to separate > that voice from the rest of sounds even though each of its > frequency components overlaps with some harmonics from the > instruments. > > The trick there is that *all* of the frequency components of > the voice change in the same way at the same time. > That is what allows you to discern the voice so clearly, and > it would not be possible if he/she were singing in a constant > pitch, like we choir-singers often try to do. >
Reply by Manuel M May 14, 20052005-05-14
axlq wrote:
> smuglr <d.mcgilvray@elec.gla.ac.uk> wrote: > >I know the general subject of pitch detection has been flogged > >to death, but I am looking specifically for pitch detection in > >a sung melody. > > I'm wondering how that will work with some opera singers, who > have such a heavy vibrato in their voice that there's no > detectable constant pitch, rather a modulated frequency. > > (While I like opera, I also never understood why they need to > sing that way. Can't they hold a note?)
Even though that last question looks like you don't really expect an answer, I've found it very interesting. As I'm both a music aficionado (amateur choir singer) and an engineering student, I'll try to give a rough answer in the DSP domain: Opera singers have to sing without any amplification and be heard over a full orchestra (often with percussion, brass, strings and everything). If we consider the orchestra as undesired interference (a very self-centered singer would), we find that we have a high-power interference, with peaks in the frequencies of interest (since the voice must be in tune with the orchestra). In order to have their voice clearly discerned from the instruments, they make it sound different by rapidly varying pitch. This creates an effect that allows your ear to separate that voice from the rest of sounds even though each of its frequency components overlaps with some harmonics from the instruments. The trick there is that *all* of the frequency components of the voice change in the same way at the same time. That is what allows you to discern the voice so clearly, and it would not be possible if he/she were singing in a constant pitch, like we choir-singers often try to do.