15 Apr
2024
15 Apr
'24
8:48 p.m.
Hello all,
I'm working on an improved version if zita-at1 which most of you
probably know as the x42-autotune plugin. The update, zita-at2,
will preserve formants while retuning.
To test and develop this I need some clean vocal tracks, in
particular of female singers and also very low (bass) males.
So if you have these available, I'd be very happy if you can
share them. They won't be used of course for any other purpose
than to improve the retuner algorithm.
TIA for anything you can provide !
--
FA
15 Apr
15 Apr
8:52 p.m.
Hi Fons,
I can upload a track or two for you. That should really get your algorithm
working. :)
Best wishes,
Jeanette
--
* Website: http://juliencoder.de - for summer is a state of sound
* Youtube: https://www.youtube.com/channel/UCMS4rfGrTwz8W7jhC1Jnv7g
* Audiobombs: https://www.audiobombs.com/users/jeanette_c
* GitHub: https://github.com/jeanette-c
And when you say those words
It's the sweetest thing I've ever heard <3
(Britney Spears)
25 Apr
25 Apr
8:02 a.m.
Hi Fons,
On 15/04/2024 20:48, Fons Adriaensen wrote:
Hello all,
I'm working on an improved version if zita-at1 which most of you
probably know as the x42-autotune plugin. The update, zita-at2,
will preserve formants while retuning.
To test and develop this I need some clean vocal tracks, in
particular of female singers and also very low (bass) males.
I think I can provide you with some female voice clips from a while ago
but they are a bit short.
Any ideal length of the phrases / clips?
For the male voices, how low is 'low'?
In general what do you define as 'clean' (e.g. I'm imagining mono
relatively closed mic with relatively little background noise?)
Also I'm imagining one singer (i.e. not choir / multiple people)?
Are song parts with words/lyrics ok or just 'aaahs' or 'ooohs' preferred
/ relevant?
I might have some of these to send and or quickly produce but might be
helpful to know any preferred specs ;-)
Lorenzo
9:36 a.m.
On Thu, Apr 25, 2024 at 08:02:19AM +0200, Lorenzo Sutton wrote:
I think I can provide you with some female voice clips
from a while ago but
they are a bit short.
Any ideal length of the phrases / clips?
Anything 10s or longer.
For the male voices, how low is 'low'?
The real 'bass' register, down to F2 or so.
In general what do you define as 'clean' (e.g.
I'm imagining mono relatively
closed mic with relatively little background noise?)
No reverb, the less background noise the better.
Also I'm imagining one singer (i.e. not choir /
multiple people)?
Yes. But I once used zita-at1 on a choir, and to my surprise it worked
quite welll...
Are song parts with words/lyrics ok or just
'aaahs' or 'ooohs' preferred /
relevant?
Both are useful.
I might have some of these to send and or quickly
produce but might be
helpful to know any preferred specs ;-)
Thanks for responding !
Ciao,
--
FA
8 May
8 May
8:57 a.m.
Hi Fons,
Took a while to catch-up with threads...
On 25/04/2024 09:36, Fons Adriaensen wrote:
I know 'someone' who is more of a baritone but can reach G# maybe G...
and I assume you're interested in the spectral qualities more than
actual performance quality right? If so could it still be interesting? :-)
Lorenzo
On Thu, Apr 25, 2024 at 08:02:19AM +0200, Lorenzo
Sutton wrote:
Alas my female-voice clips are quite a bit shorter. They were recorded
as basic material for a video 'sonification' I did years ago and are
just rather short phrases (repeated in some case like doing different
'takes') from a well-known Purcell aria. They are relatively clean
(quite close mic in a room) and unedited (no added reverbs or fx).
I'm still happy to send them for consideration though :-)
I think I can provide you with some female voice
clips from a while ago but
they are a bit short.
Any ideal length of the phrases / clips?
Anything 10s or longer. For the male voices, how low is 'low'?
The real 'bass' register, down to F2 or so.
25 Apr
25 Apr
10:08 a.m.
On Thu, Apr 25, 2024 at 08:02:19AM +0200, Lorenzo Sutton wrote:
A bit more info on some of the topics:
Also I'm imagining one singer (i.e. not choir /
multiple people)?
The occasion I referred to was a childrens choir, and they had
recorded a four-part piece, each part separately. Without taking
care of consistent tuning or tempo. It was really an effort well
beyond their capability.
I spent most of a day retuning little pieces and re-aligning them
in Ardour. Managed to get something 'presentable' that the director
was happy with in the end.
Are song parts with words/lyrics ok or just
'aaahs' or 'ooohs' preferred /
relevant?
Some vowels can have harmonics that are much stronger than
the fundamental, resulting in strong peaks in the autocorrelation
(used for pitch detection) corresponding to an harmonic. But the
autocorrelation also always has peaks at subharmonics of the
fundamental. So you can end up with very ambiguous hints at
what the actual fundamental is. Long vowels can be very useful
to test and optimise how this is handled.
Ciao,
--
FA
227
days inactive
250
days old
linux-audio-dev@lists.linuxaudio.org
5 comments
3 participants
participants (3)
-
Fons Adriaensen -
Jeanette C. -
Lorenzo Sutton