Vocalist.org archive


From:  Greypins@a...
Greypins@a...
Date:  Mon Jan 22, 2001  6:01 am
Subject:  nasality and gram50


dear everyone,

this evening i discovered how to use the gram50 program i had downloaded
a month ago. i have to warn you, i wouldn't swear on my mother's grave that
i did it right (she's not dead yet, i mean the one she'll probably have in
the very distant future. call it a 'spec' oath). anyway, after the
discussion of all that nasal, non-nasal, open velum with raised palate (i
guess that would make it possible to sing well while making milk come out of
your nose at the same time), i decided to test out the various approaches
we'd been discussing on the gram50.

the samples i tested were all on 'ah' as in barf (boston pron.) followed
by 'ay' as in schnee (one year of high school german pron.). one sample was
recorded with classical singer breath managment, one with normal sound and
'what the hell is support?' breath managment and the last with a nasalization
similar to when one says 'uh-huh' (nasalized).

these are the results i got. the numerical ranges indicate the
presence of activity measured in hz.

classical support 'ah' 0-1240 gap 2230-3675 / 'ay' 0-3800 no gap

'ah' with 'as found' approach 0-1450 gap 2470-3565 / 'ay' 0-915 gap
1300-3630

nasal 'uh-huh' style 'ah' 0-1260 gap 2680-3630 / 'ay' was a
little odd. the

attack was nearly solid from 0-3670 but then immediately went to 0-850 gap
1475-1690 gap 2315-2530 gap 2960-3675


from what little i understand of this type of analysis, it seems that
the 'singer's formant' is present in all the examples as is the fundamental.
the readout is similar to the types of readings i have seen in richard
miller's books so i think i must have everything plugged in right. as i
listen to the playbacks however, they become (with each change in technique)
significantly quieter. isn't there also a power spectrum that plays a role
in the analysis of the voice that has bearing on whether a voice is easily
audible or not?

my apologies for the form my report has taken (it brings back memories
of kindergarten) but, this is my first attempt at spectral analysis and i
have to admit, i have no idea what the hell i'm doing.

happy monday,
mike



emusic.com