[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
equal auditory steps along F1/F2
There is no way to set up an array of perceptually equidistant positions
in the F1/F2 plane. The problem is that in different languages there are
different categorization boundaries in this plane. Differences accross a
category boundary are alway perceived as larger as physically comparable
differences within a category. Being grown up with a certain language
makes us more sensitive to some steps, and less to others. I am not
aware if learning a second language in later life has at least some
influence on F1/F2 discrimination performance (I think it should).
Perhaps you could "calibrate" your vowel array, with listeners with the
same cultural background as those that you want to invite for your main
Dear List members,
I am designing an experiment and I need to synthesize an array of vowels
separated by equal perceptual magnitudes. That is to say, I need vowels
that sample the acoustic vowel space in equal auditory steps along the
continua F1 and F2.
Furthermore, I would like to justify such a magnitude, instead of just
doing a Hz to ERB/MEL/Bark etc. conversion, for instance.
I would appreciate any comment and suggestions about the relevant
Prof. Dr. Christian Kaernbach
Psychologische Methodik und computergestützte Modellierung
Institut für Psychologie
Schubertstr. 51 a