[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: speech database

Dear Joe,

I second Dick on this. It all depends on what you want to do, but often, speech databases don't include much information about the speakers (no age, or height). This is one of the reasons why we (and others) have used STRAIGHT so much in our research.


On 25 May 2016 at 06:13, Richard F. Lyon <dicklyon@xxxxxxx> wrote:
For parametric synthesis, STRAIGHT is a good package.


On Tue, May 24, 2016 at 2:55 AM, Sollini, Joseph <j.sollini@xxxxxxxxx> wrote:

Dear list,


I am looking for recommendations for speech databases of English speakers, containing male and female voices of different ages.  Ideally the voices would cover large range of speakers with both a range of similar and dissimilar sounding speakers (i.e. across different pitch, vocal tract length, formant frequencies).  Alternatively does anyone know of some realistic speech generation software that allows manipulation of these parameters and is suitable for generating large stimulus batteries?  Basically I’m hoping to have a database of speakers that allows me to (as much as possible) parametrise the characteristics of speakers voices.


I have found a few potential databases and speech generation packages but I’d really love to know which ones people prefer (and to know which ones I’ve missed).


All help welcome!






Dr Joseph Sollini

Post-doctoral researcher

UCL Ear Institute