Date: Sun, 29 Jun 1997 10:48:36 +0100 From: Alan W Black <awb@cstr.ed.ac.uk> To: hasty@rah.star-gate.com Cc: Randall Hopper <rhh@ct.picker.com>, multimedia@FreeBSD.ORG, richard@cogsci.ed.ac.uk Subject: Re: Oregon's Speech Toolkit? Message-ID: <199706290948.KAA24046@margo> In-Reply-To: <8414.199706282248@pitcairn.cogsci.ed.ac.uk> References: <8414.199706282248@pitcairn.cogsci.ed.ac.uk>
next in thread | previous in thread | raw e-mail | index | archive | help
From: Amancio Hasty <hasty@rah.star-gate.com> > > > Tnks I don't have any problems with Festival . > > Care to ask your friend to send us the patches to make Festival work > with CSLU? > My understanding is that Festival uses scheme for its command interface > and that CSLU uses tcl/tk . > > Tnks, > Amancio > > >From The Desk Of Richard Tobin : > > > |Has anyone ported the University of Oregon's Speech Toolkit? > > > > > Is that "festival" you're talking about? > > > > No, Festival is from Edinburgh University. It's largely developed > > under FreeBSD so it's not likely to need porting. I share a house with > > the main developer (Alan Black) so I can pass on any problems... > > > > -- Richard [sorry coming in on this late so I may be repeating thing already said] The OGI toolkit 1.9 already distributes a version of the Festival Speech Synthesizer which works with their toolkit. Its not the default synthesizer but may be used with the GUI dialog builder using the command fcslurp rather than cslurp (you must start a festival server first -- see the information in their festival distribution). Note their distribution includes an American English voice as the default speaker. This includes an interface between the scheme interpreter and tcl and the reverse (just naively through strings) which is now also included in our latest version. I should add I've tried (so far unsuccessfully) to get the OGI toolkit to work under FreeBSD (2.2.1). It basically has many SYSV'isms in it making it difficult to get to work under SunOS4 too. Most of the real problems are little problems in config files rather than the code (such problems are difficult to trace amongst much autogeneration of files). Their audio system is rather limited too (it only really works on older Suns). I was visiting their group last week and they have redesigned their audio io to make it easier to support new machines and integrated a number of patches that should make it easier to port in their next version (due in August). OGI have been concentrating on a Windows port of the toolkit (and Festival) as it seems that that sort of thing is necessary in that part of the country :-). I have almost convinced them that they should be aiming at the free Unixes (Linux and FreeBSD) because of the number of skilled computer users who would definitely add to their tools, which is really why they are distributing it. The toolkit when running (we use it here under Solaris) is very easy to use and you can build telephone type dialog system with simple recognition and synthesis in a few minutes. The Festival Speech Synthesis System although included in the OGI toolkit is an independent system developed at Edinburgh. Festival 1.1.1 (http://www.cstr.ed.ac.uk/projects/festival.html) is available for non-commercial use for free in source form and runs on any Unix system I can get my hands on (Solaris, FreeBSD, Linux, HPUX, OSF (Alphas), SGI Irix and SunOS). We will be releasing the next version in August with many imporvements and an American English voice for those of you who don't understand British English :-). Alan Alan W Black email: awb@cstr.ed.ac.uk Centre for Speech Technology Research http://www.cstr.ed.ac.uk/~awb University of Edinburgh tel: (44) 131 650 2787 80 South Bridge, Edinburgh, UK fax: (44) 131 650 6351
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199706290948.KAA24046>