Dear Sergio and fellow festival maintainers,
Sorry for the delay - I only bisected this yesterday due to breaking
of loading of some demo voices with an unhelpful error message.
I wish to revisit the solution for closing this bug.
The solution changes the behaviour of (voice.list).
In particular
In response to Paul's request for someone with more familiarity with flite
code base regarding
whether it is possible to access the patched function from a11y programs.
I have some knowledge of flite as I keep my own fork which mixes flite and
festival etc.
(It is not a replacement for flite or
Our git tree has been updated.
Upload is pending. (We are currently subject to a transition).
Dear All,
This email is to make it official. TTS will adopt festival-czech plus the
individual czech voices.
We will also adopt freebsoft utils which is written solely in festival's
internal dialect.
@Milan - Thanks for your reply regards default encoding. I now have the
voices working on the
closely with
the festival/spech-tools packages we maintain.
In a sense it is really the glue program between our two groups. I look
forward to your opinions.
I look forward to this continuing dialogue,
very best regards,
Peter Drysdale
of sound files from representative texts of Czech to act as a
Rosetta Stone for verifying
the voices during maintainence and to be included in our git repository.
Opinions Please?
best regards,
Peter Drysdale
In light of my desire to retire from the list of festival and speech-tools
uploaders I shall not
be able to act on this bug.
best regards,
Peter Drysdale
On Wed, May 29, 2013 at 8:56 AM, Peter Drysdale drysdalep...@gmail.comwrote:
In light of discussions with Ryan Kavanagh any solution
to understanding a solution
to the constraints of building
festival-doc from festival source.
best regards,
Peter Drysdale
Dear Samuel, Jean-Philippe, Sergio,
Unless Jean-Philippe or Sergio has an immediate objection I would like to
request
Samuel to sponsor an upload of speech-tools from the current git master.
The most important reason is to ensure compliance with gcc 4.8 which will
default in jessie later this
and try building git master with gcc 4.8.
Will update you all in the next 24 hours.
best regards,
Peter Drysdale
Dear Fellow Speech-tools maintainers,
My test sid image is updated to gcc 4.8 and I have run a build of git
master.
No build failure in master. The removal last year of esd from speech-tools
in git master
means the current version builds fine the gcc 4.8. I have done preliminary
testing with this
Dear All,
Sergio's feedback inspired me to push forward in looking at the
implementation scenarios
for building the doc binary packages from our source package. Our existing
texinfo build
dependency should make it easy enough to push out any of:- info, HTML or pdf
formats.
I am happy to do a
Dear Maintainer/Uploaders of festival and Users of festival-docs,
To clarify the situation. festival-doc currently contains a html and ps
(Postscript) version of
the Festival Manual. The version number is 1.4.2 and dates from 25th July
2001.
CMU (the current employer of Prof. Alan Black - one of
Dear Users of Festival-doc,
As one of the uploaders for festival and speech-tools I wish to start the
discussion about this recently
orphaned package.
I note that upstream has not updated the documentation since version 1.4 of
festival ( we are now up to version 2.1).
Is it really appropriate to
it accepted upstream. I have seen discussion
in the upstream support forum discussing encoding as an area where festival
can be improved.
Please report back if the czech voice maintainer are able to help you in
the first instance.
with best regards,
Peter Drysdale
Uploader for Debain festival
Dear Ubuntu6226,
I have forcibly merged your two bugs since they appear to relate the same
underlying issue with festival.
I presume from your later bug report 687822 that you have partly cured
665844. Is this true?
I have checked the files list for the festival deb for armel and it appears
to
This change has entered the Debian tts group's git repository and will be
uploaded
with the next upload of speech-tools.
best regards,
Peter
Attached is diff for debian/rules for wvstreams to fix this bug.
It has been runtime tested on two different machines.
Wvstreams has received an ITA (intent to adopt) since I first started
looking into
this bug. I presume the new maintainer (Bart Martens) will upload for this
bug in combination
Please find attached a new debian/rules file for wvstreams which appears to
fix this bug.
It is altered in two places to switch off some compiler optimizations.
It is obviously a hack.
This has been runtime tested on borrowed hardware.
This follows Fabrice's suggestions.
Proper diff and
Dear Chris,
Thanks for your testing and feedback.
Apologies about the patch being insufficient.
I have not had access to the affected hardware since my previous post.
I note work continues at Fedora on this bug which can be found at:
https://bugzilla.redhat.com/show_bug.cgi?id=812651
Instead
Dear Gregor,
Thank you for your email. It forced me to look further into Debian
procedures and policy. Between your email and details of Debian policy
I think I have found a way forward.
To avoid noise to your account I will not copy you to subsequent emails
regarding this bug where I can avoid
The purpose of this email is to set forth the reasons for raising the
severity of this bug
and describe my intentions.
1) I intend to increase the severity to release critical.
The Debian document regarding severity states:
critical makes unrelated software on the system (or the whole system)
Please find attached a draft patch refreshed in the Debian style.
This patch has passed build testing only on a Debian system.
It has passed testing on Fedora.
Run-time testing this patch on Debian will have to await either till I get
access to the hardware effected or another person affected by
?
with the very best regards,
Peter
Peter Drysdale
(Uploader for festival and speech-tools)
magic.patch
Description: Binary data
Dear Frank Leoffler (original bug opener), Javier Serrano Polo (contributor
to bug discussion), Jean-Philippe (maintainer of festival package),
I am making another attempt to close this bug.
For Frank and Javier's benefit - I am an uploader for the festival package.
This bug has been open since
Dear Matthias Klose, Cyril Brulebois and Jean-Philippe Mengual,
A patch for gcc 4.7 entered the speech-toools git tree on 2012-03-07.
The patch also enables clang 3.0 compiling of festival but not speech-tools.
Subsequently multiarch support was added by Samuel Thibault.
best regards,
Peter
:
Peter Drysdale wrote:
Dear Trent Buck and Festival users in Debian and Ubuntu,
I have applied a commit to the git repository for Debian festival,
currently
located at:
anonscm.debian.org/gitweb/?p=tts/festival.git
I can confirm this work with my favourite vox, with this change
Dear Trent Buck and Festival users in Debian and Ubuntu,
I have applied a commit to the git repository for Debian festival, currently
located at:
anonscm.debian.org/gitweb/?p=tts/festival.git
The patch provides a compatibility module (known as hts21_engine) to allow
festival 2.1 to use older
suggestion.
best regards,
Peter Drysdale
(uploader for festival)
Dear Sergio,
My testing agrees it is backward compatible.
Their appears to be merit in this patch whether of not upstream adopts it.
Thanks again. As you may know, upstream is not very active (does not
release often), although I will submit this patch once it gets into Debian
(if it is
Dear All,
I had a look at the code in this patch. It appears to follow analogously
from the existing
method by which festival finds voices. Extending this to languages does
appear to be
a natural extension.
My testing agrees it is backward compatible.
Their appears to be merit in this patch
Package: festival
Version: 2.1~release-1
Festival 2.1 comes with new integrated support for using clustergen
voices via the clustergen module.
The upstream original tar automatically includes this module when
building on a Debian system.
Festival 2.1~release-1 build system does not build this.
The equivalent bug #638691 in festival has changed from unclassified to
grave.
best regards,
Peter
This bug originated in a change in alsa-utils version of aplay.
The debian bug #638369 is the equivalent bug in alsa-utils.
For reference in Ubuntu the equivalent bug in festival is #778619.
best regards,
Peter
Dear Kingsley,
A little bit of tracking led me to discover the pause functionality in aplay
(form alsa-utils) was added last year.
This changing the old functionailty of aplay.
Takashi Iwai http://search.gmane.org/?author=Takashi+Iwaisort=datesuggested
a patch to preserve the old functionality
Dear All,
I can confirm that this has been discovered upstream and patched in the
master at the alsa project.
Version 1.0.24 of alsa-utils changed the default behaviour of aplay.
This has broken ( at least on some installations including mine) - the
festival package.
Jean-Phillipe Mengual and I
Dear Sergio,
To assist in studying your patch would you be able to clarify your
interaction with upstream.
I noticed that in Dec. 2010 you submitted a similar patch of languages.scm
to upstream via the festlang-talk
mailing list.
Would you be able to clarify if this was received favourably and
Dear Kingsley,
I was discussing this type of behaviour of festival with Kumar last week in
conjunction with my assisting in
the maintaining festival and mentioned that it appeared on my system but I
needed to test on different
hardware before filing a bug.
Like you this only occurred to me with
Yellow wrote:-
Festival rocks but I would like to notice that few information are given
regarding .festivalrc
I would like to that one user use :
festvox-rablpc16k
but however his festivalrc looks like this:
(Parameter.set 'Audio_Command aplay -q -c 1 -t raw -f s16 -r $SR $FILE)
(Parameter.set
39 matches
Mail list logo
