[Tts-project] more festival voices? [Was [DH] Planning to remove deprecated commands and compat levels in debhelper]

Paul Gevers elbrus at debian.org
Thu Nov 5 10:50:59 UTC 2015

Hi all,

I want your opinion,

On 03-11-15 12:19, Paul Gevers wrote:
> I will be migrating the voice packages to the TTS team.
> For the record, I am considering here:
> festlex-cmu
> festlex-oald
> festlex-poslex
> festvox-don
> festvox-ellpc11k
> festvox-kallpc16k
> festvox-kallpc8k
> festvox-kdlpc16k
> festvox-kdlpc8k
> festvox-rablpc16k
> festvox-rablpc8k

This has now been done. But while working on it I wondered about the

How important are these kind of voices for the accessibility community?
If deemed crucial, I am willing to spend time on improving the
situation. Like standardizing descriptions and work flow to allow more
voices (especially languages I guess) to enter.

Are the 8kHz voices still needed (looks like they may have been needed
in the past due to hardware requirements, but I expect all hardware
nowadays could run the 16kHz voices without any issue?

Upstream there a lot more voices (15 additional ones to the ones we
ship) which I expect to be of higher quality, should we package this and
improve our defaults?

Does anybody have an idea how to find additional languages (except for
just duck-duck-go-ing (the name festival doesn't help to target what I
am looking for).

Should this all be under the TTS umbrella?


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 473 bytes
Desc: OpenPGP digital signature
URL: <http://lists.alioth.debian.org/pipermail/tts-project/attachments/20151105/bbd30c0e/attachment-0001.sig>

More information about the Tts-project mailing list