Adding new voice (new language) for OSM Scout?

Hello!

I noticed there is no voice in Finnish for OSM Scout. Is there any chance to add Finnish voice from somewhere or do we need even a volunteer to record voice? If this application needs a volunteer maybe I’m ready to try but then I need some guidance. Sorry my bad English.

1 Like

I think your text to speech engine has to support it. You can install any text to speech engine with finnish i guess

2 Likes

There are no Finnish (or Swedish) voice for OSM Scout. I once started trying to create Swedish voice, but I’m afraid the whole voice project used in OSM Scout is dead. At least the important links are.

1 Like

I looked that VoiceOfMarble. It looks like there wasn’t any action in many years. Well, have to accept this until further notice.

Hi guys. Yes, Voice of Marble project seems to be really dead. I just serve copy of the samples from my server and OSM Scout is downloading them when user select some. It should not be so hard to record new voice, or generate it by some speech model. But I also considering to switch to some text to speech neural network in future, it would allow custom phrases…

4 Likes

Do you mean it maybe would be possible to make new voices without Voice Of Marble?

Go check the link @karry posted, download one of the directories and make a new recordings for every file in there. After that I assume it is just dropping the files to your phone in right directory and maybe editing some file for the meta data.

2 Likes

I have to say now I’m silly because maybe I still didn’t understand correctly. My bad :laughing:

If I understand right, this way might be one example to try to add voice: a voice maker downloads those samples, replaces samples by own new samples (keeping filenames?) and sends them to OSM Scout developer. A developer possibly can add samples to collection. I give up if I’m still clueless. I admit I want so much to help developers and other Finnish SF os users that I want really try to do something. At least I have participated in the translation work at the moment and of course voice isn’t precondition to use this application.

This is the way… :wink:

2 Likes

I’m not 100% sure it will work (didn’t try), but if you want to test your new voice files, and don’t want to dig into app code, you probably can activate any other voice from settings and simply replace those audio files with your own.
You can find the files in /home/defaultuser/.local/share/cz.karry.osmscout/OSMScout/voices.

1 Like

Thank you. I can try this when I have time

Well, I couldn’t find files that way. I mean after folder “share” I found nothing OSM Scout folders/files

First, you’ll need to download and activate at least one voice in OSM Scout’s settings. After that you should see something like the below screen. Please use File Browser from OpenRepos. The one from Jolla Store is somewhat crippled.

Check the underlined path.

1 Like

Thanks! This helped very much

I raplaced 2 voice files and I have to say they work fine.

1 Like

Yes. that’s what I thought. Great job! Now you just have to finish up and send the files to @karry , and I’m sure he will implement your voice into the app. :slightly_smiling_face:

2 Likes

Eson is right. Voice instructions are just assembled from these short recordings. When you will send me new package, I will add it to my server and create new entry in static json: https://osmscout.karry.cz/voices/list.json It is easy (for me).

2 Likes

Sounds good. I’ll send when I’m ready (soon I hope). Then we shall see what kind of result it will be.