Voice prompts

[G0NHF]

Hi Tony

One simple way to add the feature you want, is to put some spaces at the end of your channel names

Roger, I will experiment. Not sure how many characters are available in the channel name and group name fields.
Some of my group names are quite long, for example

‘23580> W.MIDLAND’
‘23526> HUBNET-UK’
‘3026> CANADA ENG’
‘204> NETHERLANDS’
‘23550> SCOT CHAT’

So they are quite long names. Anyway I will try shortening a few and adding spaces to the end and see what happens.

Thanks
Tony

[m1dyp]

Guys'

I've generated a few more English versions

Amy sounds OK, except the letters R and E. I manually modified the "R" and its a bit better, but I've not had chance to fix the E

Emma seems generally OK, but I have not tested all the Alphabet

Nicole is an AU female voice and doesnt sound too bad

Brian is a British male voice and is also quite clear

I'll do some more tomorrow.

thank you, emma uk seems the best here, great work

[IU4LEG]

Sorry guys, I can't find the function to load the speech synthesis bin file into the new cps, Roger can you tell me where it loads from please?
Thankyou 73 de IU4LEG

[W1AEX]

Along the top menu ribbon, click on Extras. From the drop down menu, select OpenGD77 Support. At the bottom of the second column of buttons, select Write voice prompts.

[G0NHF]

I have been working on adding Voice prompts to the firmware for several weeks.

The Development release now supports voice prompts on most screens.

To use the Voice prompts you have to download a prompt file for your language

https://github.com/LibreDMR/voice-promp ... /R20200606

You also need the latest CPS version, which has the option to upload the voice prompt file

https://github.com/rogerclarkmelbourne/ ... ses/latest

And of course, you need the Development / Unstable version of the firmware on your radio


Details of how to create your own voice prompts are here https://github.com/LibreDMR/voice-prompts


Notes

(1) This is a work in progress, and I can only verify if the English language version sounds OK.
(2) Its very time consuming to download and compress the synthesised speech. Around 20 minutes per language / voice. Also the source of the synthesis has daily limits based on the user's IP address, so its not practical for me to update more than one or 2 languages per day.

I was hoping that languages other than english could be generated by native speakers of those languages, but I think I will need to do a video to explain what is required in order for people to produce their own voice prompt packs

Just updated everything and it’s all sounding much much better Roger. The quality of the prompts sound clearer to me. And the eight is great mate!

Is it just me, but I can’t detect any difference in the USA and UK variants? Anyway they sound very similar to me. It might just be me though just not hearing the accent change.

The only improvements I could suggest is 1 second wait before the voice prompts time out when selecting channels or groups and a welcome message when switching on.

Thanks
Tony

Hi Roger,

Would it be feasible to have a voice mode whereby the voice prompts for channels and groups were automatic, ie no requirement to press the black button at all? It would be very handy in a mobile situation to rely on the voice in a car for instance. Basically a mode where all voice prompts would work with minimal button presses.

Just a thought.

Ps. I love the new eight
Tony

[VK3KYY]

Yes.

But I need to discuss with Ian DJ0HF the potential confusion for visual impaired operators

[N6RXT]

Yes.

But I need to discuss with Ian DJ0HF the potential confusion for visual impaired operators

Now you have a few more blind users on the forum, so, you can get a few opinions on this
This is going to come down to more of a preference thing than a confusion thing.
having the speech read as you cycle through things isn't a big deal as long as it stops when selecting an option, or you have a way to stop it.
My concern over the long term is it might not be the best thing to dedicate a button just to speech. you may need that for functions in the future. You could for example, have things read as you select them, and the 1st press of the side key silences speech and the 2nd one performs its function.
Some of this may not come out until it gets used for a few weeks and by the masses.

[VK3KYY]

Voice prompt files for all English voices have been updated

https://github.com/LibreDMR/voice-promp ... ses/latest

UK. Amy, letters "R" and "E" have now been fixed by manually downloading "Neural" synthesis versions for those letters directly from Amazon Web Services. I also replaced the words with "Call" in them with Neural versions.

I also changed the acronym "dBM" to "D B M" for all English voices as not all voices pronounced this as individual letters.

I have also added an Australian male voice "Russel", which seems to work quite well.

And, I have reduced the letters and numbers speech speed for the French voice "Lea", because I am changing all voices to use normal speed for the letters and numbers, because although it takes longer to announce the Channel name etc, at least the letters and numbers are understandable.

[IU4LEG]

Along the top menu ribbon, click on Extras. From the drop down menu, select OpenGD77 Support. At the bottom of the second column of buttons, select Write voice prompts.

OH ok! thank you!

[VK3KYY]

Along the top menu ribbon, click on Extras. From the drop down menu, select OpenGD77 Support. At the bottom of the second column of buttons, select Write voice prompts.

OH ok! thank you!

You also need to enable it in the Sound Options: -> Prompt:Voice