Voice prompts

VK3KYY
Posts: 7478
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: Voice prompts

Post by VK3KYY » Mon Jun 08, 2020 3:55 am

kd2lh wrote:
Mon Jun 08, 2020 3:43 am
Hi Roger, thanks for your work on Voice Prompts. All I can say is, "WOW!".

Is this a first for an amateur DMR HT? I know Motorola has it on commercial XPR7550a radios, but have not seen it on a DMR designed for ham use?

I have the American English voice loaded. Firmware at 6/3 unstable update.

The letter "T" is being pronounced as "D" by Matthew. Is this a vocoder issue?

Marc
Hi Mark

Yes. This is partially a vocoder problem, and also a sample rate problem, as the radio operates internally at a 8kHz audio sample rate, which gives 4kHz bandwidth at the very most.

However the problems can normally be resolved by modifying the original MP3 files a little.

I noticed some problems with "T" and "G" sounding similar, and at one point I manually edited "T.mp3" to improve this, but I have a feeling my manually edited file got overwritten when I re-downloaded all the prompts again, because of some other change that I made

I need to update my Python script which downloads the synthesised speech and does all the conversion and packing etc, so that I can mark individual prompts e.g. individual letters as having a manual override so that they don't get overwritten again

See https://github.com/LibreDMR/voice-prompts/ if you want to try building the voice pack yourself, its not that hard, you only need Python 3 and FFMPEG installed.

then run the script with the "-c config.csv" in the English_USA folder and it will download all the MP3's and do the rest.

BTW. To compress to AMBE it uses the firmware in the radio, so the GD-77 must be connected via USB



Anyway.

Let me know which letters don't sound any good and I can try a few tricks to improve this.

User avatar
IU4LEG
Posts: 191
Joined: Wed Nov 20, 2019 12:49 pm

Re: Voice prompts

Post by IU4LEG » Mon Jun 08, 2020 5:15 am

VK3KYY wrote:
Mon Jun 08, 2020 12:07 am
IU4LEG wrote:
Sun Jun 07, 2020 4:46 pm
Hi Roger,
Now that I have installed the firmware and the voice file on the radio, I realized that some words did not sound good and I made some changes to the Italian wordlist. I had also forgotten to translate "Point" to "Punto". Everything should sound better now.
I attach the new file.
Thank you, IU4LEG Fabio
You didn't put the translation for the display_background_colour
You are right, sorry. Now i added the display_background_colour translation.
Attachments
wordlist_italian.zip
(2.12 KiB) Downloaded 140 times

VK3KYY
Posts: 7478
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: Voice prompts

Post by VK3KYY » Mon Jun 08, 2020 6:55 am

OK.

I'll update the master copy and regenerate the Italian voices

User avatar
IU4LEG
Posts: 191
Joined: Wed Nov 20, 2019 12:49 pm

Re: Voice prompts

Post by IU4LEG » Mon Jun 08, 2020 7:43 am

thanks Roger, sorry for the forgetfulness

User avatar
OK2HAD
Posts: 52
Joined: Wed Jan 08, 2020 1:10 pm
Location: Uherský Brod (Czech Republic)

Re: Voice prompts

Post by OK2HAD » Mon Jun 08, 2020 8:54 am

Hi, I tried to correct the words in csv for the Czech Republic. I will gradually try Emma, Amy and Nikola. Then I'll let you know which one seems best.
František
Attachments
czech_voice1.zip
(2.51 KiB) Downloaded 127 times

VK3KYY
Posts: 7478
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: Voice prompts

Post by VK3KYY » Mon Jun 08, 2020 9:31 am

OK2HAD wrote:
Mon Jun 08, 2020 8:54 am
Hi, I tried to correct the words in csv for the Czech Republic. I will gradually try Emma, Amy and Nikola. Then I'll let you know which one seems best.
František
OK.

Let me know which voice works best.

It is possible to use phonemes, but the only TTSMP3 supports this, and there is a lower download limit on TTSMP3

Use of Phonemes is documented here

https://docs.aws.amazon.com/polly/lates ... honeme-tag

You can test phonemes on TTSMP3 e..g

<phoneme alphabet="ipa" ph="pɪˈkɑːn">pecan</phoneme>

Do not use the <Speak> tags

The wordlist data format has changed to allow phonemes

But I did not update all the languages yet with the new format.

English has been updated

https://github.com/LibreDMR/voice-promp ... ish_UK.csv

the 2 new columns PromptSpeechPrefix and PromptSpeechPostfix are for phoneme and other XML

e.g.

For <phoneme alphabet="ipa" ph="pɪˈkɑːn">pecan</phoneme>

PromptSpeechPrefix = <phoneme alphabet="ipa" ph="pɪˈkɑːn">
PromptSpeechPostfix = </phoneme>


However I think phonemes may be hard to use, so just changing the spelling of the words to make the speech sound correct is probably easier

VK3KYY
Posts: 7478
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: Voice prompts

Post by VK3KYY » Mon Jun 08, 2020 10:13 am

OK2HAD wrote:
Mon Jun 08, 2020 8:54 am
Hi, I tried to correct the words in csv for the Czech Republic. I will gradually try Emma, Amy and Nikola. Then I'll let you know which one seems best.
František
I updated the master copy with your document and the changes to the columns

Note the "Space" is actually the silence between letters, and is not any text,, it is " "

VK3KYY
Posts: 7478
Joined: Sat Nov 16, 2019 3:25 am
Location: Melbourne, Australia

Re: Voice prompts

Post by VK3KYY » Mon Jun 08, 2020 10:32 am

Guys

I've added 4 new American voices, for testing and evaluation. Joanna, Kimberly and Kendra (female) and Joey (male)
For the existing American voice "Matthew" I have removed the emphasis on the word "On" as in Calibration:On, as emphasis on this word only seems to be needed for some UK English voices


I've also updated the Italian voices with the changes from IU4LEG

See

https://github.com/LibreDMR/voice-promp ... /R20200608

User avatar
m1dyp
Posts: 601
Joined: Sat Nov 16, 2019 8:03 am
Location: Hertfordshire, U.K.
Contact:

Re: Voice prompts

Post by m1dyp » Mon Jun 08, 2020 10:52 am

thank you

User avatar
OK2HAD
Posts: 52
Joined: Wed Jan 08, 2020 1:10 pm
Location: Uherský Brod (Czech Republic)

Re: Voice prompts

Post by OK2HAD » Mon Jun 08, 2020 11:10 am

VK3KYY wrote:
Mon Jun 08, 2020 9:31 am
OK2HAD wrote:
Mon Jun 08, 2020 8:54 am
Hi, I tried to correct the words in csv for the Czech Republic. I will gradually try Emma, Amy and Nikola. Then I'll let you know which one seems best.
František
OK.

Let me know which voice works best.

It is possible to use phonemes, but the only TTSMP3 supports this, and there is a lower download limit on TTSMP3

Use of Phonemes is documented here

https://docs.aws.amazon.com/polly/lates ... honeme-tag

You can test phonemes on TTSMP3 e..g

<phoneme alphabet="ipa" ph="pɪˈkɑːn">pecan</phoneme>

Do not use the <Speak> tags

The wordlist data format has changed to allow phonemes

But I did not update all the languages yet with the new format.

English has been updated

https://github.com/LibreDMR/voice-promp ... ish_UK.csv

the 2 new columns PromptSpeechPrefix and PromptSpeechPostfix are for phoneme and other XML

e.g.

For <phoneme alphabet="ipa" ph="pɪˈkɑːn">pecan</phoneme>

PromptSpeechPrefix = <phoneme alphabet="ipa" ph="pɪˈkɑːn">
PromptSpeechPostfix = </phoneme>


However I think phonemes may be hard to use, so just changing the spelling of the words to make the speech sound correct is probably easier
Hi, it will be easier for me to edit the words in the text to make them sound believable. On "ttsmp3" I tried a few phrases with different voices. I found the best pronunciation for "Polish / Jan". I will have to adjust the numbers to match the Czech language. Give me a moment and upload a new pronunciation correction file for the pronunciation. It will use Tbin Polish / Jan for the final output of the .bin file. I can't create a .bin file myself. Will it be enough for you if I just edit the text?

Post Reply