New Script: Speak Song Title

Post a reply


In an effort to prevent automatic submissions, we require that you complete the following challenge.
Smilies
:D :) :( :o :-? 8) :lol: :x :P :oops: :cry: :evil: :roll: :wink:

BBCode is ON
[img] is ON
[flash] is OFF
[url] is ON
Smilies are ON

Topic review
   

Expand view Topic review: New Script: Speak Song Title

Lower volume instead of pause

by superluig164 » Wed Oct 19, 2016 9:22 am

Could you also add an option to lower the volume while speaking? I think that would sound really nice, especially when using MediaMonkey to play music out loud.

Thanks again!

Pronunciation

by superluig164 » Wed Oct 19, 2016 9:19 am

Do you think you could add some way to change pronunciation of some words, so for example if your program detects "R3hab" it would instead send "Rehab" to the TTS engine so that it pronounces the name correctly? I find that some of my artists don't pronounce correctly with Microsoft's TTS engine, for example it pronounces Mr. Fijiwiji as "Mr. Feejeeweyejeye."

Thank you!

Re: New Script: Speak Song Title

by proplane64 » Sat May 28, 2016 8:54 am

Does this still work? Getting VBS errors

Re: New Script: Speak Song Title

by Andreas Weichert » Thu Jun 12, 2014 2:29 am

Meander wrote:@ Andreas

Just spotted your update and testing it now. The random 'mix' of intros is working well.

For some reason the next track would start (very briefly) and then pause for the intro. But it sorted itself out after unchecking and rechecking the odd setting.

One observation: I had to copy the whole of each mask (plus separators) that I wanted to use into the Multi Test option. Is there a shorthand way of entering the mask references?

Thanks for the addition. :)
Fine that the random mix pleasures you.

I have connect the song start event from MM with my code. There is no other way.
If I get this event I pause the song (if you have enabled "pause song during speaking") and wait for the end of the speech. Then I disable the pause.
So if the computer resp. MM has not time to process the event instantly the song will be played for a short time.

What do you mean with entering the mask reference? The setting are storend in the MediaMonekey.ini
You can close MM, edit the ini-file and restart MM. Here an example

[SpeakSongTitle]
Enabled=1
StartSpeak=$title$ from $artist$ of album $album$ at year $year$
UnknownText=unknown
LangNr=1
Version=1
WithPlayPause=1
Rate=-3
Volume=100
ActiveMaskNr=2
StartSpeak_0_Name=Title with rating
StartSpeak_0_Mask=$rating$ stars for $title$|two|tree|four
StartSpeak_1_Name=Title with artist and album with pauses
StartSpeak_1_Mask=$title$ $P5$ $rating$ stars $P5$ from $artist$ $P5$ of album $album$
StartSpeak_2_Name=MultiTest
StartSpeak_2_Mask=one|two|tree|four
StartSpeak_3_Name=
StartSpeak_3_Mask=
StartSpeak_4_Name=
StartSpeak_4_Mask=
StartSpeak_5_Name=
StartSpeak_5_Mask=
StartSpeak_6_Name=
StartSpeak_6_Mask=
StartSpeak_7_Name=
StartSpeak_7_Mask=
StartSpeak_8_Name=
StartSpeak_8_Mask=
StartSpeak_9_Name=
StartSpeak_9_Mask=

Re: New Script: Speak Song Title

by Meander » Wed Jun 11, 2014 6:52 pm

@ Andreas

Just spotted your update and testing it now. The random 'mix' of intros is working well.

For some reason the next track would start (very briefly) and then pause for the intro. But it sorted itself out after unchecking and rechecking the odd setting.

One observation: I had to copy the whole of each mask (plus separators) that I wanted to use into the Multi Test option. Is there a shorthand way of entering the mask references?

Thanks for the addition. :)

Re: New Script: Speak Song Title

by Andreas Weichert » Wed Jun 11, 2014 1:03 pm

bailso wrote:Thanks for your quick response.

With ASIO and WASAPI I hear the announcement but the track does not play. No such problem with DirectSound.
Make a double-click on a track in the now-playing-list.
You hear the speech but at the end of the speaking the song does not start ?

Please disable "Pause song during speaking" ! What happens then ?

Re: New Script: Speak Song Title

by Andreas Weichert » Wed Jun 11, 2014 12:58 pm

Meander wrote:Andreas,

Your idea looks good, as you've already allowed for ten presets. So first phase might be to implement the script 'stepping' to the next (individually customised) mask between announcements - which would be great. A checkbox to enable/disable it would be good too. The random method could follow that if you think it practical.

BTW I have sapi4 (via the L&H TTS 3000 engine), as well as 5, on my machine. The voices for these are more easily obtained and are freely available. Some of the characters aren't too bad even though in a lower bitrate. Can you access those in the script?
I uploaded a new version.
With it you can edit multiple phrases in one entry. Separate them by "|" in the text.
Text1 | Text 2 | Text 3
Via random genrator one of the three will been taken. You can test it in its settings page too.
Simple press the test button and you will hear one of it.

About other voices I dont know much. My plugin demands all voices via SAPI interface and lists it. It should show all integrated from the windows system.

Re: New Script: Speak Song Title

by bailso » Thu May 29, 2014 5:47 pm

Thanks for your quick response.

With ASIO and WASAPI I hear the announcement but the track does not play. No such problem with DirectSound.

Re: New Script: Speak Song Title

by Andreas Weichert » Thu May 29, 2014 9:12 am

bailso wrote:I was able to download this addon. I use the ASIO output plugin which seems to stall when the Speak Song title is enabled.

Should the addon be compatible with ASIO and WASAPI (which I have also tried)? I only seem to be having success with Directsound.

Thanks
What do you mean with stall, what happens.

Re: New Script: Speak Song Title

by bailso » Thu May 29, 2014 3:01 am

I was able to download this addon. I use the ASIO output plugin which seems to stall when the Speak Song title is enabled.

Should the addon be compatible with ASIO and WASAPI (which I have also tried)? I only seem to be having success with Directsound.

Thanks

Re: New Script: Speak Song Title

by Meander » Wed May 21, 2014 10:09 am

Just to add.

In pursuit of more natural voices and alternative engines, I came across Festival.

It's written for Unix based systems, but can be run in windows with the aid of Cygwin.

Judging by the Technical Online demo, which can be accessed here...

http://www.cstr.ed.ac.uk/projects/festi ... oices.html

...it produces very natural character, with additional resources for creating and honing voices too.

Promising.

Re: New Script: Speak Song Title

by Meander » Tue May 20, 2014 2:35 pm

Andreas,

Your idea looks good, as you've already allowed for ten presets. So first phase might be to implement the script 'stepping' to the next (individually customised) mask between announcements - which would be great. A checkbox to enable/disable it would be good too. The random method could follow that if you think it practical.

BTW I have sapi4 (via the L&H TTS 3000 engine), as well as 5, on my machine. The voices for these are more easily obtained and are freely available. Some of the characters aren't too bad even though in a lower bitrate. Can you access those in the script?

Re: New Script: Speak Song Title

by Andreas Weichert » Tue May 20, 2014 1:18 pm

Fine that you like my script. A randomizing of the announced text would be nice idea.
Maybe a more full solution with complete changing of the sentence inclusive of the order of the mp3tags.
(This is additional more easy to integrate)

next is song xyz from abc
now you hear from abc the song xyz of the year xxxxx
and so on.

How man different task do think are sensefull aproximately ?

Re: New Script: Speak Song Title

by Meander » Tue May 20, 2014 8:04 am

Nice script - working well.

I'm currently using the following mask:

Next up $P4$ $title$ $P4$ from the album $album$ $P4$ by $artist$

...and I got to thinking it would be quite refreshing to randomise the intro. So, as well as "Next up", there could be "And now";"Here's" etc. These could be rolled up into a selectable preset called, say, 'Random Intro' for inserting into newly created masks.

The entries could be listed in the script directly (or maybe the ini entry or some other text based file for lookup) and simply cycled through each time a new track is played in a session - with, perhaps, some simple editing feature.

So, the new mask might look smething like this:

$Random Intro$ $P4$ $title$ $P4$ from the album $album$ $P4$ by $artist$

Is this feasible with the TTS engine?

Re: New Script: Speak Song Title

by rovingcowboy » Mon Nov 11, 2013 12:04 am

If they can't get the time spoken you could use ms agents to do this you would need to get all the needed software for them to work, then get the damp plug in to use as a DJ set up all the comments for spoken text like DJ's would do between songs, this would also allow you set the time tag in those comments and it would be said.
but all that is old tech and might have some hick ups on win 8. Still the information is in my post for broadcasting in this fourm. The link is number 22 on the monkeys helpful messages page in my signature

Top