Author Topic: Version 0.890 (Read 15928 times)

jitterjames · « **on:** May 22, 2011, 04:38:47 PM »

It's about time for a new version right? Well here it is.

The BIG change is that Vox is now much smarter about what needs to be rebuilt whenever you make a change to your command tree, or use GenXML. It will only rebuild the grammars that it needs to. In case it misses something, there is a "purge" option on the file menu that will clear the cache completely. Combined with multithreaded rebuilding, you should spend a lot less time waiting for VoxCommando to rebuild your commands when you are experimenting.

The problem with apostrophes ' in iTunes should be fixed.

GenXML is changing as you can see in the images in the next post. There is a new interface, you can choose where to save your generated files. You can also limit which songs will be scanned in XBMC and MediaMonkey using SQL, and both XBMC and MediaMonkey now generate Songs, Songs by Artist and Songs by Album payloadFiles.

PayloadXML files using the "song by artist" or "song by album" technique will store phrases that look like this:

value = songID#
phrase = SongTitleHere&&ArtistorAlbumNameHere

To control how Vox works with these phrases you can now access a properties window in the command editor for each payloadXML file (just double-click). See the attached images.

EDIT: the latest version is now on the main download page: http://voxcommando.com/downloads.asp

jitterjames · « **Reply #1 on:** May 22, 2011, 04:39:41 PM »

here are the new images for GenXML:

Wanilton · « **Reply #2 on:** May 22, 2011, 05:13:16 PM »

Very good, install worked fine here, thanks..

Wanilton

Kalle · « **Reply #3 on:** May 22, 2011, 05:20:54 PM »

Quote from: jitterjames on May 22, 2011, 04:38:47 PM

It's about time for a new version right? Well here it is.

The BIG change is that Vox is now much smarter about what needs to be rebuilt whenever you make a change to your command tree, or use GenXML. It will only rebuild the grammars that it needs to. In case it misses something, there is a "purge" option on the file menu that will clear the cache completely. Combined with multithreaded rebuilding, you should spend a lot less time waiting for VoxCommando to rebuild your commands when you are experimenting.

The problem with apostrophes ' in iTunes should be fixed.

GenXML is changing as you can see in the images in the next post. There is a new interface, you can choose where to save your generated files. You can also limit which songs will be scanned in XBMC and MediaMonkey using SQL, and both XBMC and MediaMonkey now generate Songs, Songs by Artist and Songs by Album payloadFiles.

PayloadXML files using the "song by artist" or "song by album" technique will store phrases that look like this:

value = songID#
phrase = SongTitleHere&&ArtistorAlbumNameHere

To control how Vox works with these phrases you can now access a properties window in the command editor for each payloadXML file (just double-click). See the attached images.

perfect, install and update works fine. Generate XML for XBMC and MM works also!
(when I click in the new GenXML editor on "save settings" it works, but better is (for me) "save and close window" (it is only cosmetic

)

Good work James

Kalle

Wanilton · « **Reply #4 on:** May 23, 2011, 07:53:37 PM »

James,

Appears one problem in this new version 0.890 for me:
Turn On the computer, I receive one message inform: Your OS not support Speech Recognition (my OS win7 64 bits English), and it´s crash voxcommando ...
I Close voxcommando, I start again manually voxcommando and work fine, no problem...
restart OS again, same problem...crash vox, appear message...

Wanilton

jitterjames · « **Reply #5 on:** May 23, 2011, 09:20:17 PM »

OK. Thanks. I'll try to figure out what is going on.

Kalle · « **Reply #6 on:** May 24, 2011, 06:49:15 AM »

Quote from: jitterjames on May 23, 2011, 09:20:17 PM

OK. Thanks. I'll try to figure out what is going on.

Hi James,
I found out one important thing. When I wanted to play newly added music in MediaMonkey with the command "play music by name", I had to change first the command "MM.LoadBySQL" see pictures, otherwise the music was not played.
That will happen only with newly added music.

Kalle

jitterjames · « **Reply #7 on:** May 24, 2011, 08:05:29 AM »

Kalle: Can you give an example of a song that doesn't work?

I tried with a clean install, did a full GenXML scan for MediaMonkey and

MM.LoadBySQL -- SongTitle Like "{1}" order by random() limit 200

works fine.

the only ones that don't work are ones with double quotes in them.

e.g. "Heroes"

If it still doesn't work, maybe you can send an image of your command history so I can try to figure out why it's not working for you?

Kalle · « **Reply #8 on:** May 24, 2011, 08:36:12 AM »

Quote from: jitterjames on May 24, 2011, 08:05:29 AM

Kalle: Can you give an example of a song that doesn't work?

I tried with a clean install, did a full GenXML scan for MediaMonkey and

MM.LoadBySQL -- SongTitle Like "{1}" order by random() limit 200

works fine.

the only ones that don't work are ones with double quotes in them.

e.g. "Heroes"

If it still doesn't work, maybe you can send an image of your command history so I can try to figure out why it's not working for you?

James, I have fixed the problem with changing the parameter in the "MM.LoadBySQL" from "Song Title Like '{1}' order by random() limit 200" to "Song Title Like "%{1}%" order by random() limit 200", thats all, work fine.

Quote

Kalle: Can you give an example of a song that doesn't work?

all newly added songs

Quote

I tried with a clean install, did a full GenXML scan for MediaMonkey and

MM.LoadBySQL -- SongTitle Like "{1}" order by random() limit 200

works fine.

It was a update-install over the Vers. 0.887 and the new music that I added has no quotes in title (New Soundtrack from Pirates of the Caribbean). I will test them with a clean install on a other PC

What I mean is, that in Vers.0.887 was the Parameter for MM.LoadBySQL -> '{1}' in the Vers. 0.890 -> "%{1}%"

jitterjames · « **Reply #9 on:** May 24, 2011, 08:49:09 AM »

the problem with using Like %{1}% is that if you want to listen to a song that is actually called "Love" you will get all songs that contain the word Love. Maybe you want this but some people don't.

in fact I have found this works fine:

MM.LoadBySQL -- SongTitle='{1}'

Kalle · « **Reply #10 on:** May 24, 2011, 09:01:45 AM »

Quote from: jitterjames on May 24, 2011, 08:49:09 AM

the problem with using Like %{1}% is that if you want to listen to a song that is actually called "Love" you will get all songs that contain the word Love. Maybe you want this but some people don't.

in fact I have found this works fine:

MM.LoadBySQL -- SongTitle='{1}'

I will test all full new install, and I have the same error as Wanilton, "Your OS does not support ...."

jitterjames · « **Reply #11 on:** May 24, 2011, 09:08:48 AM »

Quote from: Kalle on May 24, 2011, 09:01:45 AM

I will test all full new install, and I have the same error as Wanilton, "Your OS does not support ...."

Yes, I was able to recreate this and I have found the solution. I will post an update later today that fixes the reboot problem. thanks.

Kalle · « **Reply #12 on:** May 24, 2011, 09:16:43 AM »

Quote from: jitterjames on May 24, 2011, 09:08:48 AM

Yes, I was able to recreate this and I have found the solution. I will post an update later today that fixes the reboot problem. thanks.

Thanks, fine

Kalle · « **Reply #13 on:** May 24, 2011, 02:31:15 PM »

Quote from: Kalle on May 24, 2011, 09:01:45 AM

I will test all full new install, and I have the same error as Wanilton, "Your OS does not support ...."

James has solved my problem with MediaMonkey, there were spaces at the end of the song name.

Kalle

PS: Hey James, I've seen that you've done in the LCB my proposal for the skip if-block. perfect!!!

Did you watching the new THX trailer?

jitterjames · « **Reply #14 on:** May 24, 2011, 02:54:34 PM »

Oh, yeah, I forgot about that! That was a while ago.

I just watched the THX trailer that you modified and it looks great. Unfortunately I am not allowed to use the stereo at the moment...

I will check it out later. I downloaded a bunch of HD (1080p) intros in Dolby 5.1 that I am looking forward to testing after my wife finishes her final paper for the semester!

Kalle · « **Reply #15 on:** May 24, 2011, 03:12:21 PM »

Quote from: jitterjames on May 24, 2011, 02:54:34 PM

Oh, yeah, I forgot about that! That was a while ago.

I just watched the THX trailer that you modified and it looks great. Unfortunately I am not allowed to use the stereo at the moment...

I will check it out later. I downloaded a bunch of HD (1080p) intros in Dolby 5.1 that I am looking forward to testing after my wife finishes her final paper for the semester!

haha, I understand that otherwise there is trouble with your wife. The THX trailer full HD (1080p) with the VC-branding, looks very cool and sounds great (Dolby 5.1).

Kalle

Kalle · « **Reply #16 on:** May 26, 2011, 10:29:15 AM »

Quote

Post from James - EDIT: I fixed the language bug discussed below.

Works perfect!

Kalle

jitterjames · « **Reply #17 on:** May 27, 2011, 05:44:37 PM »

Here are sample command for XBMC

"play song by artist"

and

"play song from album"

I recommend you use only one of the 3 possible song commands:
"play song", "play song by artist", or "play song from album" since these use big grammars. Otherwise your load times and memory use will go up.

To trigger these commands you can say the following:

Song by artist:
- Play song (part of song name) by (part of artist name)
- Play song (part of song name) by anyone

Song from album:
- Play song (part of song name) from (part of album name)
- Play song (part of song name) from any album

(double click the payloadXML in command editor to change options for "by" and "anyone" / "any album"

jitterjames · « **Reply #18 on:** May 27, 2011, 05:54:00 PM »

and here is the same thing for MediaMonkey

Wanilton · « **Reply #19 on:** May 27, 2011, 06:05:05 PM »

Thanks James

jitterjames · « **Reply #20 on:** May 27, 2011, 06:19:57 PM »

I almost forgot to mention. You can now specify where your XBMC payload files will be saved, and I have chosen to save mine in ..\XbmcPayloads

They used to be saved directly in the VC folder.

If you are still saving them directly to the VC folder then you'll need to adjust the path for the payload files in the commands that I just uploaded.

jitterjames · « **Reply #21 on:** June 01, 2011, 08:00:53 AM »

I just uploaded 0.891 to the download page:

http://voxcommando.com/downloads.asp

- there is a minor change to the genXML XBMC (it didn't deal well with empty folder path)
- I have added the ability to make any command "always on". Any existing "pay attention" command should automatically be detected and marked as always on. Commands that are always on will work (without a prefix) even if you are in standby more. They won't work if vox if OFF - red. In the command editor "always on" commands should be shown in bold text.

jitterjames · « **Reply #22 on:** June 04, 2011, 09:36:37 AM »

there was still a bug in 0.891

I have updated to 0.892 on the downloads page:
http://voxcommando.com/downloads.asp

I have also updated the genXML commands so that they will now automatically update and restart (in the background).

jitterjames · « **Reply #23 on:** June 07, 2011, 05:30:01 PM »

I made this for another thread, but it may be useful for others. How to import groups (that have been exported to xml files)

mclough6 · « **Reply #24 on:** June 13, 2011, 10:46:57 PM »

Hello James,
The latest download is version 892, but your video on the home page references version 893. Is there an upcoming new release?

jitterjames · « **Reply #25 on:** June 13, 2011, 11:44:09 PM »

yes, I guess I started to make some minor changes and called it 0893, but I can't even remember what they were. Nothing significant yet. I will probably include some updated default configurations in the next version, at least for xbmc. As soon as I release a version, I tryr to updated my working version to a new number so I won't get confused about what version people are working with.

Are there any bugs, or features that you are currently concerned about, or looking for?

There will *always* be another version around the corner, though the time between and degree of change will obviously vary.

mclough6 · « **Reply #26 on:** June 14, 2011, 01:51:13 PM »

I noticed in your latest video that the computer voice seems much more natural than the one I am using. Is there a way to change voices?

jitterjames · « **Reply #27 on:** June 14, 2011, 04:18:55 PM »

You are probably using Anna which comes with Windows.

The voice in the demo is Julie from Neospeech

http://www.neospeech.com/

you can get tts voices from various places. As long as they are compatible with Windows TTS you can use them with VoxCommando

Loquendo has some good voices as well. There are probably some decent free ones too but I don't personally know of any.

mclough6 · « **Reply #28 on:** June 15, 2011, 08:07:05 AM »

Does neospeech require any special software to run or can it run through windows?

jitterjames · « **Reply #29 on:** June 15, 2011, 09:29:02 AM »

once the voice is installed you can just select it from the dropdown in VC options. You don't need run another program, but it may be running something in the background. I don't think so, it's just adding and accessing some dlls or something.

jitterjames · « **Reply #30 on:** June 15, 2011, 09:56:41 AM »

any voice that supports SAPI should work. Any of these voices should work with VC

http://www.nextup.com/acapela.html

I find Julie sounds quite real, but she is also kind of annoying after a while.

mclough6 · « **Reply #31 on:** June 15, 2011, 10:22:54 AM »

I guess I did not phrase my question very well. Do these voices need a software additional "engine" such as neospeak's VoiceText TTS Engine in order to work or can a voice be added to the engine that runs Anna in windows? If the latter, would you mind telling me how you go about doing that? Julie is not cheap ($60) and I am not anxious to spend additional money for a TTS engine.

mclough6 · « **Reply #32 on:** June 15, 2011, 03:38:33 PM »

Sorry James, but I did not see one of your responses where you say the voice can be accessed from within Vox. I will give that a try when I get home tonight. Thanks.

jitterjames · « **Reply #33 on:** June 16, 2011, 03:30:07 PM »

yep. there are also actions that you can use to switch voices on the fly

http://voxcommando.com/mediawiki/index.php?title=Actions#SetVoiceNum

Kalle · « **Reply #34 on:** June 16, 2011, 03:56:04 PM »

Quote from: mclough6 on June 15, 2011, 10:22:54 AM

I guess I did not phrase my question very well. Do these voices need a software additional "engine" such as neospeak's VoiceText TTS Engine in order to work or can a voice be added to the engine that runs Anna in windows? If the latter, would you mind telling me how you go about doing that? Julie is not cheap ($60) and I am not anxious to spend additional money for a TTS engine.

Hi mclough6,
take a look here http://www.naturalreaders.com/order_information.htm (scroll a little bit down for only the voices)
There are some voices for less money, also from neospeak

Kalle

mclough6 · « **Reply #35 on:** June 16, 2011, 04:31:50 PM »

Thanks Kalle, that is a better deal.

Kalle · « **Reply #36 on:** June 16, 2011, 04:35:45 PM »

Quote from: mclough6 on June 16, 2011, 04:31:50 PM

Thanks Kalle, that is a better deal.

important, that you use the acapella voice for windows, read the text on the top of the voices