VoxCommando
Help and Support (Using VoxCommando) => J River Media Center => Topic started by: nime5ter on December 31, 2013, 10:11:47 AM
-
Here is a screenshot of JRiver Plugin version 0.1.6 (which James will be releasing in the next couple of days). Note that we now have more control over how payloadXMLs are generated.
Take a look at the row for generating Songs.xml. The filter is Field=Name&Files=[Media Type]=[Audio] [rating]=>1
This will generate a payloadXML containing all of our song names. However, because we have included [rating]=>1 in our filter we will only get songs rated 2 or higher.
Also note that we have a checkbox for submatching (see http://voxcommando.com/mediawiki/index.php?title=Options#Phrase_Processing_.28PayloadXML.29). If you have a lot of songs (>20,000 titles), one way to reduce the strain on VoxCommando is to disable submatching for Songs.xml. However, in order to ask for a song you will then need to say the complete song title.
If you don't rate songs but you still want to reduce the number of songs in your payloadXML, you could try changing the filter to only include songs that you've listened to at least 3 times (just an example). For more on filters, see http://wiki.jriver.com/index.php/Smartlist_and_Search_-_Rules_and_Modifiers.
-
For those of you who don't rate your songs, but still want to limit the total number of items in your songs xml to keep it manageable, you can use play count instead. The idea being that if you've never even listened to a track, there is a pretty good chance that you won't be asking for that song by name.
Field=Name&Files=[Media Type]=[Audio] [Number Plays]=>0
I should also mention that all of these rows will generate payloadXML files with a phrase only. Playlists are the only things that require an ID and these are generated automatically as a phrase/value payloadXML file. You can see in the screenshot above that there is a subset-matching checkbox for playlists next to the generate button.
-
Note that you can filter other payloads as well, not just songs. So for artists you could modify the filter to include only artists where you have listened to at least one of their songs
Field=Artist&Files=[Media Type]=[Audio] [Number Plays]=>0
or artists that have at least one song rated higher than 3 stars
Field=Artist&Files=[Media Type]=[Audio] [Number Plays]=>0
This is a really effective way to reduce the strain on your resources (memory), reduce the load times for VC and at the same time, improve the accuracy of your recognition. If however you only have an average sized library (i.e around 10K songs or less) and it is cleanly tagged, then you may not need to use filtering at all, and can include all your songs, all your artists, all your albums etc. Even in this case you may find a use for the filtering somewhere down the road when playing with TV shows, or photos...
One thing I should point out to help avoid confusion, JRiver filters are a bit quirky when it comes to using values and the less than or greater than comparison operators.
[Number Plays] => 0 kind of looks like it means Number of Plays is greater than or equal to 0, but it doesn't, it actually mean is equal to greater than 0
if you want to filter using greater than or equal to, then you would write: =>=
-
The new version of VC (1.184) is available now on the download page and includes the updated JRiver plugin and configuration.
http://voxcommando.com/downloads.asp