put your music collection in a folder that's visible from the Internet! Read more
What's 'Related performers' and how do I configure it?
What it is:
"Related performers" makes it possible to display performers who are 'related' (f.ex. in music-style, era etc.) within your music collection.
Besides displaying related performers, AmpJuke also makes it possible to play music based on this setting. More about that in this FAQ-entry.
The functionality behind "Related performers" was made possible by using the API provided by last.fm (and some heavy coding in AmpJuke, naturally!).
"Related performers" doesn't require an access key or anything similar. It's there if you want it !
...and configure it, obviously.
How it's configured:
Configuration of "Related performers" consists of 7 parameters as follows:
Display related performers: If checked, you will allow displaying related performers when displaying any given performer.
If left blank (unchecked), the rest of the parameters will be ignored.
Max. number of related performers controls the maximum number of related performers to display.
If you set it to f.ex. 5, a maximum of 5 related performes will be shown when displaying any given performer.
Minimum match score: Each related performer is given a 'score' from 0-1 by last.fm.
Important note: This used to be a score between 0-100, so if you're upgrading, adjust the setting to a factor 10 lower (f.ex.: if you had a setting of 40 previously, it should now be 0.4).
The lower you set this, the more likely it will be that "Elvis" will be related to "Erasure" (if you catch my drift...).
Cache related performers locally for -- (days).
The last.fm site makes it perfectly clear that your IP-address will be banned from accessing this service if the number of requests are "too high".
Whatever number you enter in this field, will be cached on your own AmpJuke server for the amount of days you specify.
See the notes (*) below.
Display some tracks/samples from related performers: If checked (enabled), a configurable number of tracks will be displayed for each related performer.
If this setting is left blank (unchecked) the next two parameters will not have any effect.
Furthermore - and this might be important - if you leave this setting blank (unchecked), the related performers will show up in rows with 5 related performers in each row. Checking the setting will cause each related performer to be displayed one row at a time (one row=one related performer) incl. a number of tracks from that performer (see next setting).
Number of tracks/samples to display: Enter a value (integer) to control how many tracks/samples will displayed from each related performer.
Will only have effect if you enable "Display some tracks/samples from related performers" (previous setting).
Give priority to these tracks/samples: Controls the order of tracks from related performers.
There are 5 options: Nothing (completely random), Least played tracks, Most played tracks, Tracks not played recently and Tracks played recently.
This is basically the same options that's presented when setting up random play.
* A couple of notes on the caching stuff mentioned above:
As mentioned, AmpJuke retrieves (and stores) 'related performers' on your own AmpJuke server.
Both pictures of performers as well as XML-files are retrieved automatically.
If a request for the same performer is issued later by you or another user, AmpJuke will look in the cached (local) results for 'performer' and - if found - display that.
The contents for the performer in question will be updated in the cache, basically extending the amount of time the cached content will exist.
Should you set Cache related performers locally for... to 0, it means that the cached results will only exist for approximately one hour.
However, be aware that your IP most likely will be banned from last.fm.
Don't say I didn't warn you...
A sidenote: If you want a different, more "fine-grained" listening experience when it comes to picking "related" music, I can highly recommend setting up AmpJuke to use the Echonest API. Click here to read more.