Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[theoctavist]

in spite of having my entire library selected( 45,000 tracks or so) and shuffled multiple times, MC seems to be sampling tracks from a very small group of artists, and playing those over and over.

if there a way to make this more random, truly random? forgive me if this has been talked about!

[AndrewFG]

The definition of random is that unexpected things may happen...

[Fred1]

The definition of random is that unexpected things may happen...

Over and over again?
Always the same?

[JimH]

Have you ever driven to work and hit nothing but green lights?  Or nothing but red lights?

[fitbrit]

Are your tracks rated? I believe there is a setting to bias for more highly rated songs when playing random. Maybe you could check in options and see whether you have that activated.

[theoctavist]

tracks are unrated, which is what makes it hard to figure out!!

squeezebox touch had the same issue(although jriver is better)

http://forums.slimdevices.com/archive/index.php/t-55513.html

[6233638]

You could always use smartlists with rules that exclude tracks where the last played time is less than a certain value. (e.g. 1 hour)

[MrC]

I also find that MC is not very random, and repeats sequences or selections.

Anyway, pscriptor 1.10 just posted now includes the scriptlet MakeRandomPlaylist which will create a playlist of size N, and autostart playback to some specified zone.

[theoctavist]

I also find that MC is not very random, and repeats sequences or selections.

Anyway, pscriptor 1.10 just posted now includes the scriptlet MakeRandomPlaylist which will create a playlist of size N, and autostart playback to some specified zone.

thank you! this was a problem that plagued squeezebox touch users from the get go. never found a way out of it! *hence my ending up at jriver(huge upgrade btw)

[JimCo06]

It may seem strange but, computers really can't generate a truly random series of numbers.  Remember, computers can only due what the software tells them to do.  The Random function built into most OS's is just a complex mathematical equation that acts on a "seed" number.  If you run the function using the same seed number every time it will produce exactly the same sequence of "random" numbers every time.  The more variety there is in the seed numbers used, the more the generated sequences of numbers will appear to be random.  To get the largest possible variety in the seed numbers programmers use to preform some sort of math on elements of the system date and time.  Don't know if they still do that or not, but it appears they still aren't able to generate really good random sequences.

[MrC]

While the gist of what you say is essentially true, there are many techniques and libraries that make cryptographically secure or for all practical purposes "true" random sequences.  These do not base their algorithms on fixed or predictable seeds (such as date and time, and have much more entropy).

But the focus of this tread is not on perfect randomness - just something that is better that the results obtained thus far.