Print

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

[NanjingGuy]

Just curious. I only have a few thousand files, but when I notice the same songs being played again and again when I have a playlist that is set to Audio and [Random] I wonder just how random that function is.

So, how does MC decide where to start and how to select 'randomly'?

[KingSparta]

at least for Visual Basic the Rnd Fuction is not random when used it will select the same file each time when used for the first time.

only when you use the  "Randomize" function prior to the code is it truely Random

[KingSparta]

That Is True, but more Random to the point that a user would never know or go that deep into it.

[kiwi]

Technically, you can't generate a true *random* list programatically since they use seed sources to generate the numbers.  To be truly random, the generator can have no built-in trends or biases.

Actually, nature seems to do a reasonable job getting random data.  There are several sources of data:

The Classic Laval Lamp Random number generator
http://203.162.7.73/webs/general/AmericanScientistMagazine/lavarand.sgi.com/index.html

Updated Laval Lamp Random number generator
http://www.lavarnd.org/index.html

With radiation
http://www.fourmilab.ch/hotbits/

You can create some sort of device that monitors a completely random natural event and sends its results to the computer. For example, you could place a piece of radioactive material in front of a Geiger counter and connect the Geiger counter to a computer. Since radioactive decay is random, the Geiger counter would create truly random numbers. This approach is pretty rare, because not many people have Geiger counters connected to their machines. http://www.howstuffworks.com/question697.htm

I think that there is one that gets data from listening to space noise. (I have a vague memory of seeing a site that did just this.)

Ah, here's what I was thinking of [url]http://www.random.org/[/quote] even has interfaces to their truly random number generator that you can use in programs.

Also, we wouldn't really *want* a truly random list since that would dictate that one track could be played 30 times in a row.

I disagree.  That's just a question of whether or not you replace the chosen track into the pool before selecting your next random track.  If you don't replace it, you could still get a truly random selection of data.  

I stand to be corrected on this, but just because you have set a range for your random number does not stop it's random nature.

kiwi

[kiwi]

I should have been more specific when I said programatically...what I meant was using the random number generators built into programming languages.  Hooking your app into something organic, like the lavalamp random number generator would indeed give you truly random patterns.  

I was familiar with the lavalamp link, but I will have to check out the space link.  Chaos is an amazing thing...

The cool thing with random.org is that they have hooks that you can put into your code that let you get random numbers from them. (assuming that you are connected to the internet.)

Shame on me.  I promise to think before posting from now on...

No worries.  I always post first...  think second ... and frequently put in "edit: bla bla bla" if I can get it in before anyone else has posted.

It's usually only after I've read a post that really makes me think that I think first and then respond.

kiwi

[KingSparta]

Hooking your app into something organic, like the lavalamp random number generator would indeed give you truly random patterns

LMAO  

[NanjingGuy]

So, nobody knows, eh?

It isn't very random at all in my opinion. In fact, if I play an artist then do any auto smartlist that use [random] I will see that artist a few times. And this is out of 30 selections that were chosen.

Can you make random a bit more random?

As for a organic random generator, my old girlfriend was absolutely random. In everything. I guess I should have kept her around for this. A little hard to interface with though...

[nameless]

I agree...  the randomization in MC isn't all that great.  The people on one-way ego trips will chime in to say that humans aren't good at detecting true randomness, but that misses the mark.  When I have 100 favorites, from 10 different artists, and a "random" playist of 12 tracks includes the same artist 3 times (more or less), over and over, while always omitting other artists completely, again, over and over, something is wrong.

[modelmaker]

I agree...  the randomization in MC isn't all that great.  The people on one-way ego trips will chime in to say that humans aren't good at detecting true randomness, but that misses the mark.  When I have 100 favorites, from 10 different artists, and a "random" playist of 12 tracks includes the same artist 3 times (more or less), over and over, while always omitting other artists completely, again, over and over, something is wrong.

I agree. When I "random" play my 12000 files, random play chooses files from only 30 or so albums. I'm still running 9.0, has this gotten any better in 9.1? Jay.

[roognation]

Also see this thread for more discussion on the problems people are seeing with [Random] http://yabb.jriver.com/interact/index.php?board=3;action=display;threadid=20600

[skidoo]

at least for Visual Basic the Rnd Fuction is not random when used it will select the same file each time when used for the first time.

only when you use the  "Randomize" function prior to the code is it truely Random

Just to pick a couple of nits.... It's nearly impossible to generate a statistically random sequence with a computer alone. And absolutely impossible to generate a *truly* random sequence with a computer alone. This is a very important factor in pursuits such as, oh, cryptography.

VB's Randomize statement just sets a new seed (based on the computer's timer) for subsequent calls to the Rnd() function. Rnd() is an implementation of a linearly-recurring pseudo random number generator; in fact, it's linearlly congruential, which means it's very simple, and very difficult to use well. Or at least very difficult to get a statistically random sample from. In fact, VB's Rnd() has a high probability of having a period (non-repeating sequence) of only 2^15, which is just 32,768 digits.

Now, let a human into the equation and, if your method is significantly effective, you can get a statistically random sequence. For example, have the user wiggle the mouse around for a while to generate a seed for a good implementation of one of the more effective pseudo random number generators or asymptotically random number generators. Noise from an open microphone can provide a good seed too, if the conditions and the implentation are right.

Umm, yeah. Nevermind. Shuffling music is what we were talking about, right?

[CASELOGIC]

ive noticed this same problem with random as well.