Front Page › Forums › AUDIO & TECHZONE › Audio Chat › the maths behind BPM…?
-
AuthorPosts
-
May 16, 2006 at 9:39 pm #533996
deklinParticipanthey guyz,
dunno if this is the right place to post this, but I am wondering, what its the equation or maths involved in working out the BPM when you know the length in ms?
is it too complex or even possible?the reason i ask is because im using max msp and i want a BPM detector but ill have to make it from scratch.
any ideas?
thanks in advance!
AdSense 336x280May 16, 2006 at 9:39 pm #589907
deklinParticipanthey guyz,
dunno if this is the right place to post this, but I am wondering, what its the equation or maths involved in working out the BPM when you know the length in ms?
is it too complex or even possible?the reason i ask is because im using max msp and i want a BPM detector but ill have to make it from scratch.
any ideas?
thanks in advance!
AdSense 336x280May 16, 2006 at 9:54 pm #589908
JuggernautParticipantBPM = Beats per minute
there are 60 seconds in a minute
the amount of bass kick pulses (or beats) there are within 60 seconds will be the resulting BPM.Depending on the music you’re using, creating a BPM detector (especially in Max/MSP) is a pain in the butt. i mean if cd’s with the fancy bpm counters go nuts with something that’s not "clearly" 4/4 it’s hardly going to be a picnic in the park…
i haven’t used max/msp in a good long time (it scared me off for a while) but i’d imagine you’re going to have to define probably a frequency range for it to listen to for the said ‘pulse’ and calculate the distance between each pulse within 60 seconds. the distance=Speed/time equation comes to mind although you’d have to change that to be relevant to BPM as opposed to physical distance etc.
say we reduced the sample time to 10 seconds and for arguments sake there were 15 pulses in those 10 seconds. multiply that by 6 and you’ll get 90bpm.
that’s probably not helped at all…guess i shouldn’t think/type simultaneously.
i’m doing (or should be doing but haven’t got time at the moment) which will require getting back into grips with max/msp and getting a beat/bpm detector working to trigger something else. i’m sure there’s an equation or a way to do it…maybe i’m looking in totally the wrong direction
AdSense 336x280May 16, 2006 at 10:01 pm #589910
J.MackParticipantsounds likd tuff stuff.
AdSense 336x280May 17, 2006 at 12:27 am #589912
deklinParticipantthanks for the info. there are quite a few external objects that convert bpm to ms but i cant see them being of much use.
perhaps a counter that is triggered by a volume threshold is the way to go, although you cant be sure that every samples transient will reach about this threshold.
its a tough subject!AdSense 336x280May 17, 2006 at 4:54 am #589921
lewelschParticipantTry:
http://mp3.deepsound.net/eng/samples_calculs.phpHere is 11 very cool javascript sample calculators to help us to deal with samples, time-stretching, pitch-shifting, tempo, delays, sampling rates, sample lengths, bpm, notes…
Enjoy
LeWelschAdSense 336x280May 17, 2006 at 9:34 am #589936
JuggernautParticipantI thought that first too, but like a said before it depends on what you’re sampling. it would be great to use volume threshold on a Benny Benassi or even an eric prydz track as they’ve compressed so much out of it, the 4/4 is exceptionally prominent.
maybe most BPM counters do rely on that…i’ve always wondered (never bothered looking) about that.
i’ve just thought it would be easier and yeald more convincing results if you set an object to listen to a set frequency range where the bass kick usually is (as opposed to snare/white noise)
Tell us how you get on with it…you’ll probably help me with my project too
…when i get time for it againAdSense 336x280May 17, 2006 at 7:12 pm #589973
ProcleusParticipantLewelsch definitely knows what he’s talking about. He can clear things up about the software and its usage.
However, something I need to point out, even to the veterans, is that 120 BPM is not the same as 120.01 BPM, or even 120.001 BPM. Especially over the course of a minute with such specific arrangements. My point is, a program that tells you a whole integer shouldn’t be the end of figuring out the pure BPM.
What I do is use a soundwave editor, zoom in to one beat, and trim to the exact beginnings and ends. If a machine did the rest of the sequencing, that will be the length of every beat. Compare that to the beat of another sequence, and they tend to be different every time, even with the "same" BPM.
Definitely more work than you want, I know, but it makes a difference when blending 6-7 minutes worth of sequences…
AdSense 336x280May 18, 2006 at 2:03 am #589998
JuggernautParticipanti think you’re both missing the point…
the program the question is based in is MAX/MSP…all the stuff that you take for granted in sound editor software, you can make from scratch (or download patches from the forum). which is why the question was asked (i think?, correct me if i’m wrong).
the javascript page is interesting and one of them is going in the right direction of what i was thinking…but doesn’t quite get there (plus i don’t have max/msp installed at the mo)
probably still not making sense: http://www.cycling74.com/products/maxmsp
AdSense 336x280May 19, 2006 at 3:40 pm #590084
deklinParticipantthe javascripts are no help as I need to have the calculator self-contained within the Max program. any external programes are no use. which is why I was asking for the maths so i can build it myself.
say the time between a transient and another is 95ms and you need to work out how many times those transients occured in one minute then couldnt you divide 6000 by 95? its 63.1,would that be the BPM?
that seems too simple, though
AdSense 336x280May 19, 2006 at 4:54 pm #590087
JuggernautParticipant…seems logical that would be time= speed/distance (even though that’s not directly related to this, it kinda works in a similar manner)
or beats per minute=time(1min=6000ms)/distance between transients(95ms)
that looks like it would make sense/could work. even though it seems too simple, it’s usually the most simplest things that work the best (not much to go wrong etc.) why not try that and see if it works?
**edit** i take that back. just found a thesis by doing some googling. haven’t read it all but looks like there’s possibly some relevant stuff from page 43 in this pdf: http://eamusic.dartmouth.edu/~iroro/the … ddim.pdf**
if that link is a no no, i’m sorry mods
AdSense 336x280 -
AuthorPosts
The forum ‘Audio Chat’ is closed to new topics and replies.
