23 Mar
2012
23 Mar
'12
2:56 p.m.
I am trying to prepare a lecture on scales.
If a is say 440 Hz and r is 12th root of 2
then the usual equal tempered scale would be
a
ar^2
ar^4
ar^5
ar^7
ar^9
ar^11
ar^12 = 2a
What if we change r to 7th root of 2
and replace the powers 2 4 5 7 9 11 12 by 1 2 3 4 5 6 7?
I expect it will sound nonsensical but I really dont know...
I want to show these kind of experiments on the computer. Is csound is a
good starting place to try?
Are there any toots/templates etc for such?
Also which opcodes should I concentrate on?
[For me its been a few days of rummaging around and Ive got out a sine
wave. I expect that the actual code will be about 3-4 lines for an
instrument and 2 lines for a and ar^2 after which I can fill in the blanks.
So for a csound expert it should be 5 minutes. Yes I am being a bit lazy in
asking :-) ]
23 Mar
23 Mar
3:13 p.m.
Hi Rustom1
I suppose for the purpose of a simple sound, the sine wave is quite fine. So
your instrument might look something like this:
instr 1
iamp = 15000
iftable = 1
aout oscil iamp, p4, iftable
out aout
endin
For a stereo instrument use outs like this:
outs aout, aout
You can simply calculate the pitch in Hertz and call the instrument like
this in the score:
i1 0 2 440
i1 2 2 494
;...
e
There already are some opcodes for micro-tonal scales. Take a look at the
cpstmid, cpstun, cpstuni, cpsxpch. You should find easy examples in the manual
accompanying the explanations.
One note about the instrument I've written up there: p? means the argument
to the istrument in the score, it's like shellscript variable $1, $2,... the
first three parameters are always assigned to constant things: p1 number of
the instrument "instr 1", p2 start-time of the note, p3 length of the note
(both in seconds, can be any floatingpointnumber). From p4 onwards you can
dedicate parameters to whatever you wish.
I hope this helps.
Warm regards
Julien
=-=-=-=-=-=-=-=-=-=-=-=-
Such Is Life: Very Intensely Adorable;
Free And Jubilating Amazement Revels, Dancing On - FLOWERS!
====== Find my music at ======
http://juliencoder.de/nama/music.html
.....................................
"If you live to be 100, I hope I live to be 100 minus 1 day,
so I never have to live without you." (Winnie the Pooh)
6:11 p.m.
On Fri, Mar 23, 2012 at 7:43 PM, Julien Claassen <julien(a)mail.upb.de> wrote:
Hi Rustom1
I suppose for the purpose of a simple sound, the sine wave is quite fine.
So your instrument might look something like this:
Hi Julien,
Thanks for trying. csound is still erroring at me. Its not exactly what
numbers you gave; in either case its throwing errors. Since I understand
nothing about the intricacies could you take one more look?
Meanwhile I'll look at the opcodes you suggested
Thanks
Rusi
Heres my file
----------------------------------
<CsoundSynthesizer>
<CsOptions>
</CsOptions>
<CsInstruments>
sr = 44100
kr = 4410
ksmps = 10
nchnls = 1
instr 1
iamp = 44100
iftable = 1
aout oscil iamp, p4, iftable
out aout
endin
</CsInstruments>
<CsScore>
i1 0 2 440
i1 2 2 494
e
</CsScore>
</CsoundSynthesizer>
6:22 p.m.
Hi Rustom!
First your iamp at 44100 is rather high. Try 15000 or anything up to 32768.
Or investigate the 0dbfs header statement for giving volume in DB.
How did you call csound? With that file, youshould do:
csound -W -o out.wav test.csd (or whatever you call your file).
Better still, put something like this in the CsOptions block:
-W -o out.wav
This will generate an out.wav fie.
The essential thing, which is still missing, is the ftable, to hold the
sinewav, that the oscil opcode (a simple oscillator) will output. It doesn't
hold its own waveform, but read it from a function table. So in the CsScore
section start with:
f1 0 32768 10 1 ; ftable 1, don't know the second, but it's almost always 0
:-), the third is the number of entries, the quality or resolution of your
waveform (32768 points to form a single sinewave, 10 is the number of the
function generator, which in this case is a sine generator and 1 is its first
parameter. Says: Create the first harmonic, or whatever. You can look those up
in the csound manual as well. Probably search for gen01 to see, where the list
starts. GEN 9, 10 and 19 are the ones for sinewaves, if I'm not mistaken. But
this one is the easiest.
If that doesn't help, we can see about it off-list, since then it might get
really specific to your subject. :-) You're welcome to send me CSD-files or if
you really get interested in csound as a tool, which is good for teaching all
sorts of music related stuff, you might subscribe to the csound user's list.
Warm regards
Julien
=-=-=-=-=-=-=-=-=-=-=-=-
Such Is Life: Very Intensely Adorable;
Free And Jubilating Amazement Revels, Dancing On - FLOWERS!
====== Find my music at ======
http://juliencoder.de/nama/music.html
.....................................
"If you live to be 100, I hope I live to be 100 minus 1 day,
so I never have to live without you." (Winnie the Pooh)
6:35 p.m.
Thanks Julien
It was the function table that was missing. Both 15000 and 44100 work
On Fri, Mar 23, 2012 at 10:52 PM, Julien Claassen <julien(a)mail.upb.de>wrote;wrote:
Hi Rustom!
First your iamp at 44100 is rather high. Try 15000 or anything up to
32768. Or investigate the 0dbfs header statement for giving volume in DB.
How did you call csound? With that file, youshould do:
csound -W -o out.wav test.csd (or whatever you call your file). Better
still, put something like this in the CsOptions block:
-W -o out.wav
This will generate an out.wav fie.
The essential thing, which is still missing, is the ftable, to hold the
sinewav, that the oscil opcode (a simple oscillator) will output. It
doesn't hold its own waveform, but read it from a function table. So in the
CsScore section start with:
f1 0 32768 10 1 ; ftable 1, don't know the second, but it's almost always
0 :-), the third is the number of entries, the quality or resolution of
your waveform (32768 points to form a single sinewave, 10 is the number of
the function generator, which in this case is a sine generator and 1 is its
first parameter. Says: Create the first harmonic, or whatever. You can look
those up in the csound manual as well. Probably search for gen01 to see,
where the list starts. GEN 9, 10 and 19 are the ones for sinewaves, if I'm
not mistaken. But this one is the easiest.
If that doesn't help, we can see about it off-list, since then it might
get really specific to your subject. :-) You're welcome to send me
CSD-files or if you really get interested in csound as a tool, which is
good for teaching all sorts of music related stuff, you might subscribe to
the csound user's list.
Well I first subscribed to the list at csounds. My question is here
http://www.csounds.com/forum/2
I asked here after no answer there. Is there another list?
Warm regards
Julien
Thanks again Julien
Rusi
6:36 p.m.
Hi Rustom!
On my system 44100 for amplitude distorts a lot. But if it doesn't, good on
ya. :-)
Byebye
Julien
=-=-=-=-=-=-=-=-=-=-=-=-
Such Is Life: Very Intensely Adorable;
Free And Jubilating Amazement Revels, Dancing On - FLOWERS!
====== Find my music at ======
http://juliencoder.de/nama/music.html
.....................................
"If you live to be 100, I hope I live to be 100 minus 1 day,
so I never have to live without you." (Winnie the Pooh)
4:03 p.m.
Rustom Mody wrote:
If a is say 440 Hz and r is 12th root of 2
then the usual equal tempered scale would be
a
ar^2
ar^4
ar^5
ar^7
ar^9
ar^11
ar^12 = 2a
What if we change r to 7th root of 2
and replace the powers 2 4 5 7 9 11 12 by 1 2 3 4 5 6 7?
I expect it will sound nonsensical but I really dont know...
I assume you chose the number 7 arbitrarily.
It would be possible to compute how 'good' a scale would sound; see
<http://en.wikipedia.org/wiki/Consonance_and_dissonance#Physiological_basis_of_dissonance>.
The equal-tempered scale evolved because one of its halftone steps (r^1)
is big enough to be audibly different, and, more important, the scale
allows many consonant intervals.
To be precise, equal temperament is not as consonant as other tunings,
but it is used because it is equally usable in all scales. Wikipedia
(<http://en.wikipedia.org/wiki/Musical_tuning#Systems_for_the_twelve-note_chromatic_scale>)
suggests that for different equal-tempered scales, you should use not 7
but 19, 31, or 53 steps.
Regards,
Clemens
4681
days inactive
4681
days old
linux-audio-user@lists.linuxaudio.org
6 comments
3 participants
participants (3)
-
Clemens Ladisch -
Julien Claassen -
Rustom Mody