I find this music extremely expressive and impressive.

Art Hunkins

----- Original Message ----- 
From: "'2+" 
To: 
Sent: Sunday, August 09, 2009 5:25 PM
Subject: [Csnd] lots of notes with the instrument using ftgentmp ate up the 
ram


> hi!
>
> was quite happy with ftgentmp
> i could almost ramdomly prepare ftables for each note
> by taking advantage of using p2-p5
> which originally is used for other perpose
> (wanted to avoid using randoms)
> most of my inst was written this way
>
> instr 12
> isa = (0.22 + p2) % 0.997 + 0.001
> iri = (0.33 + p3) % 0.996 + 0.002
> iga = (0.44 + p4) % 0.995 + 0.003
> ima = (0.55 + p5) % 0.994 + 0.004
> ipa = ((0.11 + p2 + p3) * (0.66 + p5)) % 0.993 + 0.005
> ida = ((0.77 + p4) * (0.88 + p5)) % 0.992 + 0.006
> ini = ((0.99 + p4 + p5) * (0.23456789 + p2 + p3)) % 0.991 + 0.007
> iot_sa = 8 + int((22345 * isa) % 3)
> iot_ri = 5 + 4 * int((33456 * iri) % 3)
> iot_ga = 10 + 3 * int((44567 * iga) % 3)
> iot_ma = 8 + 8 * int((55678 * ima) % 3)
> iot_pa = 6 + 2 * int((66789 * ipa) % 4)
> iot_da = 8 + 5 * int((77890 * ida) % 3)
> iot_ni = 8 + 4 * int((88901 * ini) % 5)
> ift_sa  ftgentmp 0, 0, 65536, 10, 16, 8, 1/3, 4, 1/5, 1/6, 0, 2, 9 *
> isa, 1/10, 0, 1/12, 1/13, 0, 0, 1
> ift_ri  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iri), 1.0 - iga, 1.0 -
> ima, 1.0 - ipa, 1.0 - ida, 0, ida, 9 * iri, ini, 0, isa, iga, 0, 0,
> ima
> ift_ga  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iga), 1.0 - ima, 1.0 -
> ipa, 1.0 - ida, 1.0 - ini, 0, ini, 9 * iga, isa, 0, iri, ima, 0, 0,
> ipa
> ift_ma  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ima), 1.0 - ipa, 1.0 -
> ida, 1.0 - ini, 1.0 - isa, 0, isa, 9 * ima, iri, 0, iga, ipa, 0, 0,
> ida
> ift_pa  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ipa), 1.0 - ida, 1.0 -
> ini, 1.0 - isa, 1.0 - iri, 0, iri, 9 * ipa, iga, 0, ima, ida, 0, 0,
> ini
> ift_da  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ida), 1.0 - ini, 1.0 -
> isa, 1.0 - iri, 1.0 - iga, 0, iga, 9 * ida, ima, 0, ipa, ini, 0, 0,
> isa
> ift_ni  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ini), 1.0 - isa, 1.0 -
> iri, 1.0 - iga, 1.0 - ima, 0, ima, 9 * ini, ipa, 0, ida, isa, 0, 0,
> iri
> kenv_sa   linseg p4, p3, 0
> kenv_ri   linseg 0, iri * p3, p4, (1.0 - iri) * p3, 0
> kenv_ga   linseg 0, iga * p3, p4, (1.0 - iga) * p3, 0
> kenv_ma   linseg 0, ima * p3, p4, (1.0 - ima) * p3, 0
> kenv_pa   linseg 0, ipa * p3, p4, (1.0 - ipa) * p3, 0
> kenv_da   linseg 0, ida * p3, p4, (1.0 - ida) * p3, 0
> kenv_ni   linseg 0, ini * p3, p4, (1.0 - ini) * p3, 0
> asig_sa   oscil kenv_sa, p5, ift_sa, isa
> asig_ri   oscil kenv_ri, p5 * 8/9, ift_ri, iri
> asig_ga   oscil kenv_ga, p5 * 8/9, ift_ga, iga
> asig_ma   oscil kenv_ma, p5 * 9/8, ift_ma, ima
> asig_pa   oscil kenv_pa, p5 * 9/8, ift_pa, ipa
> asig_da   oscil kenv_da, p5 * iot_da/8, ift_da, ida
> asig_ni   oscil kenv_ni, p5 * iot_ni/8, ift_ni, ini
> outs    (asig_sa + asig_ri + asig_ma + asig_da) * giamp/4, (asig_sa +
> asig_ga + asig_pa + asig_ni) * giamp/4
> endin
>
> it was okay if the tune only lasts for about 1 min
> but for 3 min ... it eats up the ram
> can i still find a way to get an one and only ftable for each note?
>
> at this moment i simply replaced ftgentmp with ftgen
> so the using the same inst means using the same ftable
> the sound goes like this:
> http://sarigama.namaste.jp/loops/atrip2heaven.mp3
>
> tia!
>
> -- 
> SaRiGaMa's Oil Vending Orchestra
> is podcasting:
> http://sarigama.namaste.jp/podcast/rss.xml
> and supplying oil.py for free:
> http://oilpy.blogspot.com/
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe 
> csound" 

encouraged by that! thanx!
but do i really have to give up using ftgentmp?

On Mon, Aug 10, 2009 at 7:33 AM, Art Hunkins wrote:
> I find this music extremely expressive and impressive.
>
> Art Hunkins
>
> ----- Original Message ----- From: "'2+" 
> To: 
> Sent: Sunday, August 09, 2009 5:25 PM
> Subject: [Csnd] lots of notes with the instrument using ftgentmp ate up the
> ram
>
>
>> hi!
>>
>> was quite happy with ftgentmp
>> i could almost ramdomly prepare ftables for each note
>> by taking advantage of using p2-p5
>> which originally is used for other perpose
>> (wanted to avoid using randoms)
>> most of my inst was written this way
>>
>> instr 12
>> isa = (0.22 + p2) % 0.997 + 0.001
>> iri = (0.33 + p3) % 0.996 + 0.002
>> iga = (0.44 + p4) % 0.995 + 0.003
>> ima = (0.55 + p5) % 0.994 + 0.004
>> ipa = ((0.11 + p2 + p3) * (0.66 + p5)) % 0.993 + 0.005
>> ida = ((0.77 + p4) * (0.88 + p5)) % 0.992 + 0.006
>> ini = ((0.99 + p4 + p5) * (0.23456789 + p2 + p3)) % 0.991 + 0.007
>> iot_sa = 8 + int((22345 * isa) % 3)
>> iot_ri = 5 + 4 * int((33456 * iri) % 3)
>> iot_ga = 10 + 3 * int((44567 * iga) % 3)
>> iot_ma = 8 + 8 * int((55678 * ima) % 3)
>> iot_pa = 6 + 2 * int((66789 * ipa) % 4)
>> iot_da = 8 + 5 * int((77890 * ida) % 3)
>> iot_ni = 8 + 4 * int((88901 * ini) % 5)
>> ift_sa  ftgentmp 0, 0, 65536, 10, 16, 8, 1/3, 4, 1/5, 1/6, 0, 2, 9 *
>> isa, 1/10, 0, 1/12, 1/13, 0, 0, 1
>> ift_ri  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iri), 1.0 - iga, 1.0 -
>> ima, 1.0 - ipa, 1.0 - ida, 0, ida, 9 * iri, ini, 0, isa, iga, 0, 0,
>> ima
>> ift_ga  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iga), 1.0 - ima, 1.0 -
>> ipa, 1.0 - ida, 1.0 - ini, 0, ini, 9 * iga, isa, 0, iri, ima, 0, 0,
>> ipa
>> ift_ma  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ima), 1.0 - ipa, 1.0 -
>> ida, 1.0 - ini, 1.0 - isa, 0, isa, 9 * ima, iri, 0, iga, ipa, 0, 0,
>> ida
>> ift_pa  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ipa), 1.0 - ida, 1.0 -
>> ini, 1.0 - isa, 1.0 - iri, 0, iri, 9 * ipa, iga, 0, ima, ida, 0, 0,
>> ini
>> ift_da  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ida), 1.0 - ini, 1.0 -
>> isa, 1.0 - iri, 1.0 - iga, 0, iga, 9 * ida, ima, 0, ipa, ini, 0, 0,
>> isa
>> ift_ni  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ini), 1.0 - isa, 1.0 -
>> iri, 1.0 - iga, 1.0 - ima, 0, ima, 9 * ini, ipa, 0, ida, isa, 0, 0,
>> iri
>> kenv_sa   linseg p4, p3, 0
>> kenv_ri   linseg 0, iri * p3, p4, (1.0 - iri) * p3, 0
>> kenv_ga   linseg 0, iga * p3, p4, (1.0 - iga) * p3, 0
>> kenv_ma   linseg 0, ima * p3, p4, (1.0 - ima) * p3, 0
>> kenv_pa   linseg 0, ipa * p3, p4, (1.0 - ipa) * p3, 0
>> kenv_da   linseg 0, ida * p3, p4, (1.0 - ida) * p3, 0
>> kenv_ni   linseg 0, ini * p3, p4, (1.0 - ini) * p3, 0
>> asig_sa   oscil kenv_sa, p5, ift_sa, isa
>> asig_ri   oscil kenv_ri, p5 * 8/9, ift_ri, iri
>> asig_ga   oscil kenv_ga, p5 * 8/9, ift_ga, iga
>> asig_ma   oscil kenv_ma, p5 * 9/8, ift_ma, ima
>> asig_pa   oscil kenv_pa, p5 * 9/8, ift_pa, ipa
>> asig_da   oscil kenv_da, p5 * iot_da/8, ift_da, ida
>> asig_ni   oscil kenv_ni, p5 * iot_ni/8, ift_ni, ini
>> outs    (asig_sa + asig_ri + asig_ma + asig_da) * giamp/4, (asig_sa +
>> asig_ga + asig_pa + asig_ni) * giamp/4
>> endin
>>
>> it was okay if the tune only lasts for about 1 min
>> but for 3 min ... it eats up the ram
>> can i still find a way to get an one and only ftable for each note?
>>
>> at this moment i simply replaced ftgentmp with ftgen
>> so the using the same inst means using the same ftable
>> the sound goes like this:
>> http://sarigama.namaste.jp/loops/atrip2heaven.mp3
>>
>> tia!
>>


-- 
SaRiGaMa's Oil Vending Orchestra
is podcasting:
http://sarigama.namaste.jp/podcast/rss.xml
and supplying oil.py for free:
http://oilpy.blogspot.com/

Your problem seems strange, as ftgentmp should free memory when the note ends.
Could you send a full csd example that eats up memory?, so we can test.
all best
Oeyvind

2009/8/10 '2+ :
> encouraged by that! thanx!
> but do i really have to give up using ftgentmp?
>
> On Mon, Aug 10, 2009 at 7:33 AM, Art Hunkins wrote:
>> I find this music extremely expressive and impressive.
>>
>> Art Hunkins
>>
>> ----- Original Message ----- From: "'2+" 
>> To: 
>> Sent: Sunday, August 09, 2009 5:25 PM
>> Subject: [Csnd] lots of notes with the instrument using ftgentmp ate up the
>> ram
>>
>>
>>> hi!
>>>
>>> was quite happy with ftgentmp
>>> i could almost ramdomly prepare ftables for each note
>>> by taking advantage of using p2-p5
>>> which originally is used for other perpose
>>> (wanted to avoid using randoms)
>>> most of my inst was written this way
>>>
>>> instr 12
>>> isa = (0.22 + p2) % 0.997 + 0.001
>>> iri = (0.33 + p3) % 0.996 + 0.002
>>> iga = (0.44 + p4) % 0.995 + 0.003
>>> ima = (0.55 + p5) % 0.994 + 0.004
>>> ipa = ((0.11 + p2 + p3) * (0.66 + p5)) % 0.993 + 0.005
>>> ida = ((0.77 + p4) * (0.88 + p5)) % 0.992 + 0.006
>>> ini = ((0.99 + p4 + p5) * (0.23456789 + p2 + p3)) % 0.991 + 0.007
>>> iot_sa = 8 + int((22345 * isa) % 3)
>>> iot_ri = 5 + 4 * int((33456 * iri) % 3)
>>> iot_ga = 10 + 3 * int((44567 * iga) % 3)
>>> iot_ma = 8 + 8 * int((55678 * ima) % 3)
>>> iot_pa = 6 + 2 * int((66789 * ipa) % 4)
>>> iot_da = 8 + 5 * int((77890 * ida) % 3)
>>> iot_ni = 8 + 4 * int((88901 * ini) % 5)
>>> ift_sa  ftgentmp 0, 0, 65536, 10, 16, 8, 1/3, 4, 1/5, 1/6, 0, 2, 9 *
>>> isa, 1/10, 0, 1/12, 1/13, 0, 0, 1
>>> ift_ri  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iri), 1.0 - iga, 1.0 -
>>> ima, 1.0 - ipa, 1.0 - ida, 0, ida, 9 * iri, ini, 0, isa, iga, 0, 0,
>>> ima
>>> ift_ga  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iga), 1.0 - ima, 1.0 -
>>> ipa, 1.0 - ida, 1.0 - ini, 0, ini, 9 * iga, isa, 0, iri, ima, 0, 0,
>>> ipa
>>> ift_ma  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ima), 1.0 - ipa, 1.0 -
>>> ida, 1.0 - ini, 1.0 - isa, 0, isa, 9 * ima, iri, 0, iga, ipa, 0, 0,
>>> ida
>>> ift_pa  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ipa), 1.0 - ida, 1.0 -
>>> ini, 1.0 - isa, 1.0 - iri, 0, iri, 9 * ipa, iga, 0, ima, ida, 0, 0,
>>> ini
>>> ift_da  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ida), 1.0 - ini, 1.0 -
>>> isa, 1.0 - iri, 1.0 - iga, 0, iga, 9 * ida, ima, 0, ipa, ini, 0, 0,
>>> isa
>>> ift_ni  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ini), 1.0 - isa, 1.0 -
>>> iri, 1.0 - iga, 1.0 - ima, 0, ima, 9 * ini, ipa, 0, ida, isa, 0, 0,
>>> iri
>>> kenv_sa   linseg p4, p3, 0
>>> kenv_ri   linseg 0, iri * p3, p4, (1.0 - iri) * p3, 0
>>> kenv_ga   linseg 0, iga * p3, p4, (1.0 - iga) * p3, 0
>>> kenv_ma   linseg 0, ima * p3, p4, (1.0 - ima) * p3, 0
>>> kenv_pa   linseg 0, ipa * p3, p4, (1.0 - ipa) * p3, 0
>>> kenv_da   linseg 0, ida * p3, p4, (1.0 - ida) * p3, 0
>>> kenv_ni   linseg 0, ini * p3, p4, (1.0 - ini) * p3, 0
>>> asig_sa   oscil kenv_sa, p5, ift_sa, isa
>>> asig_ri   oscil kenv_ri, p5 * 8/9, ift_ri, iri
>>> asig_ga   oscil kenv_ga, p5 * 8/9, ift_ga, iga
>>> asig_ma   oscil kenv_ma, p5 * 9/8, ift_ma, ima
>>> asig_pa   oscil kenv_pa, p5 * 9/8, ift_pa, ipa
>>> asig_da   oscil kenv_da, p5 * iot_da/8, ift_da, ida
>>> asig_ni   oscil kenv_ni, p5 * iot_ni/8, ift_ni, ini
>>> outs    (asig_sa + asig_ri + asig_ma + asig_da) * giamp/4, (asig_sa +
>>> asig_ga + asig_pa + asig_ni) * giamp/4
>>> endin
>>>
>>> it was okay if the tune only lasts for about 1 min
>>> but for 3 min ... it eats up the ram
>>> can i still find a way to get an one and only ftable for each note?
>>>
>>> at this moment i simply replaced ftgentmp with ftgen
>>> so the using the same inst means using the same ftable
>>> the sound goes like this:
>>> http://sarigama.namaste.jp/loops/atrip2heaven.mp3
>>>
>>> tia!
>>>
>
>
> --
> SaRiGaMa's Oil Vending Orchestra
> is podcasting:
> http://sarigama.namaste.jp/podcast/rss.xml
> and supplying oil.py for free:
> http://oilpy.blogspot.com/
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

i put:
http://sarigama.namaste.jp/oil/hungry_oil.sco
http://sarigama.namaste.jp/oil/hungry_oil.orc
on the web
but wow ... didn't expect that this sco was more than 1MB
it was produced via python

maybe again it might be me doing the typo or something like that .. but anyway
will swallow my pride and show this entire ugly code to beg for help
so if i look into it carefully there might be a case that notes are not ending?

ah! .. some notes' amp can be 0.0 as a result of algo-comping and
printing it in a limited length
maybe in that case cs is skippin the process after generating the ftable?
i should fix the python code to skip spittin out these notes and test with that

On Tue, Aug 11, 2009 at 4:00 PM, Oeyvind Brandtsegg wrote:
> Your problem seems strange, as ftgentmp should free memory when the note ends.
> Could you send a full csd example that eats up memory?, so we can test.
> all best
> Oeyvind
>
> 2009/8/10 '2+ :
>> encouraged by that! thanx!
>> but do i really have to give up using ftgentmp?
>>
>> On Mon, Aug 10, 2009 at 7:33 AM, Art Hunkins wrote:
>>> I find this music extremely expressive and impressive.
>>>
>>> Art Hunkins
>>>
>>> ----- Original Message ----- From: "'2+" 
>>> To: 
>>> Sent: Sunday, August 09, 2009 5:25 PM
>>> Subject: [Csnd] lots of notes with the instrument using ftgentmp ate up the
>>> ram
>>>
>>>
>>>> hi!
>>>>
>>>> was quite happy with ftgentmp
>>>> i could almost ramdomly prepare ftables for each note
>>>> by taking advantage of using p2-p5
>>>> which originally is used for other perpose
>>>> (wanted to avoid using randoms)
>>>> most of my inst was written this way
>>>>
>>>> instr 12
>>>> isa = (0.22 + p2) % 0.997 + 0.001
>>>> iri = (0.33 + p3) % 0.996 + 0.002
>>>> iga = (0.44 + p4) % 0.995 + 0.003
>>>> ima = (0.55 + p5) % 0.994 + 0.004
>>>> ipa = ((0.11 + p2 + p3) * (0.66 + p5)) % 0.993 + 0.005
>>>> ida = ((0.77 + p4) * (0.88 + p5)) % 0.992 + 0.006
>>>> ini = ((0.99 + p4 + p5) * (0.23456789 + p2 + p3)) % 0.991 + 0.007
>>>> iot_sa = 8 + int((22345 * isa) % 3)
>>>> iot_ri = 5 + 4 * int((33456 * iri) % 3)
>>>> iot_ga = 10 + 3 * int((44567 * iga) % 3)
>>>> iot_ma = 8 + 8 * int((55678 * ima) % 3)
>>>> iot_pa = 6 + 2 * int((66789 * ipa) % 4)
>>>> iot_da = 8 + 5 * int((77890 * ida) % 3)
>>>> iot_ni = 8 + 4 * int((88901 * ini) % 5)
>>>> ift_sa  ftgentmp 0, 0, 65536, 10, 16, 8, 1/3, 4, 1/5, 1/6, 0, 2, 9 *
>>>> isa, 1/10, 0, 1/12, 1/13, 0, 0, 1
>>>> ift_ri  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iri), 1.0 - iga, 1.0 -
>>>> ima, 1.0 - ipa, 1.0 - ida, 0, ida, 9 * iri, ini, 0, isa, iga, 0, 0,
>>>> ima
>>>> ift_ga  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iga), 1.0 - ima, 1.0 -
>>>> ipa, 1.0 - ida, 1.0 - ini, 0, ini, 9 * iga, isa, 0, iri, ima, 0, 0,
>>>> ipa
>>>> ift_ma  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ima), 1.0 - ipa, 1.0 -
>>>> ida, 1.0 - ini, 1.0 - isa, 0, isa, 9 * ima, iri, 0, iga, ipa, 0, 0,
>>>> ida
>>>> ift_pa  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ipa), 1.0 - ida, 1.0 -
>>>> ini, 1.0 - isa, 1.0 - iri, 0, iri, 9 * ipa, iga, 0, ima, ida, 0, 0,
>>>> ini
>>>> ift_da  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ida), 1.0 - ini, 1.0 -
>>>> isa, 1.0 - iri, 1.0 - iga, 0, iga, 9 * ida, ima, 0, ipa, ini, 0, 0,
>>>> isa
>>>> ift_ni  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ini), 1.0 - isa, 1.0 -
>>>> iri, 1.0 - iga, 1.0 - ima, 0, ima, 9 * ini, ipa, 0, ida, isa, 0, 0,
>>>> iri
>>>> kenv_sa   linseg p4, p3, 0
>>>> kenv_ri   linseg 0, iri * p3, p4, (1.0 - iri) * p3, 0
>>>> kenv_ga   linseg 0, iga * p3, p4, (1.0 - iga) * p3, 0
>>>> kenv_ma   linseg 0, ima * p3, p4, (1.0 - ima) * p3, 0
>>>> kenv_pa   linseg 0, ipa * p3, p4, (1.0 - ipa) * p3, 0
>>>> kenv_da   linseg 0, ida * p3, p4, (1.0 - ida) * p3, 0
>>>> kenv_ni   linseg 0, ini * p3, p4, (1.0 - ini) * p3, 0
>>>> asig_sa   oscil kenv_sa, p5, ift_sa, isa
>>>> asig_ri   oscil kenv_ri, p5 * 8/9, ift_ri, iri
>>>> asig_ga   oscil kenv_ga, p5 * 8/9, ift_ga, iga
>>>> asig_ma   oscil kenv_ma, p5 * 9/8, ift_ma, ima
>>>> asig_pa   oscil kenv_pa, p5 * 9/8, ift_pa, ipa
>>>> asig_da   oscil kenv_da, p5 * iot_da/8, ift_da, ida
>>>> asig_ni   oscil kenv_ni, p5 * iot_ni/8, ift_ni, ini
>>>> outs    (asig_sa + asig_ri + asig_ma + asig_da) * giamp/4, (asig_sa +
>>>> asig_ga + asig_pa + asig_ni) * giamp/4
>>>> endin
>>>>
>>>> it was okay if the tune only lasts for about 1 min
>>>> but for 3 min ... it eats up the ram
>>>> can i still find a way to get an one and only ftable for each note?
>>>>
>>>> at this moment i simply replaced ftgentmp with ftgen
>>>> so the using the same inst means using the same ftable
>>>> the sound goes like this:
>>>> http://sarigama.namaste.jp/loops/atrip2heaven.mp3
>>>>
>>>> tia!

-- 
SaRiGaMa's Oil Vending Orchestra
is podcasting:
http://sarigama.namaste.jp/podcast/rss.xml
and supplying oil.py for free:
http://oilpy.blogspot.com/

This does look interesting

I looked at the website, & the python code there, but it wasnt particularly
informative (although I applaud the inpenetrable esotericism of the class
definitions! ; ) ...)

is it possible to have some sort of overview of the process being undertaken
/ explored here?

ways & process's to create unique function tables for each sound are also of
interest to me, esp if they contribute to "acoustic viability" / more
"natural simulations" somehow..... or evolving textures / processes in
general.....



-----
http://www.timmortimer.net


-- 
View this message in context: http://www.nabble.com/lots-of-notes-with-the-instrument-using-ftgentmp-ate-up-the-ram-tp24891384p24929733.html
Sent from the Csound - General mailing list archive at Nabble.com.

On 12/08/2009, at 2:14 PM, Tim Mortimer wrote:

>
> This does look interesting
>
> I looked at the website, & the python code there, but it wasnt  
> particularly
> informative (although I applaud the inpenetrable esotericism of the  
> class
> definitions! ; ) ...)
>
TIm!
Being informative is not compatible with inpenetrable esotericism. I  
applaud your inconsistency!
To be clear (sorry, this is an Age of Reason hangover) this code is  
not esoteric, it is simply obscure, probably due to hard-wiring of  
array indices and the lack of comments.
Such code might work, but using it for de-obscurantistic practices is  
highly likely to lead to unsatisfactory outcomes.

D.


> is it possible to have some sort of overview of the process being  
> undertaken
> / explored here?
>
> ways & process's to create unique function tables for each sound are  
> also of
> interest to me, esp if they contribute to "acoustic viability" / more
> "natural simulations" somehow..... or evolving textures / processes in
> general.....
>
>
>
> -----
> http://www.timmortimer.net
>
>
> -- 
> View this message in context: http://www.nabble.com/lots-of-notes-with-the-instrument-using-ftgentmp-ate-up-the-ram-tp24891384p24929733.html
> Sent from the Csound - General mailing list archive at Nabble.com.
>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"

________________________________________________
Dr David Worrall.
- Experimental Polymedia:	  worrall.avatar.com.au
- Sonification: www.sonifiction.com.au
- Education for Financial Independence: www.mindthemarkets.com.au

So what your trying to say is "dont worry about it, at the core its actually
pretty mundane stuff?"

if so, then im even more inclined to look for a brief, informal explanation
from the author.

i was obviously hoping the "obscurity" (conceded) would be dispelled by this
information. so the extent to which this is an "inconsistency" is exactly
the point. a bit like night & day, for example. shades of grey & dusk & dawn
not withstanding of course.....

i was, i admit, in the middle of avoiding trying to do something else a lot
more boring at the time.

i wasnt in a position to listen to the mp3 either. i might do that now
however.


David Worrall wrote:
> 
> 
> On 12/08/2009, at 2:14 PM, Tim Mortimer wrote:
> 
>>
> 
>>
> TIm!
> Being informative is not compatible with inpenetrable esotericism. I  
> applaud your inconsistency!
> To be clear (sorry, this is an Age of Reason hangover) this code is  
> not esoteric, it is simply obscure, probably due to hard-wiring of  
> array indices and the lack of comments.
> Such code might work, but using it for de-obscurantistic practices is  
> highly likely to lead to unsatisfactory outcomes.
> 
> D.
> 
> 
> 
> 
> 
> 
> 
> 
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
> csound"
> 
> 


-----
http://www.timmortimer.net


-- 
View this message in context: http://www.nabble.com/lots-of-notes-with-the-instrument-using-ftgentmp-ate-up-the-ram-tp24891384p24930980.html
Sent from the Csound - General mailing list archive at Nabble.com.

the mp3 was pretty interesting. havent listened to anything like that in a
while. very "squiggly". liked the chord at the end! ; )

reminded me a bit of the Xenakis i saw the Australian Chamber Orchestra play
last night... ill spare you my review....



-----
http://www.timmortimer.net


-- 
View this message in context: http://www.nabble.com/lots-of-notes-with-the-instrument-using-ftgentmp-ate-up-the-ram-tp24891384p24931139.html
Sent from the Csound - General mailing list archive at Nabble.com.

Hello,

I suspect the memory leak (ram eating) may have something to do with
the large amount of instrument events. It would be very useful to nail
this down, if it is a bug (or in any other case too of course), but it
would be a lot easier if you could provide a very simple example.
Maybe try to manually edit the score, or adjust the algorithm
parameters to generate something with just a few events. Then see if
the memory leak still occurs.
Just a shot in the dark: It might have to do with several "almost
simultaneous" events, I'm not sure, it should not be a problem, but
stranger things have happened.
If you have a score with approximantely 2 events per second, and let
it run for a long time, will the memory still leak ?


Also, sometimes in the process of making a simple example, the
solution (or your error) stands out.
Happens to me all the time ;-)

Oeyvind

2009/8/12 '2+ :
> i put:
> http://sarigama.namaste.jp/oil/hungry_oil.sco
> http://sarigama.namaste.jp/oil/hungry_oil.orc
> on the web
> but wow ... didn't expect that this sco was more than 1MB
> it was produced via python
>
> maybe again it might be me doing the typo or something like that .. but anyway
> will swallow my pride and show this entire ugly code to beg for help
> so if i look into it carefully there might be a case that notes are not ending?
>
> ah! .. some notes' amp can be 0.0 as a result of algo-comping and
> printing it in a limited length
> maybe in that case cs is skippin the process after generating the ftable?
> i should fix the python code to skip spittin out these notes and test with that
>
> On Tue, Aug 11, 2009 at 4:00 PM, Oeyvind Brandtsegg wrote:
>> Your problem seems strange, as ftgentmp should free memory when the note ends.
>> Could you send a full csd example that eats up memory?, so we can test.
>> all best
>> Oeyvind
>>
>> 2009/8/10 '2+ :
>>> encouraged by that! thanx!
>>> but do i really have to give up using ftgentmp?
>>>
>>> On Mon, Aug 10, 2009 at 7:33 AM, Art Hunkins wrote:
>>>> I find this music extremely expressive and impressive.
>>>>
>>>> Art Hunkins
>>>>
>>>> ----- Original Message ----- From: "'2+" 
>>>> To: 
>>>> Sent: Sunday, August 09, 2009 5:25 PM
>>>> Subject: [Csnd] lots of notes with the instrument using ftgentmp ate up the
>>>> ram
>>>>
>>>>
>>>>> hi!
>>>>>
>>>>> was quite happy with ftgentmp
>>>>> i could almost ramdomly prepare ftables for each note
>>>>> by taking advantage of using p2-p5
>>>>> which originally is used for other perpose
>>>>> (wanted to avoid using randoms)
>>>>> most of my inst was written this way
>>>>>
>>>>> instr 12
>>>>> isa = (0.22 + p2) % 0.997 + 0.001
>>>>> iri = (0.33 + p3) % 0.996 + 0.002
>>>>> iga = (0.44 + p4) % 0.995 + 0.003
>>>>> ima = (0.55 + p5) % 0.994 + 0.004
>>>>> ipa = ((0.11 + p2 + p3) * (0.66 + p5)) % 0.993 + 0.005
>>>>> ida = ((0.77 + p4) * (0.88 + p5)) % 0.992 + 0.006
>>>>> ini = ((0.99 + p4 + p5) * (0.23456789 + p2 + p3)) % 0.991 + 0.007
>>>>> iot_sa = 8 + int((22345 * isa) % 3)
>>>>> iot_ri = 5 + 4 * int((33456 * iri) % 3)
>>>>> iot_ga = 10 + 3 * int((44567 * iga) % 3)
>>>>> iot_ma = 8 + 8 * int((55678 * ima) % 3)
>>>>> iot_pa = 6 + 2 * int((66789 * ipa) % 4)
>>>>> iot_da = 8 + 5 * int((77890 * ida) % 3)
>>>>> iot_ni = 8 + 4 * int((88901 * ini) % 5)
>>>>> ift_sa  ftgentmp 0, 0, 65536, 10, 16, 8, 1/3, 4, 1/5, 1/6, 0, 2, 9 *
>>>>> isa, 1/10, 0, 1/12, 1/13, 0, 0, 1
>>>>> ift_ri  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iri), 1.0 - iga, 1.0 -
>>>>> ima, 1.0 - ipa, 1.0 - ida, 0, ida, 9 * iri, ini, 0, isa, iga, 0, 0,
>>>>> ima
>>>>> ift_ga  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - iga), 1.0 - ima, 1.0 -
>>>>> ipa, 1.0 - ida, 1.0 - ini, 0, ini, 9 * iga, isa, 0, iri, ima, 0, 0,
>>>>> ipa
>>>>> ift_ma  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ima), 1.0 - ipa, 1.0 -
>>>>> ida, 1.0 - ini, 1.0 - isa, 0, isa, 9 * ima, iri, 0, iga, ipa, 0, 0,
>>>>> ida
>>>>> ift_pa  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ipa), 1.0 - ida, 1.0 -
>>>>> ini, 1.0 - isa, 1.0 - iri, 0, iri, 9 * ipa, iga, 0, ima, ida, 0, 0,
>>>>> ini
>>>>> ift_da  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ida), 1.0 - ini, 1.0 -
>>>>> isa, 1.0 - iri, 1.0 - iga, 0, iga, 9 * ida, ima, 0, ipa, ini, 0, 0,
>>>>> isa
>>>>> ift_ni  ftgentmp 0, 0, 65536, 10, 9, 2 * (1.0 - ini), 1.0 - isa, 1.0 -
>>>>> iri, 1.0 - iga, 1.0 - ima, 0, ima, 9 * ini, ipa, 0, ida, isa, 0, 0,
>>>>> iri
>>>>> kenv_sa   linseg p4, p3, 0
>>>>> kenv_ri   linseg 0, iri * p3, p4, (1.0 - iri) * p3, 0
>>>>> kenv_ga   linseg 0, iga * p3, p4, (1.0 - iga) * p3, 0
>>>>> kenv_ma   linseg 0, ima * p3, p4, (1.0 - ima) * p3, 0
>>>>> kenv_pa   linseg 0, ipa * p3, p4, (1.0 - ipa) * p3, 0
>>>>> kenv_da   linseg 0, ida * p3, p4, (1.0 - ida) * p3, 0
>>>>> kenv_ni   linseg 0, ini * p3, p4, (1.0 - ini) * p3, 0
>>>>> asig_sa   oscil kenv_sa, p5, ift_sa, isa
>>>>> asig_ri   oscil kenv_ri, p5 * 8/9, ift_ri, iri
>>>>> asig_ga   oscil kenv_ga, p5 * 8/9, ift_ga, iga
>>>>> asig_ma   oscil kenv_ma, p5 * 9/8, ift_ma, ima
>>>>> asig_pa   oscil kenv_pa, p5 * 9/8, ift_pa, ipa
>>>>> asig_da   oscil kenv_da, p5 * iot_da/8, ift_da, ida
>>>>> asig_ni   oscil kenv_ni, p5 * iot_ni/8, ift_ni, ini
>>>>> outs    (asig_sa + asig_ri + asig_ma + asig_da) * giamp/4, (asig_sa +
>>>>> asig_ga + asig_pa + asig_ni) * giamp/4
>>>>> endin
>>>>>
>>>>> it was okay if the tune only lasts for about 1 min
>>>>> but for 3 min ... it eats up the ram
>>>>> can i still find a way to get an one and only ftable for each note?
>>>>>
>>>>> at this moment i simply replaced ftgentmp with ftgen
>>>>> so the using the same inst means using the same ftable
>>>>> the sound goes like this:
>>>>> http://sarigama.namaste.jp/loops/atrip2heaven.mp3
>>>>>
>>>>> tia!
>
> --
> SaRiGaMa's Oil Vending Orchestra
> is podcasting:
> http://sarigama.namaste.jp/podcast/rss.xml
> and supplying oil.py for free:
> http://oilpy.blogspot.com/
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"