Nabble - Csound - Dev

Re: bus opcodes

2009-11-27T07:20:20Z

This didn't seem to go through to the list though I sent it last night,
sending it again and also to the Csound list...

MKG

----- Original Message -----
From: <Michael.Gogins@...>
To: "Developer discussions" <csound-devel@...>
Sent: Thursday, November 26, 2009 11:16 PM
Subject: Re: [Cs-dev] bus opcodes

>I have mentioned before that Peter Brinkmann's jReality package, written in
> Java, for mathematics visualization and sonification can use Csound to
> attach sounds to visual objects. Here is Peter's source code, which I
> helped
> him to write, from jReality SVN for sending audio output from Csound to
> the
> audio backend of jReality. I have heard this code run. I imagine similar
> code would work in Lua, and I may verify that if I have time.
>
> The key lines are as follows, with my comments added:
>
> // SWIG-generated wrapper for an array of MYFLT.
> spout = new CsoundMYFLTArray();
> // Set the SWIG wrapper for Csound's spout buffer pointer to the MYFLT
> array
> wrapper. The "Ptr" is a SWIG wrapper for a C MYFLT * pointer.
> // In other words, in SWIG-generated code you never see a truly raw
> pointer.
> // THIS IS PROBABLY YOUR MISSING STEP, GENTLEMEN.
> spout.SetPtr(csnd.GetSpout());
> // Create a native host language buffer for sending monophonic audio to
> the
> backend.
> cumulativeBuffer = new float[ksmps];
>
> // Call on Csound to synthesize ksmps worth of audio.
> // ANOTHER POSSIBLE "GOTCHA": MUST USE csoundPerformKsmps NOT
> csoundPerformBuffer.
> csnd.PerformKsmps();
> // Transfer audio, framewise, from the SWIG wrapper for spout to my native
> buffer.
> // I.e., j is the SAMPLE FRAME index.
> for(int j=0; j<ksmps; j++) {
> // Average N-channel output to my monophonic native buffer.
> // Note that v collects only one sample if nchnls = 1.
> float v = 0;
> // I.e., k is the CHANNEL index (and ksmps is the channel "stride" in
> spout).
> for(int k=j; k<bufSize; k+=ksmps) {
> v += spout.GetValue(k);
> }
> cumulativeBuffer[j] = v/scale;
> }
> // Write my native audio buffer to my backend.
> ringBuffer.write(cumulativeBuffer, 0, ksmps);
> }
> }
>
> For N-channel final output, instead of
>
> v+= spout.GetValue(k)
>
> you should have
>
> frame[channel] = spout.getValue(k)
>
> In spite of the fact that this works, I will add to the API more direct,
> samplewise access to spout and spin. The above would be more efficient for
> ksmps above a certain size, but I'm not sure what that size would be; I
> expect for low latency (ksmps of 20 or so) samplewise access should be
> about
> as fast.
>
> I imagine the reason SWIG takes this circuitous route is that, in some
> languages, you can directly index a pointer as though it is an array,
> whereas in other languages pointers do not directly appear. The array
> wrapper makes both kinds of languages act the same way.
>
> Hope this helps,
> Mike
>
> ----- Original Message -----
> From: "jhearon" <j_hearon@...>
> To: <csound-devel@...>
> Sent: Thursday, November 26, 2009 8:51 PM
> Subject: Re: [Cs-dev] bus opcodes
>
>
>>
>>
>>
>> You may remember that i failed to get the JAVA API to access audio
>> channels last summer, and there still is no explanation
>> ==John ffitch
>>
>> Hi John,
>> This was some old java code posted a long time ago which is supposed to
>> show
>> channel communication to/from a java host using .csnd.jar methods. I'm
>> no
>> Java expert but sort of have this working. Maybe Steven Yi or J.P.
>> Lemoine
>> are better at java. The code I was working on was mostly using chnget to
>> get a k-value for frequency from the java host using the
>> csoundGetChannelPtr
>> method and passing the java wrapped value of MYFLT **p to csound.
>>
>> //Main.java
>> public class Main {
>> public static void main(String[] args) {
>> chntest_thread c = new chntest_thread();
>> }
>> }
>> //--------------
>> //chntest_thread.java
>>
>> import csnd.*;
>>
>> public class chntest_thread {
>> SWIGTYPE_p_void myvoid;
>> //SWIGTYPE_p_void myvoid = null;//if above gives compiler error
>> SWIGTYPE_p_CSOUND_ mycsound = csnd.csoundCreate(myvoid);
>> Csound csound = new Csound();
>> CsoundMYFLTArray myfltarray = new CsoundMYFLTArray();
>> CsoundMYFLTArray chn1 = new CsoundMYFLTArray();
>> CsoundMYFLTArray chn2 = new CsoundMYFLTArray();
>> CsoundMYFLTArray chn3 = new CsoundMYFLTArray();
>> CsoundMYFLTArray chn4 = new CsoundMYFLTArray();
>> CsoundMYFLTArray chn5 = new CsoundMYFLTArray();
>>
>> chntest_thread() {
>> run();
>> }
>>
>> public void run() {
>> int err, ksmps, kcnt = 0;
>> double phs = 0.0, phs_inc;
>> err = csound.Compile("chnTest.csd");
>> if (err == 0) {
>> err |= csnd.csoundGetChannelPtr(mycsound, chn1.GetPtr(), "k_in",
>> csndConstants.CSOUND_CONTROL_CHANNEL
>> | csndConstants.CSOUND_INPUT_CHANNEL);
>> err |= csnd.csoundGetChannelPtr(mycsound, chn2.GetPtr(), "a_in",
>> csndConstants.CSOUND_AUDIO_CHANNEL
>> | csndConstants.CSOUND_INPUT_CHANNEL);
>> err |= csnd.csoundGetChannelPtr(mycsound, chn3.GetPtr(), "S_in",
>> csndConstants.CSOUND_STRING_CHANNEL
>> | csndConstants.CSOUND_INPUT_CHANNEL);
>> err |= csnd.csoundGetChannelPtr(mycsound, chn4.GetPtr(), "k_out",
>> csndConstants.CSOUND_CONTROL_CHANNEL
>> | csndConstants.CSOUND_OUTPUT_CHANNEL);
>> err |= csnd.csoundGetChannelPtr(mycsound, chn5.GetPtr(), "S_out",
>> csndConstants.CSOUND_STRING_CHANNEL
>> | csndConstants.CSOUND_OUTPUT_CHANNEL);
>> if (err == 0) {
>> chn1.SetValue(0, 3.14159265);
>> chn3.SetStringValue("String Input Value",
>> csound.GetStrVarMaxLen());
>> phs_inc = 2.0 * 3.14159265 * 440.0 / csound.GetSr();
>> ksmps = csound.GetKsmps();
>> do {
>> for (int i = 0; i < ksmps; i++) {
>> chn2.SetValue(i, 0.7071 * Math.sin(phs));
>> phs += phs_inc;
>> }
>> err = csound.PerformKsmps();
>> if (kcnt == 0) {
>> String s = chn5.GetStringValue();
>> System.err.printf("Values sent: %g, '%s'\n", chn4.GetValue(0),
>> s);
>> }
>> kcnt++;
>> } while (err == 0);
>> }
>> }
>> csound.Reset();
>> java.lang.System.exit(1);
>> } //end run method
>> }//ends chntest_thread
>> //-----------
>> //chnTest.csd
>>
>> <CsoundSynthesizer>
>> <CsOptions>
>> -s -d -+rtaudio=ALSA -odac:hw:0,0 -b1024 -B16384
>> </CsOptions>
>> <CsInstruments>
>> sr = 44100
>> ksmps = 32
>> nchnls = 2
>> 0dbfs = 1
>> chn_k "k_in", 1
>> chn_a "a_in", 1
>> chn_S "S_in", 1
>> chn_k "k_out", 2
>> chn_S "S_out", 2
>>
>> chnset 2.71828, "k_out"
>> chnset "Output String Value", "S_out"
>>
>> instr 1
>> ichn1 chnget "k_in"
>> achn2 chnget "a_in"
>> Schn3 chnget "S_in"
>>
>> printf_i "Values received: %g, '%s'\n", 1, ichn1, Schn3
>> outs achn2, achn2
>> endin
>>
>> </CsInstruments>
>> <CsScore>
>> i 1 0 2
>> e
>> </CsScore>
>> </CsoundSynthesizer>
>>
>> --
>> View this message in context:
>> http://old.nabble.com/bus-opcodes-tp26362662p26535852.html
>> Sent from the Csound - Dev mailing list archive at Nabble.com.
>>
>>
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
>> 30-Day
>> trial. Simplify your report design, integration and deployment - and
>> focus
>> on
>> what you do best, core application coding. Discover what's new with
>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@...
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>
>

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: bus opcodes

2009-11-26T20:16:26Z

I have mentioned before that Peter Brinkmann's jReality package, written in
Java, for mathematics visualization and sonification can use Csound to
attach sounds to visual objects. Here is Peter's source code, which I helped
him to write, from jReality SVN for sending audio output from Csound to the
audio backend of jReality. I have heard this code run. I imagine similar
code would work in Lua, and I may verify that if I have time.

The key lines are as follows, with my comments added:

// SWIG-generated wrapper for an array of MYFLT.
spout = new CsoundMYFLTArray();
// Set the SWIG wrapper for Csound's spout buffer pointer to the MYFLT array
wrapper. The "Ptr" is a SWIG wrapper for a C MYFLT * pointer.
// In other words, in SWIG-generated code you never see a truly raw pointer.
// THIS IS PROBABLY YOUR MISSING STEP, GENTLEMEN.
spout.SetPtr(csnd.GetSpout());
// Create a native host language buffer for sending monophonic audio to the
backend.
cumulativeBuffer = new float[ksmps];

// Call on Csound to synthesize ksmps worth of audio.
// ANOTHER POSSIBLE "GOTCHA": MUST USE csoundPerformKsmps NOT
csoundPerformBuffer.
csnd.PerformKsmps();
// Transfer audio, framewise, from the SWIG wrapper for spout to my native
buffer.
// I.e., j is the SAMPLE FRAME index.
for(int j=0; j<ksmps; j++) {
// Average N-channel output to my monophonic native buffer.
// Note that v collects only one sample if nchnls = 1.
float v = 0;
// I.e., k is the CHANNEL index (and ksmps is the channel "stride" in
spout).
for(int k=j; k<bufSize; k+=ksmps) {
v += spout.GetValue(k);
}
cumulativeBuffer[j] = v/scale;
}
// Write my native audio buffer to my backend.
ringBuffer.write(cumulativeBuffer, 0, ksmps);
}
}

For N-channel final output, instead of

v+= spout.GetValue(k)

you should have

frame[channel] = spout.getValue(k)

In spite of the fact that this works, I will add to the API more direct,
samplewise access to spout and spin. The above would be more efficient for
ksmps above a certain size, but I'm not sure what that size would be; I
expect for low latency (ksmps of 20 or so) samplewise access should be about
as fast.

I imagine the reason SWIG takes this circuitous route is that, in some
languages, you can directly index a pointer as though it is an array,
whereas in other languages pointers do not directly appear. The array
wrapper makes both kinds of languages act the same way.

Hope this helps,
Mike

----- Original Message -----
From: "jhearon" <j_hearon@...>
To: <csound-devel@...>
Sent: Thursday, November 26, 2009 8:51 PM
Subject: Re: [Cs-dev] bus opcodes

>
>
>
> You may remember that i failed to get the JAVA API to access audio
> channels last summer, and there still is no explanation
> ==John ffitch
>
> Hi John,
> This was some old java code posted a long time ago which is supposed to
> show
> channel communication to/from a java host using .csnd.jar methods. I'm no
> Java expert but sort of have this working. Maybe Steven Yi or J.P.
> Lemoine
> are better at java. The code I was working on was mostly using chnget to
> get a k-value for frequency from the java host using the
> csoundGetChannelPtr
> method and passing the java wrapped value of MYFLT **p to csound.
>
> //Main.java
> public class Main {
> public static void main(String[] args) {
> chntest_thread c = new chntest_thread();
> }
> }
> //--------------
> //chntest_thread.java
>
> import csnd.*;
>
> public class chntest_thread {
> SWIGTYPE_p_void myvoid;
> //SWIGTYPE_p_void myvoid = null;//if above gives compiler error
> SWIGTYPE_p_CSOUND_ mycsound = csnd.csoundCreate(myvoid);
> Csound csound = new Csound();
> CsoundMYFLTArray myfltarray = new CsoundMYFLTArray();
> CsoundMYFLTArray chn1 = new CsoundMYFLTArray();
> CsoundMYFLTArray chn2 = new CsoundMYFLTArray();
> CsoundMYFLTArray chn3 = new CsoundMYFLTArray();
> CsoundMYFLTArray chn4 = new CsoundMYFLTArray();
> CsoundMYFLTArray chn5 = new CsoundMYFLTArray();
>
> chntest_thread() {
> run();
> }
>
> public void run() {
> int err, ksmps, kcnt = 0;
> double phs = 0.0, phs_inc;
> err = csound.Compile("chnTest.csd");
> if (err == 0) {
> err |= csnd.csoundGetChannelPtr(mycsound, chn1.GetPtr(), "k_in",
> csndConstants.CSOUND_CONTROL_CHANNEL
> | csndConstants.CSOUND_INPUT_CHANNEL);
> err |= csnd.csoundGetChannelPtr(mycsound, chn2.GetPtr(), "a_in",
> csndConstants.CSOUND_AUDIO_CHANNEL
> | csndConstants.CSOUND_INPUT_CHANNEL);
> err |= csnd.csoundGetChannelPtr(mycsound, chn3.GetPtr(), "S_in",
> csndConstants.CSOUND_STRING_CHANNEL
> | csndConstants.CSOUND_INPUT_CHANNEL);
> err |= csnd.csoundGetChannelPtr(mycsound, chn4.GetPtr(), "k_out",
> csndConstants.CSOUND_CONTROL_CHANNEL
> | csndConstants.CSOUND_OUTPUT_CHANNEL);
> err |= csnd.csoundGetChannelPtr(mycsound, chn5.GetPtr(), "S_out",
> csndConstants.CSOUND_STRING_CHANNEL
> | csndConstants.CSOUND_OUTPUT_CHANNEL);
> if (err == 0) {
> chn1.SetValue(0, 3.14159265);
> chn3.SetStringValue("String Input Value",
> csound.GetStrVarMaxLen());
> phs_inc = 2.0 * 3.14159265 * 440.0 / csound.GetSr();
> ksmps = csound.GetKsmps();
> do {
> for (int i = 0; i < ksmps; i++) {
> chn2.SetValue(i, 0.7071 * Math.sin(phs));
> phs += phs_inc;
> }
> err = csound.PerformKsmps();
> if (kcnt == 0) {
> String s = chn5.GetStringValue();
> System.err.printf("Values sent: %g, '%s'\n", chn4.GetValue(0),
> s);
> }
> kcnt++;
> } while (err == 0);
> }
> }
> csound.Reset();
> java.lang.System.exit(1);
> } //end run method
> }//ends chntest_thread
> //-----------
> //chnTest.csd
>
> <CsoundSynthesizer>
> <CsOptions>
> -s -d -+rtaudio=ALSA -odac:hw:0,0 -b1024 -B16384
> </CsOptions>
> <CsInstruments>
> sr = 44100
> ksmps = 32
> nchnls = 2
> 0dbfs = 1
> chn_k "k_in", 1
> chn_a "a_in", 1
> chn_S "S_in", 1
> chn_k "k_out", 2
> chn_S "S_out", 2
>
> chnset 2.71828, "k_out"
> chnset "Output String Value", "S_out"
>
> instr 1
> ichn1 chnget "k_in"
> achn2 chnget "a_in"
> Schn3 chnget "S_in"
>
> printf_i "Values received: %g, '%s'\n", 1, ichn1, Schn3
> outs achn2, achn2
> endin
>
> </CsInstruments>
> <CsScore>
> i 1 0 2
> e
> </CsScore>
> </CsoundSynthesizer>
>
> --
> View this message in context:
> http://old.nabble.com/bus-opcodes-tp26362662p26535852.html
> Sent from the Csound - Dev mailing list archive at Nabble.com.
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
> 30-Day
> trial. Simplify your report design, integration and deployment - and focus
> on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

CsoundSource.java (2K) Download Attachment

Re: bus opcodes

2009-11-26T18:57:36Z

Hi Jim,

I think the problems for John and Java were for working with a-sig
variables and not k-vars. I haven't had much luck myself but was going
to look into it. (Got sidetracked)

Steven

On 11/26/09, jhearon <j_hearon@...> wrote:

>
>
>
> You may remember that i failed to get the JAVA API to access audio
> channels last summer, and there still is no explanation
> ==John ffitch
>
> Hi John,
> This was some old java code posted a long time ago which is supposed to show
> channel communication to/from a java host using .csnd.jar methods. I'm no
> Java expert but sort of have this working. Maybe Steven Yi or J.P. Lemoine
> are better at java. The code I was working on was mostly using chnget to
> get a k-value for frequency from the java host using the csoundGetChannelPtr
> method and passing the java wrapped value of MYFLT **p to csound.
>
> //Main.java
> public class Main {
> public static void main(String[] args) {
> chntest_thread c = new chntest_thread();
> }
> }
> //--------------
> //chntest_thread.java
>
> import csnd.*;
>
> public class chntest_thread {
> SWIGTYPE_p_void myvoid;
> //SWIGTYPE_p_void myvoid = null;//if above gives compiler error
> SWIGTYPE_p_CSOUND_ mycsound = csnd.csoundCreate(myvoid);
> Csound csound = new Csound();
> CsoundMYFLTArray myfltarray = new CsoundMYFLTArray();
> CsoundMYFLTArray chn1 = new CsoundMYFLTArray();
> CsoundMYFLTArray chn2 = new CsoundMYFLTArray();
> CsoundMYFLTArray chn3 = new CsoundMYFLTArray();
> CsoundMYFLTArray chn4 = new CsoundMYFLTArray();
> CsoundMYFLTArray chn5 = new CsoundMYFLTArray();
>
> chntest_thread() {
> run();
> }
>
> public void run() {
> int err, ksmps, kcnt = 0;
> double phs = 0.0, phs_inc;
> err = csound.Compile("chnTest.csd");
> if (err == 0) {
> err |= csnd.csoundGetChannelPtr(mycsound, chn1.GetPtr(), "k_in",
> csndConstants.CSOUND_CONTROL_CHANNEL
> | csndConstants.CSOUND_INPUT_CHANNEL);
> err |= csnd.csoundGetChannelPtr(mycsound, chn2.GetPtr(), "a_in",
> csndConstants.CSOUND_AUDIO_CHANNEL
> | csndConstants.CSOUND_INPUT_CHANNEL);
> err |= csnd.csoundGetChannelPtr(mycsound, chn3.GetPtr(), "S_in",
> csndConstants.CSOUND_STRING_CHANNEL
> | csndConstants.CSOUND_INPUT_CHANNEL);
> err |= csnd.csoundGetChannelPtr(mycsound, chn4.GetPtr(), "k_out",
> csndConstants.CSOUND_CONTROL_CHANNEL
> | csndConstants.CSOUND_OUTPUT_CHANNEL);
> err |= csnd.csoundGetChannelPtr(mycsound, chn5.GetPtr(), "S_out",
> csndConstants.CSOUND_STRING_CHANNEL
> | csndConstants.CSOUND_OUTPUT_CHANNEL);
> if (err == 0) {
> chn1.SetValue(0, 3.14159265);
> chn3.SetStringValue("String Input Value", csound.GetStrVarMaxLen());
> phs_inc = 2.0 * 3.14159265 * 440.0 / csound.GetSr();
> ksmps = csound.GetKsmps();
> do {
> for (int i = 0; i < ksmps; i++) {
> chn2.SetValue(i, 0.7071 * Math.sin(phs));
> phs += phs_inc;
> }
> err = csound.PerformKsmps();
> if (kcnt == 0) {
> String s = chn5.GetStringValue();
> System.err.printf("Values sent: %g, '%s'\n", chn4.GetValue(0),
> s);
> }
> kcnt++;
> } while (err == 0);
> }
> }
> csound.Reset();
> java.lang.System.exit(1);
> } //end run method
> }//ends chntest_thread
> //-----------
> //chnTest.csd
>
> <CsoundSynthesizer>
> <CsOptions>
> -s -d -+rtaudio=ALSA -odac:hw:0,0 -b1024 -B16384
> </CsOptions>
> <CsInstruments>
> sr = 44100
> ksmps = 32
> nchnls = 2
> 0dbfs = 1
> chn_k "k_in", 1
> chn_a "a_in", 1
> chn_S "S_in", 1
> chn_k "k_out", 2
> chn_S "S_out", 2
>
> chnset 2.71828, "k_out"
> chnset "Output String Value", "S_out"
>
> instr 1
> ichn1 chnget "k_in"
> achn2 chnget "a_in"
> Schn3 chnget "S_in"
>
> printf_i "Values received: %g, '%s'\n", 1, ichn1, Schn3
> outs achn2, achn2
> endin
>
> </CsInstruments>
> <CsScore>
> i 1 0 2
> e
> </CsScore>
> </CsoundSynthesizer>
>
> --
> View this message in context:
> http://old.nabble.com/bus-opcodes-tp26362662p26535852.html
> Sent from the Csound - Dev mailing list archive at Nabble.com.
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus
> on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

--
Sent from my mobile device

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: bus opcodes

2009-11-26T17:51:46Z

You may remember that i failed to get the JAVA API to access audio
channels last summer, and there still is no explanation
==John ffitch

Hi John,
This was some old java code posted a long time ago which is supposed to show channel communication to/from a java host using .csnd.jar methods. I'm no Java expert but sort of have this working. Maybe Steven Yi or J.P. Lemoine are better at java. The code I was working on was mostly using chnget to get a k-value for frequency from the java host using the csoundGetChannelPtr method and passing the java wrapped value of MYFLT **p to csound.

//Main.java
public class Main {
public static void main(String[] args) {
chntest_thread c = new chntest_thread();
}
}
//--------------
//chntest_thread.java

import csnd.*;

public class chntest_thread {
SWIGTYPE_p_void myvoid;
//SWIGTYPE_p_void myvoid = null;//if above gives compiler error
SWIGTYPE_p_CSOUND_ mycsound = csnd.csoundCreate(myvoid);
Csound csound = new Csound();
CsoundMYFLTArray myfltarray = new CsoundMYFLTArray();
CsoundMYFLTArray chn1 = new CsoundMYFLTArray();
CsoundMYFLTArray chn2 = new CsoundMYFLTArray();
CsoundMYFLTArray chn3 = new CsoundMYFLTArray();
CsoundMYFLTArray chn4 = new CsoundMYFLTArray();
CsoundMYFLTArray chn5 = new CsoundMYFLTArray();

chntest_thread() {
run();
}

public void run() {
int err, ksmps, kcnt = 0;
double phs = 0.0, phs_inc;
err = csound.Compile("chnTest.csd");
if (err == 0) {
err |= csnd.csoundGetChannelPtr(mycsound, chn1.GetPtr(), "k_in",
csndConstants.CSOUND_CONTROL_CHANNEL
| csndConstants.CSOUND_INPUT_CHANNEL);
err |= csnd.csoundGetChannelPtr(mycsound, chn2.GetPtr(), "a_in",
csndConstants.CSOUND_AUDIO_CHANNEL
| csndConstants.CSOUND_INPUT_CHANNEL);
err |= csnd.csoundGetChannelPtr(mycsound, chn3.GetPtr(), "S_in",
csndConstants.CSOUND_STRING_CHANNEL
| csndConstants.CSOUND_INPUT_CHANNEL);
err |= csnd.csoundGetChannelPtr(mycsound, chn4.GetPtr(), "k_out",
csndConstants.CSOUND_CONTROL_CHANNEL
| csndConstants.CSOUND_OUTPUT_CHANNEL);
err |= csnd.csoundGetChannelPtr(mycsound, chn5.GetPtr(), "S_out",
csndConstants.CSOUND_STRING_CHANNEL
| csndConstants.CSOUND_OUTPUT_CHANNEL);
if (err == 0) {
chn1.SetValue(0, 3.14159265);
chn3.SetStringValue("String Input Value", csound.GetStrVarMaxLen());
phs_inc = 2.0 * 3.14159265 * 440.0 / csound.GetSr();
ksmps = csound.GetKsmps();
do {
for (int i = 0; i < ksmps; i++) {
chn2.SetValue(i, 0.7071 * Math.sin(phs));
phs += phs_inc;
}
err = csound.PerformKsmps();
if (kcnt == 0) {
String s = chn5.GetStringValue();
System.err.printf("Values sent: %g, '%s'\n", chn4.GetValue(0), s);
}
kcnt++;
} while (err == 0);
}
}
csound.Reset();
java.lang.System.exit(1);
} //end run method
}//ends chntest_thread
//-----------
//chnTest.csd

<CsoundSynthesizer>
<CsOptions>
-s -d -+rtaudio=ALSA -odac:hw:0,0 -b1024 -B16384
</CsOptions>
<CsInstruments>
sr = 44100
ksmps = 32
nchnls = 2
0dbfs = 1
chn_k "k_in", 1
chn_a "a_in", 1
chn_S "S_in", 1
chn_k "k_out", 2
chn_S "S_out", 2

chnset 2.71828, "k_out"
chnset "Output String Value", "S_out"

instr 1
ichn1 chnget "k_in"
achn2 chnget "a_in"
Schn3 chnget "S_in"

printf_i "Values received: %g, '%s'\n", 1, ichn1, Schn3
outs achn2, achn2
endin

</CsInstruments>
<CsScore>
i 1 0 2
e
</CsScore>
</CsoundSynthesizer>

Re: can't open device

2009-11-26T14:13:55Z

On Thu, 2009-11-26 at 19:27 +0000, victor wrote:
> Sounds like a dmix issue. There was a patch fixing the buffer size issue for
> use with the dmix only last week or the week before.

Ah, I see that CVS auto-corrects the -b and -B values. I've checked the
sources, though, and thought the default for alsa should change from
"plughw" to "default". I'm not quite sure how to make portaudio use the
default alsa interface instead of whatever it is choosing. Does anyone
object to that change?

--
Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

signature.asc (853 bytes) Download Attachment

Re: can't open device

2009-11-26T11:27:39Z

Sounds like a dmix issue. There was a patch fixing the buffer size issue for
use with the dmix only last week or the week before.

Victor
----- Original Message -----
From: "Felipe Sateler" <fsateler@...>
To: "Developer discussions" <csound-devel@...>
Sent: Thursday, November 26, 2009 6:22 PM
Subject: [Cs-dev] can't open device

> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
> 30-Day
> trial. Simplify your report design, integration and deployment - and focus
> on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july

--------------------------------------------------------------------------------

> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

can't open device

2009-11-26T10:22:12Z

When I have my music player paused (or any program with audio, but not
necessarily audio running), csound can't play, either portaudio or alsa.
When using portaudio I get:

PortAudio V19-devel (built Sep 1 2009)
PortAudio: available output devices:
0: HDA Intel: STAC92xx Digital (hw:0,1)
1: iec958
2: spdif
3: default
4: dmix
PortAudio: failed to obtain device info.
Failed to initialize real time audio output

And with alsa I get

*** Can't open device 'plughw' for audio output: Device or resource busy

Now, I could say that the ALSA driver is buggy and doesn't allow
multiple apps at the same time, but I can open any number of other audio
applications at the same time (eg, mplayer, vlc, youtube videos) without
problems.

Any hints on why is this? My soundcard is an Intel HDA (the integrated
chip on my inspiron 1420 notebook).

Hmm, I just discovered that if I use -odac:default -B 1881 -b 1881 I can
get sound. csound insisted that I set -B 1881 and the only way to do
that was to set -b 1881, since otherwise csound rounds up to the next
multiple of -b.

--
Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

signature.asc (853 bytes) Download Attachment

Re: bus opcodes

2009-11-26T03:01:23Z

You may remember that i failed to get the JAVA API to access audio
channels last summer, and there still is no explanation
==John ffitch

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: UB binary of csound framework

2009-11-26T00:40:20Z

I don't want to discourage you completely, I think if you feel strongly about it you should

have a go. The main place to look in are files in the InOut directory. The best thing is to try to keep

the behaviour of FileOpen, FileOpen2, etc unchanged. I think most of the code will access files through

these functions.

Victor

----- Original Message -----

From: matt@...

To: csound-devel@...

Sent: Wednesday, November 25, 2009 10:19 PM

Subject: Re: [Cs-dev] UB binary of csound framework

On Nov 25, 2009, at 1:38 PM, victor wrote:

I think they are both (difficult and dangerous). Also will require an API bump.

As for "in", the same could be done with an opcode.

fair enough

If you want to send in the filename,

a macro would do the job. But you can also run lots of soundfiles in one csound run. My opinion

of -i and in (regarding soundfiles) does not change, it's not that useful or flexible. IMHO it is a

legacy feature.

well the MacCsound soundfile effects feature does rely heavily on it..

Implementing MP3 output as a plugin is (IMHO) the best option.

perhaps, but it definitely would be a smoother user experience to have mp3 work just like any other file in all places throughout the program. anyway, i was just citing mp3 as ONE EXAMPLE why

having the flexibility to optionally not use libsndfile seems desirable.

but if it is going to be a pain in the butt i don't see it as worth the effort at this point and i actually was more interested in just being able to compile and/or run csound without it.

-m

matt ingalls

matt@...

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july

_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: UB binary of csound framework

2009-11-25T14:19:51Z

On Nov 25, 2009, at 1:38 PM, victor wrote:

I think they are both (difficult and dangerous). Also will require an API bump.

As for "in", the same could be done with an opcode.

fair enough

If you want to send in the filename,
a macro would do the job. But you can also run lots of soundfiles in one csound run. My opinion
of -i and in (regarding soundfiles) does not change, it's not that useful or flexible. IMHO it is a
legacy feature.

well the MacCsound soundfile effects feature does rely heavily on it..

Implementing MP3 output as a plugin is (IMHO) the best option.

perhaps, but it definitely would be a smoother user experience to have mp3 work just like any other file in all places throughout the program. anyway, i was just citing mp3 as ONE EXAMPLE why

having the flexibility to optionally not use libsndfile seems desirable.

but if it is going to be a pain in the butt i don't see it as worth the effort at this point and i actually was more interested in just being able to compile and/or run csound without it.

-m

matt ingalls

matt@...

Re: bus opcodes

2009-11-25T13:52:55Z

Hi Matt,

On Wed, Nov 25, 2009 at 12:05 AM, matt ingalls <matt@...> wrote:
> and strings, please.
>
> at least at one point you could put strings inside of the invalue/outvalue.
> although it was kind of a hack and it might have been removed
> (it doesn't seem to be working for me right now)

It's working, I use it in QuteCsound.

Cheers,
Andrés

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: bus opcodes

2009-11-25T13:50:21Z

I think you could do a bit of a hack using fractional instrument
numbers, and then doing something like:

Schan sprinfk "id_%f", p1
kval invalue Schan

But I do agree the situation with invalue vs. chnrecv is not ideal. I
would think that since the functionality of invalue and outvalue is
covered and extended by chnrecv and chnsend, they should be merged, to
use a single code base, marking one of the API functions to set the
callbacks as deprecated to remove in the future (and discouranging use
of the string passing hack in invalue/outvalue).

Cheers,
Andrés

On Tue, Nov 24, 2009 at 1:40 PM, Jonatan Liljedahl <lijon@...> wrote:

> Regarding this... there's a lot:
>
> chnset/chnget is not callback based, both k and a-rate.
>
> chani/chano is not callback based, both k and a-rate.
>
> invalue/outvalue is callback based but doesn't support a-rate.
>
> chnrecv/chnsend is callback and supports a-rate according to csound.h
> but they are not documented in the manual, do they exist?
>
> This surely is a bit messy. Also there are differences regarding channel
> names vs channel numbers. It's a lot easier to dynamically change a
> channel if it's a number, like passing in 5 in a p4 and use p4+0, p4+1,
> p4+2, p4+3 as channelnumbers in the instrument...
>
> I'm currently looking for a way to dynamically connect csound instrument
> _instances_ to/from each other and to/from the host through the API to
> have a timeline-based graphical signal flow language. (AlgoScore2)
>
> This means I need to use a callback-based approach, and that the
> callback needs to identify which instr instance is sending/getting a
> value. This doesn't seem to be possible right now?
>
> One work-around would be to pass in an unique ID in a p-field and in the
> instr use something like this:
>
> chnsend "_id", p4
> k1 chnrecv "foo"
> a1 chnrecv "bar"
>
> The callback would then react on a value on channel "_id" by storing the
> ID in a variable in the Csound host data, so that in the other calls it
> will know which instr instance is asking for data.
>
> Surely this would work. But then comes another problem: if this is used
> to send stuff between instr instances, then how do I control the order
> of their performance? If I have A->B->C then they must be performed in
> that order for the data to be available for the next instance down the
> chain. Note that this is about instrument _instances_, not the static
> instrument declarations. A,B, and C could be instances of the same
> instr. Or if it's different instruments, they could later on be used
> A->C->B, etc.. So the order of instrument declarations in the orchestra
> file is of no use.
>
> The only thing I can think of is if the order of events sent to csound
> affects their order of performance, is this the case?
>
> i1 2 1
> i1 0 3
>
> does the second instance perform after the first when both instances is
> performing? A quick test indicates no. :(
>
> Any ideas? How about the new signal flow opcodes, do they work on
> instance-level or only a global instrument-level?
>
> jpff@... wrote:
>> Perhaps it is time to remove chani and chano as no one else seems to
>> use/want them.
>>
>>> And because Csound does everything for you (creates channels, keeps
>>> track of bus names etc).
>>>
>>> On 16 Nov 2009, at 07:19, Andres Cabrera wrote:
>>>
>>>> Hi James,
>>>>
>>>> There are two channel busses. One using chnget/chnset and the other
>>>> using invalue/outvalue. I prefer invalue/outvalue as they are callback
>>>> based, so you can know and act whenever a value is changed on either
>>>> side, while chnget/chnset are updated at k-rate. The chnget/chnset
>>>> system is better in the way it passes strings, so this would be a
>>>> reason to use it.
>>>>
>>>> Cheers,
>>>> Andrés
>>>>
>>>> On Sun, Nov 15, 2009 at 8:21 PM, Victor Lazzarini
>>>> <Victor.Lazzarini@...> wrote:
>>>>> in C:
>>>>> int csoundGetChannelPtr(CSOUND *,
>>>>> MYFLT **p, const char *name, int
>>>>> type);
>>>>> in C++:
>>>>> Csound::GetChannel() and Csound::SetChannel() (not audio rate)
>>>>> or
>>>>> Csound::GetChannelPtr() (any rate)
>>>>>
>>>>>
>>>>> On 15 Nov 2009, at 19:55, James Hearon wrote:
>>>>>
>>>>> Hi,
>>>>> I compiled vers. 5.11 and am looking at API trying to figure out
>>>>> what is
>>>>> appropriate method for control and events using bus opcodes chnget,
>>>>> chnset?
>>>>>
>>>>> Thanks
>>>>>
>>>>>
>>>>>
>>>>> ________________________________
>>>>> Hotmail: Trusted email with powerful SPAM protection. Sign up
>>>>> now.
>>>>> ------------------------------------------------------------------------------
>>>>> Let Crystal Reports handle the reporting - Free Crystal Reports
>>>>> 2008 30-Day
>>>>> trial. Simplify your report design, integration and deployment -
>>>>> and focus
>>>>> on
>>>>> what you do best, core application coding. Discover what's new with
>>>>> Crystal Reports now.
>>>>> http://p.sf.net/sfu/bobj-july_______________________________________________
>>>>> Csound-devel mailing list
>>>>> Csound-devel@...
>>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------------
>>>>> Let Crystal Reports handle the reporting - Free Crystal Reports
>>>>> 2008 30-Day
>>>>> trial. Simplify your report design, integration and deployment -
>>>>> and focus
>>>>> on
>>>>> what you do best, core application coding. Discover what's new with
>>>>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>>>>> _______________________________________________
>>>>> Csound-devel mailing list
>>>>> Csound-devel@...
>>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>>
>>>>
>>>> Andrés
>>>>
>>>> ------------------------------------------------------------------------------
>>>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
>>>> 30-Day
>>>> trial. Simplify your report design, integration and deployment - and
>>>> focus on
>>>> what you do best, core application coding. Discover what's new with
>>>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>>>> _______________________________________________
>>>> Csound-devel mailing list
>>>> Csound-devel@...
>>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>
>>> ------------------------------------------------------------------------------
>>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
>>> 30-Day
>>> trial. Simplify your report design, integration and deployment - and focus
>>> on
>>> what you do best, core application coding. Discover what's new with
>>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@...
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>
>>>
>>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
>> trial. Simplify your report design, integration and deployment - and focus on
>> what you do best, core application coding. Discover what's new with
>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@...
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

--

Andrés

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: UB binary of csound framework

2009-11-25T13:38:07Z

I think they are both (difficult and dangerous). Also will require an API bump.

As for "in", the same could be done with an opcode. If you want to send in the filename,

a macro would do the job. But you can also run lots of soundfiles in one csound run. My opinion

of -i and in (regarding soundfiles) does not change, it's not that useful or flexible. IMHO it is a

legacy feature.

Implementing MP3 output as a plugin is (IMHO) the best option. But we won't be able to distribute

it in soucerforge AFAIK.

Victor

maybe it was not clear -- i was offering to do this work myself - i was

just asking if any of you thought it was too difficult or dangerous.

matt ingalls

matt@...

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july

_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: UB binary of csound framework

2009-11-25T11:38:42Z

Using in with soundfiles is not particularly useful,

what? passing an input file on the commandline is *definitely* useful!

it is great for any kind of file processing -- i use this a lot for post-

render. like this normalizing instr:

instr 1

ipeak filepeak "-i"

iscale = (ipeak == 0 ? 1 : 0dbfs/ipeak)

aout1, aout2 ins

outs iscale*aout1, iscale*aout2

endin

Not too much of a limitation, IMHO. Certainly not a reason to ditch

libsndfile.

i never suggested to "ditch" libsndfile!

all i was suggesting is to not have it hard-wired to csound, so one

would have the FLEXIBILITY to not include it for whatever sensible

or stupid reason.

it really surprised me that you do not find this suggestion reasonable or useful.

maybe it was not clear -- i was offering to do this work myself - i was

just asking if any of you thought it was too difficult or dangerous.

matt ingalls

matt@...

Re: bus opcodes

2009-11-25T09:11:51Z

I never answered Victor's initial reply to this thread. Sorry for the late reply. I should have asked a more detailed question to avoid you guys having to rewrite what has been very well described here before about the bus opcodes.

What I was confused about in the latest API for 5.11 which is 2.20, I think, was how to pass the MYFLT **p param from the csoundGetChannelPtr method to the bus using chnget in the .csd. It's not so hard in c/c++ but I was trying with Java, then Lua.

PUBLIC int csoundGetChannelPtr(CSOUND *,
MYFLT **p, const char *name, int type);
-------------
chn_k "pitch", 1

instr 1
kval chnget "pitch"
printk2 kval
a1 oscil 3000, kval, 1
outs a1, a1
endin
----------------
With C, I can pass MYFLT by giving a value to the pointer.

MYFLT *pvalue;
double myhz = 600.00;

while(csound->PerformKsmps()==0)
{

if (csound->GetChannelPtr (*&pvalue, "pitch", CSOUND_INPUT_CHANNEL | CSOUND_CONTROL_CHANNEL) == 0 )
*pvalue = myhz;
}
--------
But with Java, SetValue disappeared for a while but is now back and working.

int result = csound.Compile(args.argc(), args.argv());
int myinputch = csndConstants.CSOUND_INPUT_CHANNEL;
int mycontch = csndConstants.CSOUND_CONTROL_CHANNEL;
SWIGTYPE_p_p_double myptr = myfltarray.GetPtr();

if (result == 0) {
while (csound.PerformKsmps() == 0) {
if (csnd.csoundGetChannelPtr(mycsound, myptr, "pitch", myinputch | mycontch) == 0) //int index, double value
{
//myfltarray.SetValue(0, myvalue);
csound.SetChannel("pitch" , myvalue); //or use SetChannel
}
------
I'm still trying to figure it out with Lua and having problems with MYFLT, but Mike has said he would add better bus support in the wrapper for Lua, so hopefully that will become clearer soon.

I guess to chime in on the rest of the thread, I'd be up for more streamlined bus opcode functionality too; but would hope the construction would translate well thru SWIG to the interface wrappers, since for some methods the path back to the original method from the library can be really narly.

Re: table methods API

2009-11-25T04:35:53Z

I think these should be OK to be called after compilation. But I need to check.

On 25 Nov 2009, at 03:46, James Hearon wrote:

Hi,
Working with table methods from API csoundTableLength, and csoundGetTable. It seems they need to be called while csound is performing to return the proper values, else I get -1.

I'm wondering what is best practice for doing that? I'm using a while statement to perform csound, for ex., but when I call one of the table methods from within the loop it really causes problems with the performance.

    int cs_performing = 1;
    while( cs_performing )
    {
        cs_performing = !csoundPerformKsmps(csound);
    }
    csoundCleanup(csound);
    return 1;

I came up with a kludge to just perform one control sample to return the table values, then continue on to the regular performance by resetting the cs_performing flag variable and continuing with the while loop above, but this is not very elegant.

    MYFLT *pvalue;
    if ( cs_performing = 1)
    {
    csoundPerformKsmps(csound);

    int csound_tablelength = csoundTableLength(csound, 1);
    printf("---RESULT of TABLE LENGTH: %d\n",csound_tablelength);

    int csound_gettable = csoundGetTable(csound, &pvalue, 2);
    printf("---RESULT of GET TABLE LENGTH: %d\n",csound_gettable);

    cs_performing = 2;
    }
    cs_performing = 1; //reset flag to "1"

I'm wondering how best to approach API methods which seem to only return proper values when csound is performing? The is within one thread. Do I need separate threads?

Regards,
Jim Hearon

Windows 7: It works the way you want. Learn more. ------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: UB binary of csound framework

2009-11-25T04:16:13Z

There is already MP3input opcode and GEN. One could construct an output
opcode, or use the piping facility that already exists to use lane or some
other external compressor.

> Add a new GEN, which can be a plugin too.
>
> Using in with soundfiles is not particularly useful, and while out is
> used for soundfiles,
> it is also handy to write files with opcodes like fout.
>
> Not too much of a limitation, IMHO. Certainly not a reason to ditch
> libsndfile.
>
> Victor
>
>
> On 25 Nov 2009, at 00:27, matt ingalls wrote:
>
>>
>> On Nov 23, 2009, at 1:14 PM, Victor Lazzarini wrote:
>>
>>> No, all you need is to add an opcode linked to an mp3 lib of your
>>> choice.
>>
>> you mean a csound external opcode?
>> how will this work for Gen01, in/out family opcodes, etc?
>>
>>
>> matt ingalls
>> matt@...
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
>> 30-Day
>> trial. Simplify your report design, integration and deployment - and
>> focus on
>> what you do best, core application coding. Discover what's new with
>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@...
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
> 30-Day
> trial. Simplify your report design, integration and deployment - and focus
> on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
>

Re: table methods API

2009-11-25T04:12:34Z

Note that this will always be true.

On 25 Nov 2009, at 03:46, James Hearon wrote:

if ( cs_performing = 1)

Re: UB binary of csound framework

2009-11-25T03:28:47Z

Add a new GEN, which can be a plugin too.

Using in with soundfiles is not particularly useful, and while out is
used for soundfiles,
it is also handy to write files with opcodes like fout.

Not too much of a limitation, IMHO. Certainly not a reason to ditch
libsndfile.

Victor

On 25 Nov 2009, at 00:27, matt ingalls wrote:

>
> On Nov 23, 2009, at 1:14 PM, Victor Lazzarini wrote:
>
>> No, all you need is to add an opcode linked to an mp3 lib of your
>> choice.
>
> you mean a csound external opcode?
> how will this work for Gen01, in/out family opcodes, etc?
>
>
> matt ingalls
> matt@...
>
>
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
> 30-Day
> trial. Simplify your report design, integration and deployment - and
> focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel

table methods API

2009-11-24T19:46:20Z

Hi,
Working with table methods from API csoundTableLength, and csoundGetTable. It seems they need to be called while csound is performing to return the proper values, else I get -1.

I'm wondering what is best practice for doing that? I'm using a while statement to perform csound, for ex., but when I call one of the table methods from within the loop it really causes problems with the performance.

    int cs_performing = 1;
    while( cs_performing )
    {
        cs_performing = !csoundPerformKsmps(csound);
    }
    csoundCleanup(csound);
    return 1;

I came up with a kludge to just perform one control sample to return the table values, then continue on to the regular performance by resetting the cs_performing flag variable and continuing with the while loop above, but this is not very elegant.

    MYFLT *pvalue;
    if ( cs_performing = 1)
    {
    csoundPerformKsmps(csound);

    int csound_tablelength = csoundTableLength(csound, 1);
    printf("---RESULT of TABLE LENGTH: %d\n",csound_tablelength);

    int csound_gettable = csoundGetTable(csound, &pvalue, 2);
    printf("---RESULT of GET TABLE LENGTH: %d\n",csound_gettable);

    cs_performing = 2;
    }
    cs_performing = 1; //reset flag to "1"

I'm wondering how best to approach API methods which seem to only return proper values when csound is performing? The is within one thread. Do I need separate threads?

Regards,
Jim Hearon

Windows 7: It works the way you want. Learn more.
------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: bus opcodes

2009-11-24T17:50:19Z

matt ingalls wrote:

>> Say you have a generic sampleplayer instrument, and at some point the
>> score creates two instances of this. One you want to connect to an
>> output-instrument, and the other through some effect-instrument. This is
>> not possible?
>>
>
> i'm in the middle of doing this kind of thing myself.
> the 2 solutions i came up with was:
>
>
> to either pass a unique ID as a p-field to the instrument
>
> instr 1
>
> iID = p7
>
> outvalue iID+1, kval1
> outvalue iID+2, kval2
> ...
> endin

This also needs to know how many channels an instrument needs, so that
the next ID doesn't overlap.. Or have a fixed MAX_CHANS=100 and pass
id*MAX_CHANS in the p-field.

> or use fractional instrument numbers and then use the fractional part as a unique ID
>
> ; host sends something like csoundInputMessage(myCsoundInstance, "i4.03 0 10");
> instr 4
>
> iID = 100*(p1-4)
> outvalue iID+1, kval1
> outvalue iID+2, kval2
> ...
> endin
>
>
> i like the fractional p1 a bit better because that is also a way to turnoff the instrument instance.
>
>
> either way the host app will need to keep track of all allocation/deallocation of instruments

Yep, this is how I did it in current AlgoScore, to be able to have instr
instances plot k-values graphically.
But I plan to pass an ID in a p-field in AlgoScore2 (when that project
happens), since I think the fractional thing is a bit too hackish. and
I'll have special algoscore-csound-instrument snippets that are
assembled into an orc file, and a custom opcode for communication with
AlgoScore (since outvalue/invalue doesn't support audio and I need
callback-approach for per-instance communication).

Would be better if csoundInputMessage() returned a unique ID (for
example, the pointervalue to the event) and that the CSOUND struct
contained a current_event field that callbacks could match against..

But this will probably never happen, maybe I should just write my own
DSP language/system based on JavaScript, LADSPA and Faust :)

/Jonatan

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: UB binary of csound framework

2009-11-24T16:27:55Z

On Nov 23, 2009, at 1:14 PM, Victor Lazzarini wrote:

> No, all you need is to add an opcode linked to an mp3 lib of your
> choice.

you mean a csound external opcode?
how will this work for Gen01, in/out family opcodes, etc?

matt ingalls
matt@...

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: bus opcodes

2009-11-24T16:23:30Z

Say you have a generic sampleplayer instrument, and at some point the
score creates two instances of this. One you want to connect to an
output-instrument, and the other through some effect-instrument. This is
not possible?

i'm in the middle of doing this kind of thing myself.

the 2 solutions i came up with was:

to either pass a unique ID as a p-field to the instrument

instr 1

iID = p7

outvalue iID+1, kval1

outvalue iID+2, kval2

...

endin

or use fractional instrument numbers and then use the fractional part as a unique ID

; host sends something like csoundInputMessage(myCsoundInstance, "i4.03 0 10");

instr 4

iID = 100*(p1-4)

outvalue iID+1, kval1

outvalue iID+2, kval2

...

endin

i like the fractional p1 a bit better because that is also a way to turnoff the instrument instance.

either way the host app will need to keep track of all allocation/deallocation of instruments

matt ingalls

matt@...

Re: bus opcodes

2009-11-24T16:05:31Z

and strings, please.

at least at one point you could put strings inside of the invalue/outvalue.

although it was kind of a hack and it might have been removed

(it doesn't seem to be working for me right now)

but the comment is still in the csound.h:

/**

* Called by external software to set a function for Csound to

* fetch input control values. The 'invalue' opcodes will

* directly call this function. If 'channelName' starts with a

* '$', then 'invalue' opcode is expecting a C string, to be copied

* to 'value', with maximum size csoundGetStrVarMaxLen().

PUBLIC void csoundSetInputValueCallback(CSOUND *,

void (*inputValueCalback_)(CSOUND *,

const char *channelName,

MYFLT *value));

/**

* Called by external software to set a function for Csound to

* send output control values. The 'outvalue' opcodes will

* directly call this function. If 'channelName' starts with a

* '$', then the 'outvalue' opcode is sending a string appended

* to channelName in the format: "$channelName$stringOutput".

* and 'value' will be the index number into 'channelName' where

* the stringOutput begins.

PUBLIC void csoundSetOutputValueCallback(CSOUND *,

void (*outputValueCalback_)(CSOUND *,

const char *channelName,

MYFLT value));

matt ingalls

matt@...

Re: bus opcodes

2009-11-24T14:57:46Z

and strings, please.

Am 24.11.2009 um 21:59 schrieb Jonatan Liljedahl:

> It would be nice if invalue/outvalue supported a-rate signals too.
>
> Victor Lazzarini wrote:
>> My opinion is that the two systems complement each other, one
>> callback
>> based, Matt's and the other polling (well kind of)-based, Istvan's.
>> What I meant to say below is that Csound manages the channel lists,
>> whereas with the callback system, it's the host that needs to hold
>> them.
>> Also the chn system does not depend on a host supporting it, you can
>> use channels to pass data between instruments as well, and it
>> supports
>> audio signals too.
>>
>> There are pros and cons of both systems. I think it's worth keeping
>> them all, IMHO, of course.
>>
>> Victor
>>
>> On 18 Nov 2009, at 19:39, matt ingalls wrote:
>>
>>> hi everyone
>>> ( recently unemployed and looking forward to get back into csound
>>> dev! )
>>>
>>>>> And because Csound does everything for you (creates channels,
>>>>> keeps
>>>>> track of bus names etc).
>>>
>>> if there is interest, i could easily add code to interface files
>>> like
>>> CppSound (or cs_glue -- what's the difference?) to handle the
>>> invalue/
>>> outvalue channel management.
>>>
>>> i personally am bothered by all the different channel bus methods --
>>> it is a pain for host apps to support all of them -- what about
>>> consolidating them internally at least? this way hosts only need to
>>> support one API and then users can use any opcodes they want. the
>>> duplicating opcodes could eventually be deprecated, or not (see
>>> below)
>>>
>>>
>>>> I do think that people keep using old opcodes instead of newer,
>>>> better
>>>> ones simply out of habit, so maybe we should move deprecated
>>>> opcodes
>>>> to an appendix in the reference manual.
>>> what about making a library of legacy opcodes that would be an
>>> option
>>> for users to install?
>>> then you can mark opcodes as deprecated and then 2 releases later or
>>> something like that move them to the legacy library.
>>> if a user wants to run an old orc they still can with one extra
>>> step,
>>> but everyone else gets to have a leaner and forward-looking csound.
>>>
>>>
>>> matt ingalls
>>> matt@...
>>>
>>>
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
>>> 30-Day
>>> trial. Simplify your report design, integration and deployment - and
>>> focus on
>>> what you do best, core application coding. Discover what's new with
>>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@...
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>
>>
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal Reports
>> 2008 30-Day
>> trial. Simplify your report design, integration and deployment -
>> and focus on
>> what you do best, core application coding. Discover what's new with
>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@...
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
> 30-Day
> trial. Simplify your report design, integration and deployment - and
> focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

Re: bus opcodes

2009-11-24T13:41:09Z

All it needs is a couple of new API functions to set callbacks for
audio signals.

Victor
On 24 Nov 2009, at 20:59, Jonatan Liljedahl wrote:

Re: bus opcodes

2009-11-24T13:13:10Z

> I'm currently looking for a way to dynamically connect csound instrument
> _instances_ to/from each other and to/from the host through the API to
> have a timeline-based graphical signal flow language. (AlgoScore2)
>
> This means I need to use a callback-based approach, and that the
> callback needs to identify which instr instance is sending/getting a
> value. This doesn't seem to be possible right now?
>
> One work-around would be to pass in an unique ID in a p-field and in the
> instr use something like this:
>
> chnsend "_id", p4
> k1 chnrecv "foo"
> a1 chnrecv "bar"
>
> The callback would then react on a value on channel "_id" by storing the
> ID in a variable in the Csound host data, so that in the other calls it
> will know which instr instance is asking for data.

I took a look at the source and it seems chnrecv/send is using the
software busses declared by chnexport/chn_* opcodes, which is global and
I want per-instrument declared busses. And since invalue/outvalue
doesn't support a-rate I will probably write yet-another-bus-opcode for
my host, and throw in the ID-stuff so it can be passed to the callback
for instance identification:

instr 1
host_id p4
ka host_in "amp"
kp host_in "pitch"
kw host_in "pwm"
a1 vco2 ka, kp, 4, kw
host_out "out", a1
endin

or can an opcode have per-instance private data?

/Jonatan

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: bus opcodes

2009-11-24T12:59:41Z

It would be nice if invalue/outvalue supported a-rate signals too.

Victor Lazzarini wrote:

> My opinion is that the two systems complement each other, one callback
> based, Matt's and the other polling (well kind of)-based, Istvan's.
> What I meant to say below is that Csound manages the channel lists,
> whereas with the callback system, it's the host that needs to hold them.
> Also the chn system does not depend on a host supporting it, you can
> use channels to pass data between instruments as well, and it supports
> audio signals too.
>
> There are pros and cons of both systems. I think it's worth keeping
> them all, IMHO, of course.
>
> Victor
>
> On 18 Nov 2009, at 19:39, matt ingalls wrote:
>
>> hi everyone
>> ( recently unemployed and looking forward to get back into csound
>> dev! )
>>
>>>> And because Csound does everything for you (creates channels, keeps
>>>> track of bus names etc).
>>
>> if there is interest, i could easily add code to interface files like
>> CppSound (or cs_glue -- what's the difference?) to handle the
>> invalue/
>> outvalue channel management.
>>
>> i personally am bothered by all the different channel bus methods --
>> it is a pain for host apps to support all of them -- what about
>> consolidating them internally at least? this way hosts only need to
>> support one API and then users can use any opcodes they want. the
>> duplicating opcodes could eventually be deprecated, or not (see below)
>>
>>
>>> I do think that people keep using old opcodes instead of newer,
>>> better
>>> ones simply out of habit, so maybe we should move deprecated opcodes
>>> to an appendix in the reference manual.
>> what about making a library of legacy opcodes that would be an option
>> for users to install?
>> then you can mark opcodes as deprecated and then 2 releases later or
>> something like that move them to the legacy library.
>> if a user wants to run an old orc they still can with one extra step,
>> but everyone else gets to have a leaner and forward-looking csound.
>>
>>
>> matt ingalls
>> matt@...
>>
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008
>> 30-Day
>> trial. Simplify your report design, integration and deployment - and
>> focus on
>> what you do best, core application coding. Discover what's new with
>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@...
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: Cell phone Csound

2009-11-24T11:30:49Z

Michael Gogins wrote:

> I note that for Android phones, there is a native development kit
> (NDK) enabling the use of C and C++ code on Android phones.

Another possibilty is the Palm Pre which seems a little more open
than the Android phones. The Palm Pre comes with libsndfile
version 1.0.17 pre-installed (as a shared library).

Erik
--
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: Cell phone Csound

2009-11-24T09:36:51Z

That's great news! I'll be very interested to see how well this
performs when it can be compiled!

On Tue, Nov 24, 2009 at 12:27 PM, Michael Gogins
<michael.gogins@...> wrote:

> I should add, the "bionic" libc that the NDK contains will
> automatically use the FPU, if it exists, if the code is built for
> "soft" float.
>
> I'd bet that whether or not MY phone has an FPU, or my wife's more
> recent Eris phone has an FPU, the next one I get in a year or so
> definitely will have an FPU and the NDK will support it.
>
> Regards,
> Mike
>
> On Tue, Nov 24, 2009 at 12:09 PM, Michael Gogins
> <michael.gogins@...> wrote:
>> I'm trying to figure this out for my own phone, the MyTouch 3G (aka
>> HTC Magic or Sapphire). The ARM1136 CPU that the phone uses has an
>> optional FPU. I'm trying to figure out if the phone contains the
>> optional FPU or not.
>>
>> Regards,
>> Mike
>>
>> On Tue, Nov 24, 2009 at 9:53 AM, chris kummerer <chris@...> wrote:
>>> re Michael,
>>>
>>> I read:
>>>> I note that for Android phones, there is a native development kit
>>>> (NDK) enabling the use of C and C++ code on Android phones.
>>>
>>> at last!
>>>
>>>> I will
>>>> investigate further. Because Csound has a Java API, it probably is
>>>> possible to produce a fairly complete Csound port for the Android
>>>> phones.
>>>
>>> please correct me if I'm wrong but so far all the actual android
>>> phones I found lack an FPU ...
>>>
>>> regards,
>>>
>>> x
>>>
>>> --
>>> chris@... Postmodernism is german romanticism with better
>>> http://pilot.fm/ special effects. (Jeff Keuss / via ctheory.net)
>>>
>>> ------------------------------------------------------------------------------
>>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
>>> trial. Simplify your report design, integration and deployment - and focus on
>>> what you do best, core application coding. Discover what's new with
>>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>>> _______________________________________________
>>> Csound-devel mailing list
>>> Csound-devel@...
>>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>>
>>
>>
>>
>> --
>> Michael Gogins
>> Irreducible Productions
>> http://www.michael-gogins.com
>> Michael dot Gogins at gmail dot com
>>
>
>
>
> --
> Michael Gogins
> Irreducible Productions
> http://www.michael-gogins.com
> Michael dot Gogins at gmail dot com
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

Re: Cell phone Csound

2009-11-24T09:27:19Z

I should add, the "bionic" libc that the NDK contains will
automatically use the FPU, if it exists, if the code is built for
"soft" float.

I'd bet that whether or not MY phone has an FPU, or my wife's more
recent Eris phone has an FPU, the next one I get in a year or so
definitely will have an FPU and the NDK will support it.

Regards,
Mike

On Tue, Nov 24, 2009 at 12:09 PM, Michael Gogins
<michael.gogins@...> wrote:

> I'm trying to figure this out for my own phone, the MyTouch 3G (aka
> HTC Magic or Sapphire). The ARM1136 CPU that the phone uses has an
> optional FPU. I'm trying to figure out if the phone contains the
> optional FPU or not.
>
> Regards,
> Mike
>
> On Tue, Nov 24, 2009 at 9:53 AM, chris kummerer <chris@...> wrote:
>> re Michael,
>>
>> I read:
>>> I note that for Android phones, there is a native development kit
>>> (NDK) enabling the use of C and C++ code on Android phones.
>>
>> at last!
>>
>>> I will
>>> investigate further. Because Csound has a Java API, it probably is
>>> possible to produce a fairly complete Csound port for the Android
>>> phones.
>>
>> please correct me if I'm wrong but so far all the actual android
>> phones I found lack an FPU ...
>>
>> regards,
>>
>> x
>>
>> --
>> chris@... Postmodernism is german romanticism with better
>> http://pilot.fm/ special effects. (Jeff Keuss / via ctheory.net)
>>
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
>> trial. Simplify your report design, integration and deployment - and focus on
>> what you do best, core application coding. Discover what's new with
>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@...
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>>
>
>
>
> --
> Michael Gogins
> Irreducible Productions
> http://www.michael-gogins.com
> Michael dot Gogins at gmail dot com
>

--
Michael Gogins
Irreducible Productions
http://www.michael-gogins.com
Michael dot Gogins at gmail dot com

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

Re: Cell phone Csound

2009-11-24T09:25:34Z

Very.

Mike

On Tue, Nov 24, 2009 at 12:04 PM, Felipe Sateler <fsateler@...> wrote:

> How hard would it be to use integer for internal processing? Very?
>
> On Tue, 2009-11-24 at 11:35 -0500, Steven Yi wrote:
>> Last I looked at the NDK, there were limitations on what libraries
>> were being included. It looks like the 1.6r1 NDK has included a few
>> more things:
>>
>> * libc (C library) headers
>> * libm (math library) headers
>> * JNI interface headers
>> * libz (Zlib compression) headers
>> * liblog (Android logging) header
>> * OpenGL ES 1.1 (3D graphics library) headers
>> * A Minimal set of headers for C++ support
>>
>> I imagine floating point will be slow but haven't seen any metrics on it.
>>
>> On Tue, Nov 24, 2009 at 9:53 AM, chris kummerer <chris@...> wrote:
>> > re Michael,
>> >
>> > I read:
>> >> I note that for Android phones, there is a native development kit
>> >> (NDK) enabling the use of C and C++ code on Android phones.
>> >
>> > at last!
>> >
>> >> I will
>> >> investigate further. Because Csound has a Java API, it probably is
>> >> possible to produce a fairly complete Csound port for the Android
>> >> phones.
>> >
>> > please correct me if I'm wrong but so far all the actual android
>> > phones I found lack an FPU ...
>> >
>> > regards,
>> >
>> > x
>> >
>> > --
>> > chris@... Postmodernism is german romanticism with better
>> > http://pilot.fm/ special effects. (Jeff Keuss / via ctheory.net)
>> >
>> > ------------------------------------------------------------------------------
>> > Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
>> > trial. Simplify your report design, integration and deployment - and focus on
>> > what you do best, core application coding. Discover what's new with
>> > Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> > _______________________________________________
>> > Csound-devel mailing list
>> > Csound-devel@...
>> > https://lists.sourceforge.net/lists/listinfo/csound-devel
>> >
>>
>> ------------------------------------------------------------------------------
>> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
>> trial. Simplify your report design, integration and deployment - and focus on
>> what you do best, core application coding. Discover what's new with
>> Crystal Reports now. http://p.sf.net/sfu/bobj-july
>> _______________________________________________
>> Csound-devel mailing list
>> Csound-devel@...
>> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>
> --
> Saludos,
> Felipe Sateler
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>
>

Re: Cell phone Csound

2009-11-24T09:09:23Z

I'm trying to figure this out for my own phone, the MyTouch 3G (aka
HTC Magic or Sapphire). The ARM1136 CPU that the phone uses has an
optional FPU. I'm trying to figure out if the phone contains the
optional FPU or not.

Regards,
Mike

On Tue, Nov 24, 2009 at 9:53 AM, chris kummerer <chris@...> wrote:

> re Michael,
>
> I read:
>> I note that for Android phones, there is a native development kit
>> (NDK) enabling the use of C and C++ code on Android phones.
>
> at last!
>
>> I will
>> investigate further. Because Csound has a Java API, it probably is
>> possible to produce a fairly complete Csound port for the Android
>> phones.
>
> please correct me if I'm wrong but so far all the actual android
> phones I found lack an FPU ...
>
> regards,
>
> x
>
> --
> chris@... Postmodernism is german romanticism with better
> http://pilot.fm/ special effects. (Jeff Keuss / via ctheory.net)
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel
>

Re: Cell phone Csound

2009-11-24T09:04:52Z

How hard would it be to use integer for internal processing? Very?

On Tue, 2009-11-24 at 11:35 -0500, Steven Yi wrote:

> Last I looked at the NDK, there were limitations on what libraries
> were being included. It looks like the 1.6r1 NDK has included a few
> more things:
>
> * libc (C library) headers
> * libm (math library) headers
> * JNI interface headers
> * libz (Zlib compression) headers
> * liblog (Android logging) header
> * OpenGL ES 1.1 (3D graphics library) headers
> * A Minimal set of headers for C++ support
>
> I imagine floating point will be slow but haven't seen any metrics on it.
>
> On Tue, Nov 24, 2009 at 9:53 AM, chris kummerer <chris@...> wrote:
> > re Michael,
> >
> > I read:
> >> I note that for Android phones, there is a native development kit
> >> (NDK) enabling the use of C and C++ code on Android phones.
> >
> > at last!
> >
> >> I will
> >> investigate further. Because Csound has a Java API, it probably is
> >> possible to produce a fairly complete Csound port for the Android
> >> phones.
> >
> > please correct me if I'm wrong but so far all the actual android
> > phones I found lack an FPU ...
> >
> > regards,
> >
> > x
> >
> > --
> > chris@... Postmodernism is german romanticism with better
> > http://pilot.fm/ special effects. (Jeff Keuss / via ctheory.net)
> >
> > ------------------------------------------------------------------------------
> > Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> > trial. Simplify your report design, integration and deployment - and focus on
> > what you do best, core application coding. Discover what's new with
> > Crystal Reports now. http://p.sf.net/sfu/bobj-july
> > _______________________________________________
> > Csound-devel mailing list
> > Csound-devel@...
> > https://lists.sourceforge.net/lists/listinfo/csound-devel
> >
>
> ------------------------------------------------------------------------------
> Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
> trial. Simplify your report design, integration and deployment - and focus on
> what you do best, core application coding. Discover what's new with
> Crystal Reports now. http://p.sf.net/sfu/bobj-july
> _______________________________________________
> Csound-devel mailing list
> Csound-devel@...
> https://lists.sourceforge.net/lists/listinfo/csound-devel

--
Saludos,
Felipe Sateler

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day
trial. Simplify your report design, integration and deployment - and focus on
what you do best, core application coding. Discover what's new with
Crystal Reports now. http://p.sf.net/sfu/bobj-july
_______________________________________________
Csound-devel mailing list
Csound-devel@...
https://lists.sourceforge.net/lists/listinfo/csound-devel

signature.asc (853 bytes) Download Attachment

Re: Cell phone Csound

2009-11-24T08:35:53Z

Last I looked at the NDK, there were limitations on what libraries
were being included. It looks like the 1.6r1 NDK has included a few
more things:

* libc (C library) headers
* libm (math library) headers
* JNI interface headers
* libz (Zlib compression) headers
* liblog (Android logging) header
* OpenGL ES 1.1 (3D graphics library) headers
* A Minimal set of headers for C++ support

I imagine floating point will be slow but haven't seen any metrics on it.

On Tue, Nov 24, 2009 at 9:53 AM, chris kummerer <chris@...> wrote: