headers
James McCartney | 1 Jun 2004 02:33
Picon

Re: Re: [sc-users] Post and flush


Flushing from the language is generally not a good thing to do because 
it calls the UI, and the language runs in real time.
All text is queued asynchronously. If the language were forced to wait 
for a flush during sequencing, it would ruin timing.
Flushing happens periodically from a timer thread.

On May 31, 2004, at 3:02 PM, James Harkins wrote:

> Let me bump this one, and switch it over to dev... there's 
> Meta_Post-flush, but it just calls "this.flushPostBuf."
>
> Post is only a subclass of Object. Is this right?
> hjh
>
> On Wednesday, May 19, 2004, at 09:49 US/Eastern, nikolai collinsky 
> wrote:
>
>> I'm trying to get posted feedback during intensive
>> algorithm churning, any recommendations on forcing
>> updates to display? At the moment nothing prints until
>> the calculations are over.
>>
>> Cheers
>>
>> CALL STACK:
>> 	Object-doesNotUnderstand
>> 		arg this = class Post
>> No implementations of 'flushPostBuf'.
>>
(Continue reading)

Permalink | Reply | 3 comments |
headers
ccos | 1 Jun 2004 02:50
Picon

melbourne colliders

we are having a meeting on the last saturday of June.
if you're in melbourne you should stop by.

http://swiki.hfbk-hamburg.de:8888/MusicTechnology/514

_c
Permalink | Reply | 0 comments |
headers
James Harkins | 1 Jun 2004 04:29
Picon
Favicon

Re: Re: [sc-users] Post and flush

Should Meta_Post-flush be removed then? It has only one possible 
outcome... "'flushPostBuf' not understood."
hjh

On Monday, May 31, 2004, at 20:33 US/Eastern, James McCartney wrote:

>
> Flushing from the language is generally not a good thing to do because 
> it calls the UI, and the language runs in real time.
> All text is queued asynchronously. If the language were forced to wait 
> for a flush during sequencing, it would ruin timing.
> Flushing happens periodically from a timer thread.
>
> On May 31, 2004, at 3:02 PM, James Harkins wrote:
>
>> Let me bump this one, and switch it over to dev... there's 
>> Meta_Post-flush, but it just calls "this.flushPostBuf."
>>
>> Post is only a subclass of Object. Is this right?
>> hjh
>>
>> On Wednesday, May 19, 2004, at 09:49 US/Eastern, nikolai collinsky 
>> wrote:
>>
>>> I'm trying to get posted feedback during intensive
>>> algorithm churning, any recommendations on forcing
>>> updates to display? At the moment nothing prints until
>>> the calculations are over.
>>>
>>> Cheers
(Continue reading)

Permalink | Reply | 1 comment |
headers
Sciss | 1 Jun 2004 06:31
Picon
Gravatar

Severe NRT DiskIn Trouble

hi all,

i spent the night debugging my NRT control files only to find out that 
there must be a severe bug in scsynth, namely the DiskIn operation. i 
broke it down the kind of the simplest test configuration:

SynthDef( "even-more-simpler", {
	Out.ar( bus: 0, channelsArray: DiskIn.ar( numChannels: 1, bufnum: 666 ));
}).writeDefFile;

f = File( "/tmp/ich-debug4.tmp", "w" );
c = [ 0.1, ["/d_load", 
"/Applications/SuperCollider_f/synthdefs/even-more-simpler.scsyndef"]].asRawOSC;
f.write( c.size );
f.write( c );
c = [ 0.2, ["/b_alloc", 666, 4096, 1, 0 ]].asRawOSC;
f.write( c.size );
f.write( c );
c = [ 0.3, ["/b_read", 666, 
"/Volumes/Edgard/audio/IchnoTest/DadaFlavourKrummeDauer2.aif", 0, 4096, 
0, 1 ]].asRawOSC;
f.write( c.size );
f.write( c );
c = [ 0.4, ["/s_new", "even-more-simpler", 1001, 1, 0 ]].asRawOSC;
f.write( c.size );
f.write( c );
c = [ 32.4, ["/n_free", 1001 ]].asRawOSC;
f.write( c.size );
f.write( c );
f.close;
(Continue reading)

Permalink | Reply | 2 comments |
headers
Joshua Parmenter | 1 Jun 2004 06:46

Re: Severe NRT DiskIn Trouble

DiskIn doesn't work in NRT.  PlayBuf works fine.

Josh

On May 31, 2004, at 9:31 PM, Sciss wrote:

> hi all,
>
> i spent the night debugging my NRT control files only to find out that  
> there must be a severe bug in scsynth, namely the DiskIn operation. i  
> broke it down the kind of the simplest test configuration:
>
>
> SynthDef( "even-more-simpler", {
> 	Out.ar( bus: 0, channelsArray: DiskIn.ar( numChannels: 1, bufnum: 666  
> ));
> }).writeDefFile;
>
>
> f = File( "/tmp/ich-debug4.tmp", "w" );
> c = [ 0.1, ["/d_load",  
> "/Applications/SuperCollider_f/synthdefs/even-more- 
> simpler.scsyndef"]].asRawOSC;
> f.write( c.size );
> f.write( c );
> c = [ 0.2, ["/b_alloc", 666, 4096, 1, 0 ]].asRawOSC;
> f.write( c.size );
> f.write( c );
> c = [ 0.3, ["/b_read", 666,  
> "/Volumes/Edgard/audio/IchnoTest/DadaFlavourKrummeDauer2.aif", 0,
(Continue reading)

Permalink | Reply | 0 comments |
headers
James McCartney | 1 Jun 2004 08:17
Picon

Re: Severe NRT DiskIn Trouble


DiskIn is written for asynchronous real time streaming only. It might 
not be hard to make it work for NRT, but I've not done that.
Permalink | Reply | 0 comments |
headers
James McCartney | 1 Jun 2004 08:55
Picon

Re: Re: [sc-users] Post and flush


yes.

On May 31, 2004, at 7:29 PM, James Harkins wrote:

> Should Meta_Post-flush be removed then? It has only one possible 
> outcome... "'flushPostBuf' not understood."
Permalink | Reply | 0 comments |
headers
Sciss | 1 Jun 2004 09:40
Picon
Gravatar

Re: Severe NRT DiskIn Trouble

p.s. it seems to be a problem of sc running without a "clock" beat. the 
problem disappears when the cmd file contains osc bundle a couple of 
times per second, i can go down with the diskbuf size again to say 32k 
whereas without intermitting osc messages the stuttering would only 
disappear for diskbufsizes of 4mb or so ;o...)   so well, at least seems 
to work now for me (though there are still some obscure noises sometimes 
in the output, but just rarely and of very low intensity).

ciao, -sciss-
Permalink | Reply | 0 comments |
headers
Thomas Charbonnel | 2 Jun 2004 09:34

Re: Getting the sample rate from the language

The redundancy point you made was bugging me too, so here's another
proposal, based on newton armstrong's method :

+ Server {

        getSampleRate { arg condition;
                var c, id, def, resp, rate;
                condition.test = false;
                c = Condition.new;
                resp = OSCresponderNode(addr, '/tr', { |t, r, msg|
                        r.remove;
                        rate = msg[3];
                        c.test = true;
                        c.signal;
                }).add;
                def = SynthDef("get-sample-rate", { |t_trig = 0|
                        SendTrig.kr(t_trig, 0, SampleRate.ir(0));
                });
                Routine.run {
                        id = this.nextNodeID;
                        this.sendMsgSync(c, "/d_recv", def.asBytes,
                                                        ["/s_new",
"get-sample-rate",
id, 1, 1]);
                        c.test = false;
                        this.sendMsg("/n_set", id, 't_trig', 1);
                        c.wait;
                        this.sendMsg("/n_free", id);
                        condition.test = true;
                        condition.signal;
(Continue reading)

Permalink | Reply | 0 comments |
headers
David Thall | 2 Jun 2004 21:26
Picon

slider & controlspec mapping question

Hello,
	I have a set of sliders with an exponential warp from 0.001 -> 10.  
I'd like to make the sliders go from 0 -> 10, but retain the 
exponential warp from 0.001 -> 10.  Is this possible?
	My current solution is to set a threshold (e.g., 0.001, or some value 
slightly smaller), and then set the output value to 0 if the output of 
the slider crosses this boundary.  The problem with this solution is I 
have to rescale my entire range to take into consideration a threshold 
slightly below my required range.
	Any suggestions / ideas / solutions would be greatly appreciated...

-David
Permalink | Reply | 1 comment |

Gmane