Cycling 74 » Topic: Does Poly~ really consume less CPU ?

Does Poly~ really consume less CPU ?

Cheng Chien-Wen — Thu, 02 Mar 2006 17:55:42 +0000

Hi,
I built a FM symthesizer with poly~ carrying 16 instances, but when I was observing the CPU consumption rate in the DSP setup menu, it consumed about 20 percent of my CPU even if I did not send any message to trigger the instance. Another thing I found is that when I sent MIDI message to trigger the synthesizer built with poly~, it did not really influece the CPU consumption rate even if 16 instances are trigger at the same time.

This seems to mean that the poly~ does not really turn off the DSP processing within each instance even when no message is sent to the poly~ patch. Does anybody have the similar experiences ?

Thanks.

Chien-Wen

Re: Does Poly~ really consume less CPU ?

Isjtar — Thu, 02 Mar 2006 17:58:46 +0000

Quote: Cheng Chien-Wem wrote on Thu, 02 March 2006 10:55
—————————————————-

> This seems to mean that the poly~ does not really turn off the DSP processing within each instance even when no message is sent to the poly~ patch. Does anybody have the similar experiences ?
—————————————————-

i think you need to use the mute command to turn off dsp in one voice.

Re: Does Poly~ really consume less CPU ?

mzed — Thu, 02 Mar 2006 18:14:33 +0000

Quote: Cheng Chien-Wem wrote on Thu, 02 March 2006 09:55
—————————————————-
> This seems to mean that the poly~ does not really turn off the DSP processing within each instance even when no message is sent to the poly~ patch. Does anybody have the similar experiences ?
—————————————————-

Right – you have to do it yourself. Look at the thispoly~ help file.

Re: Does Poly~ really consume less CPU ?

Cheng Chien-Wen — Thu, 02 Mar 2006 20:51:34 +0000

I still feel confused. The help file of thispoly~ says I can use the sig~ object to turn on/off the busy state as an alternative to mute message. If they function the same, then when I use the volume output signal to control on/off of busy state, DSP on/off should also work, doesn’t it ?

My questions are:

1. Does “busy” state on/off also mean DSP on/off for each instance ?

2. I have seen people use “mute 1, 0″ and “mute 0, 1″ message to control “thispoly~”. What do “1, 0″, and “0, 1″ mean ?
The help file only have “mute 1″, “mute 0″ usage. What are the differences between the two-number command and one-number command ?

Thanks.

Re: Does Poly~ really consume less CPU ?

Jean-Francois Charles — Thu, 02 Mar 2006 21:07:19 +0000

Re: Does Poly~ really consume less CPU ?

Stefan Tiedje — Thu, 02 Mar 2006 22:17:26 +0000

Cheng Chien-Wem wrote:
> This seems to mean that the poly~ does not really turn off the DSP
> processing within each instance even when no message is sent to the
> poly~ patch. Does anybody have the similar experiences ?

You have to explicitly mute the instance if its not used. The midinote
messages are only used to route it to the correct instance.

Hope this helps…

Stefan

–

[][] [][][] [][] [][][]
[][][][][][][][][][][][][][][]

Stefan Tiedje
Klanggestalter
Electronic Composition
&
Improvisation

/~~~~~
\ /|() ()|
))))) )| | |( \
/// _/)/ )))))
___/ ///

————————-x—-
–_____———–|———–
–(_|_ —-|—–|—–()—-
– _|_)—-|—–()———–
———-()————x—–

14, Av. Pr. Franklin Roosevelt,
94320 Thiais, France
Phone at CCMIX +33-1-49 77 51 72

Re: Does Poly~ really consume less CPU ?

mzed — Thu, 02 Mar 2006 22:18:33 +0000

>
> > 2. I have seen people use “mute 1, 0″ and “mute 0, 1″ message to control
> > “thispoly~”. What do “1, 0″, and “0, 1″ mean ?
> > The help file only have “mute 1″, “mute 0″ usage. What are the differences
> > between the two-number command and one-number command ?
>
> mute 1 or mute 0 are used inside a poly-voice, in the subpatch.
> Use a 2-number command to speak to the poly~ box, one of the number telling
> which voice to address, or 0 to speak to all voices.
>

I’m sorry, but this is inaccurate.

In a message box, the comma separates the contents into separate messages. So, “mute 1, 0″ would send the message “mute 1″ and then the message “0″. Try connecting one to a print object to see how it works.

In the thispoly~ case, this is an easy way to control the mute state *and* the busy state in one go. So, “mute 1, 0″ would turn the mute state *on* (“mute 1″) and the busy state off (“0″). And of course “mute 0, 1″ does the opposite.

More clarification:
Busy refers to polyphony management. Busy == 0 means “I’m not busy, send me something.” Busy == 1 means “I’m busy, don’t send me anything to do.”

Mute refers to DSP. Mute == 0 means “I’m not muted, and am consuming CPU cycles.” Mute == 1 means “I am muted, and not consuming CPU.”

While you often set these messages at the same time, they aren’t doing the same thing. Poly~ is one of the trickier things in MaxMSP programming, so don’t be alarmed if it takes a bit of work to master. The payoff comes with more effecient patches.

good luck,

mzed

Re: Does Poly~ really consume less CPU ?

Roman Thilenius — Fri, 03 Mar 2006 01:12:59 +0000

> 2. I have seen people use “mute 1, 0″ and “mute 0, 1″ message to control “thispoly~”. What do “1, 0″, and “0, 1″ mean ?
> The help file only have “mute 1″, “mute 0″ usage. What are the differences between the two-number command and one-number command ?

one of these two numbers is the instance (“voice”)
if the patch which should be affected by the mute
message when i am not wrong.

Re: Does Poly~ really consume less CPU ?

Roman Thilenius — Fri, 03 Mar 2006 01:13:51 +0000

> 1. Does “busy” state on/off also mean DSP on/off for each instance ?

busy turns off midi where mute turns of audio

Re: Does Poly~ really consume less CPU ?

Roman Thilenius — Fri, 03 Mar 2006 01:15:48 +0000

> one of these two numbers is the instance (“voice”)
> if the patch which should be affected by the mute
> message when i am not wrong.

i am wrong.

Re: Does Poly~ really consume less CPU ?

Cheng Chien-Wen — Sat, 04 Mar 2006 20:38:07 +0000

Thank you very much for the help.