Hi,

There is also an issue with Portmidi not being reentrant which
sometime causes crashes after the first run. If you can use alsa MIDI
it will be a lot more stable.

Removing the scopes will also be a big improvement.

Cheers,
Andres

On Wed, Feb 23, 2011 at 2:24 AM, Aaron Krister Johnson
 wrote:
> Hi all,
>
> I was wondering if any one had a similar experience.
>
> I'm trying to play, with a MIDI keyboard, the 'pipe synth' example (and some
> other examples like the FM synth) in qutecsound.
>
> I'm finding that the output is *extremely* choppy, not matter what buffer
> and ksmps setting I try, and furthermore, qutecsound crashes everytime I hit
> the stop button. Ditto jack vs. alsa. (I'm on Arch Linux, BTW)
>
> Is this just really buggy, or am I missing something in my settings. Without
> the QT gui, in FLTK GUIs, etc, Csound normally seems reasonably dependable
> for real-time work, but I've never had as poor a response as with
> qutecsound.
>
> Any help would be appreciated; I think the interface looks promising, but I
> need it to be dependable and responsive!
>
> AKJ
>
> --
> Aaron Krister Johnson
> http://www.akjmusic.com
> http://www.untwelve.org
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hi Aaron,

Csound doesn't use alsaseq, so it shouldn't interfere at all with
aconnect. It's not working probably because you need to load the
module snd-vir-midi.

And your report made me find something very silly I was doing in the
Csound thread, which causes the choppyness you describe, even on my
system. Just remove line 311 which reads:
qApp->processEvents();

from qcsperfthread.cpp. Please let me know if this improves performance for you.

Cheers,
Andres

On Wed, Feb 23, 2011 at 4:34 PM, Aaron Krister Johnson
 wrote:
> Andres,
>
> Thanks for the quick reply.
>
> Indeed, the use of Alsa MIDI instead of Portmidi was more stable, and
> avoided the crash I reported after hitting the 'Stop' button. However, I had
> the strange issue of qutecsound grabbing and managing the connections,
> effectively blocking jackctl.py (which uses aconnect) from setting up midi
> connections. In addition, my USB midi keyboard was seen as an alsa device
> (hw:2,0,0), but csound/qutecsound reported an error trying to use the device
> when it was listed as such in the configure dialog. Strange, esp. when
> Portmidi runs the same USB keyboard directly, no problem. maybe it was a
> 'grab' issue like the other; I'll try again when I get the chance.
>
> Taking away the oscilloscope did in fact result in better performance,
> although it was still quite choppy. Granted, I am doing this all on a
> non-interrupt optimial EEEPC netbook, but considering that other Linux
> softsynths with GUIs run flawlessy, I can only come to the unfortunate
> conclusion that qutecsound/csound, at least with certain complex orchestras
> (in particular, physical models) are not production stable/ready :( I wonder
> if there are other settings I might try---running csound in a separate
> terminal thread instead of using the API?
>
> Best,
> AKJ
>
> On Wed, Feb 23, 2011 at 1:57 AM, Andres Cabrera 
> wrote:
>>
>> Hi,
>>
>> There is also an issue with Portmidi not being reentrant which
>> sometime causes crashes after the first run. If you can use alsa MIDI
>> it will be a lot more stable.
>>
>> Removing the scopes will also be a big improvement.
>>
>> Cheers,
>> Andres
>>
>> On Wed, Feb 23, 2011 at 2:24 AM, Aaron Krister Johnson
>>  wrote:
>> > Hi all,
>> >
>> > I was wondering if any one had a similar experience.
>> >
>> > I'm trying to play, with a MIDI keyboard, the 'pipe synth' example (and
>> > some
>> > other examples like the FM synth) in qutecsound.
>> >
>> > I'm finding that the output is *extremely* choppy, not matter what
>> > buffer
>> > and ksmps setting I try, and furthermore, qutecsound crashes everytime I
>> > hit
>> > the stop button. Ditto jack vs. alsa. (I'm on Arch Linux, BTW)
>> >
>> > Is this just really buggy, or am I missing something in my settings.
>> > Without
>> > the QT gui, in FLTK GUIs, etc, Csound normally seems reasonably
>> > dependable
>> > for real-time work, but I've never had as poor a response as with
>> > qutecsound.
>> >
>> > Any help would be appreciated; I think the interface looks promising,
>> > but I
>> > need it to be dependable and responsive!
>> >
>> > AKJ
>> >
>> > --
>> > Aaron Krister Johnson
>> > http://www.akjmusic.com
>> > http://www.untwelve.org
>> >
>> >
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>
>
>
> --
> Aaron Krister Johnson
> http://www.akjmusic.com
> http://www.untwelve.org
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hello, 

I do not know very well C + +, but I want to know it better ...

 And generally interesting to know "how it all works. "

 I looked a bit at the source code (but I do not have anything from the
development tools ...), and of course, did not get good results ... :)

___________________
qApp->processEvents(); 
___________________


 If you can comment on this line and why it is so important (theoretically?)
for fast response, it would be cool.

Best regards,

Alexander Tasev, Russia, SPb

--
View this message in context: http://csound.1045644.n5.nabble.com/poor-performance-from-qutecsound-0-6-1-tp3396482p3407723.html
Sent from the Csound - General mailing list archive at Nabble.com.


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hi,

This line tells the qt toolkit to process any events (mouse, keyboard,
etc.), and do anything it needs to do like redrawing, updating, etc.
The problem is that it was being called from within the csound thread,
which caused performance to degrade. The line is actually only needed
if Csound is not running in a separate thread. There was a significant
improvement in performance on my system, but it seems it's
unfortunately not enough for Aaron. I'm looking very seriously into
this, as it is a very important matter.

Cheers,
Andres

On Thu, Mar 3, 2011 at 10:08 AM, Alesandro  wrote:
> Hello,
>
> I do not know very well C + +, but I want to know it better ...
>
>  And generally interesting to know "how it all works. "
>
>  I looked a bit at the source code (but I do not have anything from the
> development tools ...), and of course, did not get good results ... :)
>
> ___________________
> qApp->processEvents();
> ___________________
>
>
>  If you can comment on this line and why it is so important (theoretically?)
> for fast response, it would be cool.
>
> Best regards,
>
> Alexander Tasev, Russia, SPb
>
> --
> View this message in context: http://csound.1045644.n5.nabble.com/poor-performance-from-qutecsound-0-6-1-tp3396482p3407723.html
> Sent from the Csound - General mailing list archive at Nabble.com.
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Hi Aaron,

I've now made some changes in svn, which allow you to tune the
performance of QuteCsound by disabling the record buffer, the printing
of messages and the evaluation of realtime events. This improves
rendering of Trapped by 20%, leaving the time now much closer (only
about 10% or less above running it on the command line). Can you fetch
svn and try setting these performance tweaks in the preferences to see
if performance improves for you?

Cheers,
Andres

On Fri, Feb 25, 2011 at 9:26 PM, Aaron Krister Johnson
 wrote:
> Hi Andres,
>
> Sorry to report that after making the correction tot eh source code, I'm
> still get rather poor, choppy performance from qutecsound....hope this can
> be fixed at some point :)
>
> AKJ
>
> On Wed, Feb 23, 2011 at 12:39 PM, Andres Cabrera 
> wrote:
>>
>> Hi Aaron,
>>
>> Csound doesn't use alsaseq, so it shouldn't interfere at all with
>> aconnect. It's not working probably because you need to load the
>> module snd-vir-midi.
>>
>> And your report made me find something very silly I was doing in the
>> Csound thread, which causes the choppyness you describe, even on my
>> system. Just remove line 311 which reads:
>> qApp->processEvents();
>>
>> from qcsperfthread.cpp. Please let me know if this improves performance
>> for you.
>>
>> Cheers,
>> Andres
>>
>> On Wed, Feb 23, 2011 at 4:34 PM, Aaron Krister Johnson
>>  wrote:
>> > Andres,
>> >
>> > Thanks for the quick reply.
>> >
>> > Indeed, the use of Alsa MIDI instead of Portmidi was more stable, and
>> > avoided the crash I reported after hitting the 'Stop' button. However, I
>> > had
>> > the strange issue of qutecsound grabbing and managing the connections,
>> > effectively blocking jackctl.py (which uses aconnect) from setting up
>> > midi
>> > connections. In addition, my USB midi keyboard was seen as an alsa
>> > device
>> > (hw:2,0,0), but csound/qutecsound reported an error trying to use the
>> > device
>> > when it was listed as such in the configure dialog. Strange, esp. when
>> > Portmidi runs the same USB keyboard directly, no problem. maybe it was a
>> > 'grab' issue like the other; I'll try again when I get the chance.
>> >
>> > Taking away the oscilloscope did in fact result in better performance,
>> > although it was still quite choppy. Granted, I am doing this all on a
>> > non-interrupt optimial EEEPC netbook, but considering that other Linux
>> > softsynths with GUIs run flawlessy, I can only come to the unfortunate
>> > conclusion that qutecsound/csound, at least with certain complex
>> > orchestras
>> > (in particular, physical models) are not production stable/ready :( I
>> > wonder
>> > if there are other settings I might try---running csound in a separate
>> > terminal thread instead of using the API?
>> >
>> > Best,
>> > AKJ
>> >
>> > On Wed, Feb 23, 2011 at 1:57 AM, Andres Cabrera 
>> > wrote:
>> >>
>> >> Hi,
>> >>
>> >> There is also an issue with Portmidi not being reentrant which
>> >> sometime causes crashes after the first run. If you can use alsa MIDI
>> >> it will be a lot more stable.
>> >>
>> >> Removing the scopes will also be a big improvement.
>> >>
>> >> Cheers,
>> >> Andres
>> >>
>> >> On Wed, Feb 23, 2011 at 2:24 AM, Aaron Krister Johnson
>> >>  wrote:
>> >> > Hi all,
>> >> >
>> >> > I was wondering if any one had a similar experience.
>> >> >
>> >> > I'm trying to play, with a MIDI keyboard, the 'pipe synth' example
>> >> > (and
>> >> > some
>> >> > other examples like the FM synth) in qutecsound.
>> >> >
>> >> > I'm finding that the output is *extremely* choppy, not matter what
>> >> > buffer
>> >> > and ksmps setting I try, and furthermore, qutecsound crashes
>> >> > everytime I
>> >> > hit
>> >> > the stop button. Ditto jack vs. alsa. (I'm on Arch Linux, BTW)
>> >> >
>> >> > Is this just really buggy, or am I missing something in my settings.
>> >> > Without
>> >> > the QT gui, in FLTK GUIs, etc, Csound normally seems reasonably
>> >> > dependable
>> >> > for real-time work, but I've never had as poor a response as with
>> >> > qutecsound.
>> >> >
>> >> > Any help would be appreciated; I think the interface looks promising,
>> >> > but I
>> >> > need it to be dependable and responsive!
>> >> >
>> >> > AKJ
>> >> >
>> >> > --
>> >> > Aaron Krister Johnson
>> >> > http://www.akjmusic.com
>> >> > http://www.untwelve.org
>> >> >
>> >> >
>> >>
>> >>
>> >> Send bugs reports to the Sourceforge bug tracker
>> >>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> >> Discussions of bugs and features can be posted here
>> >> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>> >> "unsubscribe
>> >> csound"
>> >>
>> >
>> >
>> >
>> > --
>> > Aaron Krister Johnson
>> > http://www.akjmusic.com
>> > http://www.untwelve.org
>> >
>> >
>>
>>
>> Send bugs reports to the Sourceforge bug tracker
>>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
>> Discussions of bugs and features can be posted here
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>> csound"
>>
>
>
>
> --
> Aaron Krister Johnson
> http://www.akjmusic.com
> http://www.untwelve.org
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

That'll be a great feature to have Andres. Nice one.

On 11 March 2011 17:31, Andres Cabrera  wrote:
> Hi Aaron,
>
> I've now made some changes in svn, which allow you to tune the
> performance of QuteCsound by disabling the record buffer, the printing
> of messages and the evaluation of realtime events. This improves
> rendering of Trapped by 20%, leaving the time now much closer (only
> about 10% or less above running it on the command line). Can you fetch
> svn and try setting these performance tweaks in the preferences to see
> if performance improves for you?
>
> Cheers,
> Andres
>
> On Fri, Feb 25, 2011 at 9:26 PM, Aaron Krister Johnson
>  wrote:
>> Hi Andres,
>>
>> Sorry to report that after making the correction tot eh source code, I'm
>> still get rather poor, choppy performance from qutecsound....hope this can
>> be fixed at some point :)
>>
>> AKJ
>>
>> On Wed, Feb 23, 2011 at 12:39 PM, Andres Cabrera 
>> wrote:
>>>
>>> Hi Aaron,
>>>
>>> Csound doesn't use alsaseq, so it shouldn't interfere at all with
>>> aconnect. It's not working probably because you need to load the
>>> module snd-vir-midi.
>>>
>>> And your report made me find something very silly I was doing in the
>>> Csound thread, which causes the choppyness you describe, even on my
>>> system. Just remove line 311 which reads:
>>> qApp->processEvents();
>>>
>>> from qcsperfthread.cpp. Please let me know if this improves performance
>>> for you.
>>>
>>> Cheers,
>>> Andres
>>>
>>> On Wed, Feb 23, 2011 at 4:34 PM, Aaron Krister Johnson
>>>  wrote:
>>> > Andres,
>>> >
>>> > Thanks for the quick reply.
>>> >
>>> > Indeed, the use of Alsa MIDI instead of Portmidi was more stable, and
>>> > avoided the crash I reported after hitting the 'Stop' button. However, I
>>> > had
>>> > the strange issue of qutecsound grabbing and managing the connections,
>>> > effectively blocking jackctl.py (which uses aconnect) from setting up
>>> > midi
>>> > connections. In addition, my USB midi keyboard was seen as an alsa
>>> > device
>>> > (hw:2,0,0), but csound/qutecsound reported an error trying to use the
>>> > device
>>> > when it was listed as such in the configure dialog. Strange, esp. when
>>> > Portmidi runs the same USB keyboard directly, no problem. maybe it was a
>>> > 'grab' issue like the other; I'll try again when I get the chance.
>>> >
>>> > Taking away the oscilloscope did in fact result in better performance,
>>> > although it was still quite choppy. Granted, I am doing this all on a
>>> > non-interrupt optimial EEEPC netbook, but considering that other Linux
>>> > softsynths with GUIs run flawlessy, I can only come to the unfortunate
>>> > conclusion that qutecsound/csound, at least with certain complex
>>> > orchestras
>>> > (in particular, physical models) are not production stable/ready :( I
>>> > wonder
>>> > if there are other settings I might try---running csound in a separate
>>> > terminal thread instead of using the API?
>>> >
>>> > Best,
>>> > AKJ
>>> >
>>> > On Wed, Feb 23, 2011 at 1:57 AM, Andres Cabrera 
>>> > wrote:
>>> >>
>>> >> Hi,
>>> >>
>>> >> There is also an issue with Portmidi not being reentrant which
>>> >> sometime causes crashes after the first run. If you can use alsa MIDI
>>> >> it will be a lot more stable.
>>> >>
>>> >> Removing the scopes will also be a big improvement.
>>> >>
>>> >> Cheers,
>>> >> Andres
>>> >>
>>> >> On Wed, Feb 23, 2011 at 2:24 AM, Aaron Krister Johnson
>>> >>  wrote:
>>> >> > Hi all,
>>> >> >
>>> >> > I was wondering if any one had a similar experience.
>>> >> >
>>> >> > I'm trying to play, with a MIDI keyboard, the 'pipe synth' example
>>> >> > (and
>>> >> > some
>>> >> > other examples like the FM synth) in qutecsound.
>>> >> >
>>> >> > I'm finding that the output is *extremely* choppy, not matter what
>>> >> > buffer
>>> >> > and ksmps setting I try, and furthermore, qutecsound crashes
>>> >> > everytime I
>>> >> > hit
>>> >> > the stop button. Ditto jack vs. alsa. (I'm on Arch Linux, BTW)
>>> >> >
>>> >> > Is this just really buggy, or am I missing something in my settings.
>>> >> > Without
>>> >> > the QT gui, in FLTK GUIs, etc, Csound normally seems reasonably
>>> >> > dependable
>>> >> > for real-time work, but I've never had as poor a response as with
>>> >> > qutecsound.
>>> >> >
>>> >> > Any help would be appreciated; I think the interface looks promising,
>>> >> > but I
>>> >> > need it to be dependable and responsive!
>>> >> >
>>> >> > AKJ
>>> >> >
>>> >> > --
>>> >> > Aaron Krister Johnson
>>> >> > http://www.akjmusic.com
>>> >> > http://www.untwelve.org
>>> >> >
>>> >> >
>>> >>
>>> >>
>>> >> Send bugs reports to the Sourceforge bug tracker
>>> >>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> >> Discussions of bugs and features can be posted here
>>> >> To unsubscribe, send email sympa@lists.bath.ac.uk with body
>>> >> "unsubscribe
>>> >> csound"
>>> >>
>>> >
>>> >
>>> >
>>> > --
>>> > Aaron Krister Johnson
>>> > http://www.akjmusic.com
>>> > http://www.untwelve.org
>>> >
>>> >
>>>
>>>
>>> Send bugs reports to the Sourceforge bug tracker
>>>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
>>> Discussions of bugs and features can be posted here
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe
>>> csound"
>>>
>>
>>
>>
>> --
>> Aaron Krister Johnson
>> http://www.akjmusic.com
>> http://www.untwelve.org
>>
>>
>
>
> Send bugs reports to the Sourceforge bug tracker
>            https://sourceforge.net/tracker/?group_id=81968&atid=564599
> Discussions of bugs and features can be posted here
> To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"
>
>


Send bugs reports to the Sourceforge bug tracker
            https://sourceforge.net/tracker/?group_id=81968&atid=564599
Discussions of bugs and features can be posted here
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"