All of my instruments run in real-time but just! And that's on a quad 
core and 8 GB of RAM.

Well, imagine writing pvs data to tables and looping through the tables 
with conditionals, and on every FFT frame, using FFT size of 8192 and 
overlap of 8192/8.

Now multiply this by eight (I am working in 8 channels) and at 96k.

It runs, but would I use it in a live performance? Probably not. But 
mostly because this doesn't run on my laptop

Best,

Peiman



On 05/06/2011 21:10, Alex Weiss wrote:
> Hi,
>
> I was contemplating whether this belongs on the csound-dev list 
> instead, but I figured I'd try my luck here first...
>
> I'm looking for ways to benchmark various versions of my realtime IO 
> engine of my csound host, as well as csound's internal modules. 
> Specifically, I'm looking for instruments/opcodes that are known to be 
> problematic in a RT context and whose complexity can be scaled, so 
> that I can see how far I can go with the engines. Unfortunately, I 
> only have little realtime experience with csound, so I'm not sure 
> what's supposed to bring the system to its knees. So I was hoping I 
> could gather some input from users on this list; has anybody ever 
> tried to benchmark their system? What's usually most problematic 
> during realtime performances?
>
> Thanks,
>
> Alex



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"

Ah, pvs is a good one. Haven't thought of that, thanks!

Your results are pretty impressive, though... Makes me wonder if I should finally get a new computer.

Are you using portaudio for realtime output?

Alex

On Sun, Jun 5, 2011 at 10:18 PM, peiman khosravi <peimankhosravi@gmail.com> wrote:

All of my instruments run in real-time but just! And that's on a quad core and 8 GB of RAM.

Well, imagine writing pvs data to tables and looping through the tables with conditionals, and on every FFT frame, using FFT size of 8192 and overlap of 8192/8.

Now multiply this by eight (I am working in 8 channels) and at 96k.

It runs, but would I use it in a live performance? Probably not. But mostly because this doesn't run on my laptop

Best,

Peiman

On 05/06/2011 21:10, Alex Weiss wrote:

Hi,

I was contemplating whether this belongs on the csound-dev list instead, but I figured I'd try my luck here first...

I'm looking for ways to benchmark various versions of my realtime IO engine of my csound host, as well as csound's internal modules. Specifically, I'm looking for instruments/opcodes that are known to be problematic in a RT context and whose complexity can be scaled, so that I can see how far I can go with the engines. Unfortunately, I only have little realtime experience with csound, so I'm not sure what's supposed to bring the system to its knees. So I was hoping I could gather some input from users on this list; has anybody ever tried to benchmark their system? What's usually most problematic during realtime performances?

Thanks,

Alex

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"

Are you using float or double?

Best,

Alex

On Sun, Jun 5, 2011 at 10:53 PM, peiman khosravi <peimankhosravi@gmail.com> wrote:

Yes portaudio and large buffers (1024).  I have lots of maxmsp GUI running too, all at the same time (mostly writing to tables at k-rate).

I remember the day I thought to myself if I'm gonna get any deeper into FFT without compromising quality I need a new computer!

Best,

Peiman

On 05/06/2011 21:25, Alex Weiss wrote:

Ah, pvs is a good one. Haven't thought of that, thanks!

Your results are pretty impressive, though... Makes me wonder if I should finally get a new computer.

Are you using portaudio for realtime output?

Alex

On Sun, Jun 5, 2011 at 10:18 PM, peiman khosravi <peimankhosravi@gmail.com> wrote:

All of my instruments run in real-time but just! And that's on a quad core and 8 GB of RAM.

Well, imagine writing pvs data to tables and looping through the tables with conditionals, and on every FFT frame, using FFT size of 8192 and overlap of 8192/8.

Now multiply this by eight (I am working in 8 channels) and at 96k.

It runs, but would I use it in a live performance? Probably not. But mostly because this doesn't run on my laptop

Best,

Peiman

On 05/06/2011 21:10, Alex Weiss wrote:

Hi,

I was contemplating whether this belongs on the csound-dev list instead, but I figured I'd try my luck here first...

I'm looking for ways to benchmark various versions of my realtime IO engine of my csound host, as well as csound's internal modules. Specifically, I'm looking for instruments/opcodes that are known to be problematic in a RT context and whose complexity can be scaled, so that I can see how far I can go with the engines. Unfortunately, I only have little realtime experience with csound, so I'm not sure what's supposed to bring the system to its knees. So I was hoping I could gather some input from users on this list; has anybody ever tried to benchmark their system? What's usually most problematic during realtime performances?

Thanks,

Alex

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"

Are you using float or double?

Best,

Alex

On Sun, Jun 5, 2011 at 10:53 PM, peiman khosravi <peimankhosravi@gmail.com> wrote:

Yes portaudio and large buffers (1024).  I have lots of maxmsp GUI running too, all at the same time (mostly writing to tables at k-rate).

I remember the day I thought to myself if I'm gonna get any deeper into FFT without compromising quality I need a new computer!

Best,

Peiman

On 05/06/2011 21:25, Alex Weiss wrote:

Ah, pvs is a good one. Haven't thought of that, thanks!

Your results are pretty impressive, though... Makes me wonder if I should finally get a new computer.

Are you using portaudio for realtime output?

Alex

On Sun, Jun 5, 2011 at 10:18 PM, peiman khosravi <peimankhosravi@gmail.com> wrote:

All of my instruments run in real-time but just! And that's on a quad core and 8 GB of RAM.

Well, imagine writing pvs data to tables and looping through the tables with conditionals, and on every FFT frame, using FFT size of 8192 and overlap of 8192/8.

Now multiply this by eight (I am working in 8 channels) and at 96k.

It runs, but would I use it in a live performance? Probably not. But mostly because this doesn't run on my laptop

Best,

Peiman

On 05/06/2011 21:10, Alex Weiss wrote:

Hi,

I was contemplating whether this belongs on the csound-dev list instead, but I figured I'd try my luck here first...

I'm looking for ways to benchmark various versions of my realtime IO engine of my csound host, as well as csound's internal modules. Specifically, I'm looking for instruments/opcodes that are known to be problematic in a RT context and whose complexity can be scaled, so that I can see how far I can go with the engines. Unfortunately, I only have little realtime experience with csound, so I'm not sure what's supposed to bring the system to its knees. So I was hoping I could gather some input from users on this list; has anybody ever tried to benchmark their system? What's usually most problematic during realtime performances?

Thanks,

Alex

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"