Dear List,
I came accross this "Stompbox design workshop" ( https://ccrma.stanford.edu/wiki/Stompbox_2011). They use Pd on a beagleboard to make stompboxes. They talk about the older version of the beagleboard with a CPU running at around 700 MHz. This is the same speed as the RPi's.
I'm wondering what kind of latency they achieved on the beagleboard. Leaving aside the problem of the lack of audio in, don't you think we should be able to get the same performance on a raspberry pi ? Also, I wonder why JACK runs on the beagleboard and not on the RPi.
Cheers,
Pierre.
I could be wrong, but I believe the Beagleboard is ARMv7 and the Pi is ARM v6 - so BB could easily have twice the floating-point erformance of Pi. I'm not sure why and whether there would be any difference in Jack for the two. I believe (but am not sure at all) that BB doesn't have the same USB problems the Pi has. All in all it's still a very attractive possibliity, perticularly since the Stanford people have tested and exercised it thoroughly.
cheers Miller
On Mon, Dec 17, 2012 at 04:38:40PM +0100, Pierre Massat wrote:
Dear List,
I came accross this "Stompbox design workshop" ( https://ccrma.stanford.edu/wiki/Stompbox_2011). They use Pd on a beagleboard to make stompboxes. They talk about the older version of the beagleboard with a CPU running at around 700 MHz. This is the same speed as the RPi's.
I'm wondering what kind of latency they achieved on the beagleboard. Leaving aside the problem of the lack of audio in, don't you think we should be able to get the same performance on a raspberry pi ? Also, I wonder why JACK runs on the beagleboard and not on the RPi.
Cheers,
Pierre.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Hi Miller,
Thank you for your reply. I didn't know there could be such a difference in performance between two ARM chips. It seems very attractive indeed, and i think i'll contact the people at CCRMA to ask about the latency they achieved. If it's low enough i'll definitely give it a try.
Cheers,
Pierre.
2012/12/17 Miller Puckette msp@ucsd.edu
I could be wrong, but I believe the Beagleboard is ARMv7 and the Pi is ARM v6 - so BB could easily have twice the floating-point erformance of Pi. I'm not sure why and whether there would be any difference in Jack for the two. I believe (but am not sure at all) that BB doesn't have the same USB problems the Pi has. All in all it's still a very attractive possibliity, perticularly since the Stanford people have tested and exercised it thoroughly.
cheers Miller
On Mon, Dec 17, 2012 at 04:38:40PM +0100, Pierre Massat wrote:
Dear List,
I came accross this "Stompbox design workshop" ( https://ccrma.stanford.edu/wiki/Stompbox_2011). They use Pd on a beagleboard to make stompboxes. They talk about the older version of the beagleboard with a CPU running
at
around 700 MHz. This is the same speed as the RPi's.
I'm wondering what kind of latency they achieved on the beagleboard. Leaving aside the problem of the lack of audio in, don't you think we should be able to get the same performance on a raspberry pi ? Also, I wonder why JACK runs on the beagleboard and not on the RPi.
Cheers,
Pierre.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management ->
On Mon, Dec 17, 2012 at 7:35 PM, Miller Puckette msp@ucsd.edu wrote:
I could be wrong, but I believe the Beagleboard is ARMv7 and the Pi is ARM v6 - so BB could easily have twice the floating-point erformance of Pi. I'm not sure why and whether there would be any difference in Jack for the two. I believe (but am not sure at all) that BB doesn't have the same USB problems the Pi has. All in all it's still a very attractive possibliity, perticularly since the Stanford people have tested and exercised it thoroughly.
Yes, BB has a _LOT_ better performance than RasPi ..
Btw, if you want a cheaper board with BB like performance but RasPi like price, look at Cubieboard .. ( http://cubieboard.org/ )
-- buZz
Hi,
I wrote an e-mail to the guys who made the Satellite CCRMA image for the BB, asking about latency and quality of the built-in audio interface. I didn't get a reply to any of these questions, so I guess i'll have to buy one and figure out myself.
One interesting news though is that they are preparing an image for the Raspberry Pi. It should be the first image for this machine optimized for audio, beside the Raspbian image recently tweaked by Miller Puckette.
And finally, just so you know : "I haven't published about it, but I can safely say that if you don't use the satellite CCRMA images, you will be sorry later on down the road. As far as I know, no one else is bothering to make good images." I'm not sure what to make of these statements ;)
Cheers,
Pierre.
2012/12/20 Bastiaan van den Berg buzz@spacedout.nl
On Mon, Dec 17, 2012 at 7:35 PM, Miller Puckette msp@ucsd.edu wrote:
I could be wrong, but I believe the Beagleboard is ARMv7 and the Pi is ARM v6 - so BB could easily have twice the floating-point erformance of Pi. I'm not sure why and whether there would be any difference in Jack for the two. I believe (but am not sure at all) that BB doesn't have the same USB problems the Pi has. All in all it's still a very attractive possibliity, perticularly since the Stanford people have tested and exercised it thoroughly.
Yes, BB has a _LOT_ better performance than RasPi ..
Btw, if you want a cheaper board with BB like performance but RasPi like price, look at Cubieboard .. ( http://cubieboard.org/ )
-- buZz
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
I just tried out the latency.pd patch in 7.stuff/tools on my beagleboard and [timer] shifts between 72.5 and 75.4, if this answers your question. I've attended a workshop on the BB with Pd by the buys at CCRMA, it was a rather nice one.
On Thu, Dec 20, 2012 at 4:54 PM, Pierre Massat pimassat@gmail.com wrote:
Hi,
I wrote an e-mail to the guys who made the Satellite CCRMA image for the BB, asking about latency and quality of the built-in audio interface. I didn't get a reply to any of these questions, so I guess i'll have to buy one and figure out myself.
One interesting news though is that they are preparing an image for the Raspberry Pi. It should be the first image for this machine optimized for audio, beside the Raspbian image recently tweaked by Miller Puckette.
And finally, just so you know : "I haven't published about it, but I can safely say that if you don't use the satellite CCRMA images, you will be sorry later on down the road. As far as I know, no one else is bothering to make good images." I'm not sure what to make of these statements ;)
Cheers,
Pierre.
2012/12/20 Bastiaan van den Berg buzz@spacedout.nl
On Mon, Dec 17, 2012 at 7:35 PM, Miller Puckette msp@ucsd.edu wrote:
I could be wrong, but I believe the Beagleboard is ARMv7 and the Pi is ARM v6 - so BB could easily have twice the floating-point erformance of Pi. I'm not sure why and whether there would be any difference in Jack for the two. I believe (but am not sure at all) that BB doesn't have the same USB problems the Pi has. All in all it's still a very attractive possibliity, perticularly since the Stanford people have tested and exercised it thoroughly.
Yes, BB has a _LOT_ better performance than RasPi ..
Btw, if you want a cheaper board with BB like performance but RasPi like price, look at Cubieboard .. ( http://cubieboard.org/ )
-- buZz
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Wow! Are you using their satellite distro ? What version of the beagleboard are you using ? 72 ms is definitely unusable for realtime effects. I can't seem to find any info about the latency they achieve and this is a bit strange. A low latency is a sine qua non condition if one wants to build a stompbox.
Cheers,
Pierre.
2012/12/20 Alexandros Drymonitis adrcki@gmail.com
I just tried out the latency.pd patch in 7.stuff/tools on my beagleboard and [timer] shifts between 72.5 and 75.4, if this answers your question. I've attended a workshop on the BB with Pd by the buys at CCRMA, it was a rather nice one.
On Thu, Dec 20, 2012 at 4:54 PM, Pierre Massat pimassat@gmail.com wrote:
Hi,
I wrote an e-mail to the guys who made the Satellite CCRMA image for the BB, asking about latency and quality of the built-in audio interface. I didn't get a reply to any of these questions, so I guess i'll have to buy one and figure out myself.
One interesting news though is that they are preparing an image for the Raspberry Pi. It should be the first image for this machine optimized for audio, beside the Raspbian image recently tweaked by Miller Puckette.
And finally, just so you know : "I haven't published about it, but I can safely say that if you don't use the satellite CCRMA images, you will be sorry later on down the road. As far as I know, no one else is bothering to make good images." I'm not sure what to make of these statements ;)
Cheers,
Pierre.
2012/12/20 Bastiaan van den Berg buzz@spacedout.nl
On Mon, Dec 17, 2012 at 7:35 PM, Miller Puckette msp@ucsd.edu wrote:
I could be wrong, but I believe the Beagleboard is ARMv7 and the Pi is ARM v6 - so BB could easily have twice the floating-point erformance of Pi. I'm not sure why and whether there would be any difference in Jack for the two. I believe (but am not sure at all) that BB doesn't have the same USB problems the Pi has. All in all it's still a very attractive possibliity, perticularly since the Stanford people have tested and exercised it thoroughly.
Yes, BB has a _LOT_ better performance than RasPi ..
Btw, if you want a cheaper board with BB like performance but RasPi like price, look at Cubieboard .. ( http://cubieboard.org/ )
-- buZz
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Yup, I'm using their distro as I got it from them at the workshop. Not sure which version it is, the workshop took place last June, so it shouldn't be old... Building a stombox out of a machine as powerful as a smart phone...dunno. Edgar Berahl himself said that the BB is as powerful as an iPhone 4, or something. How much could it take?...And we did have some CPU issues during the workshop, especially with [expr~]..
On Thu, Dec 20, 2012 at 5:15 PM, Pierre Massat pimassat@gmail.com wrote:
Wow! Are you using their satellite distro ? What version of the beagleboard are you using ? 72 ms is definitely unusable for realtime effects. I can't seem to find any info about the latency they achieve and this is a bit strange. A low latency is a sine qua non condition if one wants to build a stompbox.
Cheers,
Pierre.
2012/12/20 Alexandros Drymonitis adrcki@gmail.com
I just tried out the latency.pd patch in 7.stuff/tools on my beagleboard and [timer] shifts between 72.5 and 75.4, if this answers your question. I've attended a workshop on the BB with Pd by the buys at CCRMA, it was a rather nice one.
On Thu, Dec 20, 2012 at 4:54 PM, Pierre Massat pimassat@gmail.comwrote:
Hi,
I wrote an e-mail to the guys who made the Satellite CCRMA image for the BB, asking about latency and quality of the built-in audio interface. I didn't get a reply to any of these questions, so I guess i'll have to buy one and figure out myself.
One interesting news though is that they are preparing an image for the Raspberry Pi. It should be the first image for this machine optimized for audio, beside the Raspbian image recently tweaked by Miller Puckette.
And finally, just so you know : "I haven't published about it, but I can safely say that if you don't use the satellite CCRMA images, you will be sorry later on down the road. As far as I know, no one else is bothering to make good images." I'm not sure what to make of these statements ;)
Cheers,
Pierre.
2012/12/20 Bastiaan van den Berg buzz@spacedout.nl
On Mon, Dec 17, 2012 at 7:35 PM, Miller Puckette msp@ucsd.edu wrote:
I could be wrong, but I believe the Beagleboard is ARMv7 and the Pi is ARM v6 - so BB could easily have twice the floating-point erformance of Pi. I'm not sure why and whether there would be any difference in Jack for the two. I believe (but am not sure at all) that BB doesn't have the same USB problems the Pi has. All in all it's still a very attractive possibliity, perticularly since the Stanford people have tested and exercised it thoroughly.
Yes, BB has a _LOT_ better performance than RasPi ..
Btw, if you want a cheaper board with BB like performance but RasPi like price, look at Cubieboard .. ( http://cubieboard.org/ )
-- buZz
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Am 20.12.2012 um 16:09 schrieb Alexandros Drymonitis adrcki@gmail.com:
I just tried out the latency.pd patch in 7.stuff/tools on my beagleboard and [timer] shifts between 72.5 and 75.4, if this answers your question. I've attended a workshop on the BB with Pd by the buys at CCRMA, it was a rather nice one.
AFAIK you also need to say what the delay setting was in the audio properties of Pd. It makes a difference for the numbers you get.
m.
15 ms, by default, meaning I never touched it, maybe Edgar did the whole setup.
On Thu, Dec 20, 2012 at 5:16 PM, Max abonnements@revolwear.com wrote:
Am 20.12.2012 um 16:09 schrieb Alexandros Drymonitis adrcki@gmail.com:
I just tried out the latency.pd patch in 7.stuff/tools on my beagleboard
and [timer] shifts between 72.5 and 75.4, if this answers your question. I've attended a workshop on the BB with Pd by the buys at CCRMA, it was a rather nice one.
AFAIK you also need to say what the delay setting was in the audio properties of Pd. It makes a difference for the numbers you get.
m.
Hm, ok. So if it's the satellite distro then Pd should run in rt with JACK, which means that the latency you mention is fairly close to the minimum one can get. Am I wrong ? Even 50 ms is unusable for a realtime effect processor. Why call it a "stompbox" workshop then ? :(
Thank you Alexandros!
2012/12/20 Alexandros Drymonitis adrcki@gmail.com
15 ms, by default, meaning I never touched it, maybe Edgar did the whole setup.
On Thu, Dec 20, 2012 at 5:16 PM, Max abonnements@revolwear.com wrote:
Am 20.12.2012 um 16:09 schrieb Alexandros Drymonitis adrcki@gmail.com:
I just tried out the latency.pd patch in 7.stuff/tools on my
beagleboard and [timer] shifts between 72.5 and 75.4, if this answers your question. I've attended a workshop on the BB with Pd by the buys at CCRMA, it was a rather nice one.
AFAIK you also need to say what the delay setting was in the audio properties of Pd. It makes a difference for the numbers you get.
m.
Hi,
just so people know, there is some hardware, not open, but way more powerful available for cheap: look for "MK808 tv box" on geekbuying. Dual arm cortex A9 running Android, 1GB ram, at less than 60USD.
Maybe can pd runs on these.
Pierre Massat wrote:
Dear List,
I came accross this "Stompbox design workshop" ( https://ccrma.stanford.edu/wiki/Stompbox_2011). They use Pd on a beagleboard to make stompboxes. They talk about the older version of the beagleboard with a CPU running at around 700 MHz. This is the same speed as the RPi's.
One thing to remember with all these boards is that ARM boards are not PCs. That means that they are not easily interchangeable. There is not standardized techniques for booting, for example. That means you better be sure you can get the distro that you want running on that board. A Debian image running on a Beagleboard most likely will not boot on a RPi or the MK808.
.hc
On Dec 20, 2012, at 10:36 AM, Charles Goyard wrote:
Hi,
just so people know, there is some hardware, not open, but way more powerful available for cheap: look for "MK808 tv box" on geekbuying. Dual arm cortex A9 running Android, 1GB ram, at less than 60USD.
Maybe can pd runs on these.
Pierre Massat wrote:
Dear List,
I came accross this "Stompbox design workshop" ( https://ccrma.stanford.edu/wiki/Stompbox_2011). They use Pd on a beagleboard to make stompboxes. They talk about the older version of the beagleboard with a CPU running at around 700 MHz. This is the same speed as the RPi's.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Hi all,
We'll present the RPi version of the Satellite in a lab at TEI'13 ( http://www.tei-conf.org/13/studios, studio 2), which is basically a reduced version of that workshop. So, I guess the distro will be available by then, in 5 weeks.
Quim
2012/12/20 Hans-Christoph Steiner hans@at.or.at
One thing to remember with all these boards is that ARM boards are not PCs. That means that they are not easily interchangeable. There is not standardized techniques for booting, for example. That means you better be sure you can get the distro that you want running on that board. A Debian image running on a Beagleboard most likely will not boot on a RPi or the MK808.
.hc
On Dec 20, 2012, at 10:36 AM, Charles Goyard wrote:
Hi,
just so people know, there is some hardware, not open, but way more powerful available for cheap: look for "MK808 tv box" on geekbuying. Dual arm cortex A9 running Android, 1GB ram, at less than 60USD.
Maybe can pd runs on these.
Pierre Massat wrote:
Dear List,
I came accross this "Stompbox design workshop" ( https://ccrma.stanford.edu/wiki/Stompbox_2011). They use Pd on a beagleboard to make stompboxes. They talk about the older version of the beagleboard with a CPU running
at
around 700 MHz. This is the same speed as the RPi's.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management ->
http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
On Fri, Dec 21, 2012 at 12:36 AM, Charles Goyard cg@fsck.fr wrote:
Hi,
just so people know, there is some hardware, not open, but way more powerful available for cheap: look for "MK808 tv box" on geekbuying. Dual arm cortex A9 running Android, 1GB ram, at less than 60USD.
Maybe can pd runs on these.
There are also ARM-v7/A9 Exynos based boards from a company here in Korea that apparently run Ubuntu. These are the same chips that power devices like the Samsung Galaxy SIII and Note2. They are not exactly impulse buys like the Pi, but a quad core based board with 2GB ram running at 1.7Ghz for under $100 still seems like a pretty good deal to me.
See here: http://www.hardkernel.com/renewal_2011/products/prdt_info.php?g_code=G135235...
-m
-
Alexandros Drymonitis -
Bastiaan van den Berg -
Charles Goyard -
Hans-Christoph Steiner -
Max -
michael noble -
Miller Puckette -
Pierre Massat -
Quim Llimona