sorry, i've mistaken logitech and logilink for the same think, they are both "logi", but different brands

2013/1/31 Alexandre Torres Porres porres@gmail.com

by the way, there's a list of peripherals here with some sound cards verified, I guess you people know about that :)

http://elinux.org/RPi_VerifiedPeripherals

there's another cheap logilink one there http://www.computeruniverse.net/en/products/90378340/logilink-ua0053.asp

2013/1/31 Alexandre Torres Porres porres@gmail.com

...

does it work for line input or just mic? thanks!

Message: 1

...

Date: Wed, 30 Jan 2013 19:31:04 -0500 From: "me.grimm" megrimm@gmail.com Subject: [PD] Working RPI Soundcards (was raspberry pi user experience) To: "Peter P." p8rpp@aol.com Cc: Pd-list@iem.at Message-ID: <CACE5Q14VcCKKd9PM+6Ppm7g=_ BkG6fpfZ5uoAUFmGQE3qj7Ryw@mail.gmail.com> Content-Type: text/plain; charset=ISO-8859-1

i just tested this card:

Logitech USB To 3.5mm Jack Audio Adapter ASIN: B0058P0I2C http://www.amazon.com/gp/product/B0058P0I2C/ref=oh_details_o02_s00_i00

w/ dwc_otg.speed=1 in /boot/cmdline.txt

works between 10-20ms with no clicks with adc~ running various patches from manual (reverb, octave.divider, etc)

also works in conjunction with my edimax wifi adapter + vnc screensharing (no need for ethernet cable)

i added it and the beringer to the wiki. others should be added as discovered....

http://puredata.info/docs/raspberry-pi/FrontPage

m

That's great, can't wait till you try them all, it'd be just too expensive for me to import them all and try here, so I'll wait for the winner. Are you trying any of the cards from DealExtreme?

cheers

2013/1/31 me.grimm megrimm@gmail.com

i just tried a cheap-o mic but im assuming line-in would be the same no? i mean it was a passive mic so what the difference would be between that and a direct guitar line, for example, i do not know.

again, i ordered like 7 or 8 of these cheap-o line ins. ill test them as they get delivered and report back.

yes! i know of this http://elinux.org/RPi_VerifiedPeripherals#USB_Sound_Cards but not all of these test well with pd i am assuming by all the recent sound card crackling probs. also there are more that have not been tested or/and have been but not listed here im sure....

m

On Thu, Jan 31, 2013 at 9:56 AM, Alexandre Torres Porres porres@gmail.com wrote:

...

does it work for line input or just mic? thanks!

...

Message: 1 Date: Wed, 30 Jan 2013 19:31:04 -0500 From: "me.grimm" megrimm@gmail.com Subject: [PD] Working RPI Soundcards (was raspberry pi user experience) To: "Peter P." p8rpp@aol.com Cc: Pd-list@iem.at Message-ID:

CACE5Q14VcCKKd9PM+6Ppm7g=_BkG6fpfZ5uoAUFmGQE3qj7Ryw@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1

i just tested this card:

Logitech USB To 3.5mm Jack Audio Adapter ASIN: B0058P0I2C http://www.amazon.com/gp/product/B0058P0I2C/ref=oh_details_o02_s00_i00

w/ dwc_otg.speed=1 in /boot/cmdline.txt

works between 10-20ms with no clicks with adc~ running various patches from manual (reverb, octave.divider, etc)

also works in conjunction with my edimax wifi adapter + vnc screensharing (no need for ethernet cable)

i added it and the beringer to the wiki. others should be added as discovered....

http://puredata.info/docs/raspberry-pi/FrontPage

m

-- ____________________ m.e.grimm | m.f.a | ed.m. megrimm@gmail.com _________________________________

me? guide? to what?

Anyway, regarding RPI, are these cheap soundcards being tested supposed to help in the CPU load?

I've testing some stuff with my audio card here on the macbook air, and it's a RME Multiface II (via PCI-Express Card) and it doesn't seem to affect so much the CPU load (as measured by pd's meter). I can see a better improvement on latency though.

So, regarding the RPI, is the concern only about the latency? And how much can a sound card help after all in any occasion/computer/systrem.

cheers

2013/1/31 me.grimm megrimm@gmail.com

...

trying any of the cards from DealExtreme?

no no all. just the ones on amazon. i submit my receipts to the uni so they pay me back. the students can use what ever i get that doesnt work.

BTW. if you write a basic intro guide that a 1st or 2nd year student can follow (in english :)) ill use it in my class....

m

On Thu, Jan 31, 2013 at 10:56 AM, Alexandre Torres Porres porres@gmail.com wrote:

...

That's great, can't wait till you try them all, it'd be just too

expensive

...

for me to import them all and try here, so I'll wait for the winner. Are

you

...

trying any of the cards from DealExtreme?

cheers

2013/1/31 me.grimm megrimm@gmail.com

...

i just tried a cheap-o mic but im assuming line-in would be the same no? i mean it was a passive mic so what the difference would be between that and a direct guitar line, for example, i do not know.

again, i ordered like 7 or 8 of these cheap-o line ins. ill test them as they get delivered and report back.

yes! i know of this http://elinux.org/RPi_VerifiedPeripherals#USB_Sound_Cards but not all of these test well with pd i am assuming by all the recent sound card crackling probs. also there are more that have not been tested or/and have been but not listed here im sure....

m

On Thu, Jan 31, 2013 at 9:56 AM, Alexandre Torres Porres porres@gmail.com wrote:

...

does it work for line input or just mic? thanks!

...

Message: 1 Date: Wed, 30 Jan 2013 19:31:04 -0500 From: "me.grimm" megrimm@gmail.com Subject: [PD] Working RPI Soundcards (was raspberry pi user experience) To: "Peter P." p8rpp@aol.com Cc: Pd-list@iem.at Message-ID:

CACE5Q14VcCKKd9PM+6Ppm7g=_BkG6fpfZ5uoAUFmGQE3qj7Ryw@mail.gmail.com Content-Type: text/plain; charset=ISO-8859-1

i just tested this card:

Logitech USB To 3.5mm Jack Audio Adapter ASIN: B0058P0I2C

http://www.amazon.com/gp/product/B0058P0I2C/ref=oh_details_o02_s00_i00

...

...

...

...

w/ dwc_otg.speed=1 in /boot/cmdline.txt

works between 10-20ms with no clicks with adc~ running various

patches

...

...

...

...

from manual (reverb, octave.divider, etc)

also works in conjunction with my edimax wifi adapter + vnc screensharing (no need for ethernet cable)

i added it and the beringer to the wiki. others should be added as discovered....

http://puredata.info/docs/raspberry-pi/FrontPage

m

-- ____________________ m.e.grimm | m.f.a | ed.m. megrimm@gmail.com _________________________________

-- ____________________ m.e.grimm | m.f.a | ed.m. megrimm@gmail.com _________________________________

On Fre, 2013-02-01 at 04:14 -0200, Alexandre Torres Porres wrote:

by the way, since I've been abusing so much of the GUI to build me fancy interfaces, I can see how easily it can choke on its update and flood the CPU.

Now, could an external video card help on that? What is the issue?

Since the GUI process (called pd-gui) is computed on the CPU, a dedicated video card wouldn't help.

I'm using a macbook air and I see that the video processor is built-in the CPU or something like that, hence no proper videocard. I figure the RPI has gotta be something like that.

So, does it mean a macbook pro or other laptop / computer with some badass video processor or card can run the patches better even with same clock?

No, as nothing of current Pd can be hardware accelerated. A fast GPU would only provide a benefit for Gem. However, pd and pd-gui are two separate processes and they run on separate cores, if there is more than one. Most laptops (netbooks not included) have at least two cores already. This means on many laptops, pd-gui does not eat away CPU power from the pd process.

Beware that there is also another bottleneck: the communication between pd and pd-gui. A possible cause for audio-dropouts is the saturation of the TCP link between the two as it might be triggered by frequent array updates or similar. If you experience troubles, try to make pd and pd-gui talk less to each other (for instance, by reducing the update rates for your GUIs).

Roman

On 01/02/13 21:22, Alexandre Torres Porres wrote:

...

No, as nothing of current Pd can be hardware accelerated. A fast GPU would only provide a benefit for Gem.

Therefore, you can improve the performance of GEM on a RPI with an external video card, cool.

??? I don't think you could add an external card in any way that could provide openGL. The built in GPU and its drivers don't provide openGL either, so no GEM at all on the RPI (unless GEM is ported to ES, or some GL to ES bridge is written). The GPU parts of the hardware are closed source, so no hacking the driver or such possible either. Same applies to Beagleboards and other similar devices as well, they all use ES rather than GL.

Sticking to the Pi issue, since it is so economic in CPU power, what are the efficiency benefits from audio cards more than just number of inputs and outputs?

They are needed on the Pi ... there is no analogue input and the built-in analogue out is very poor, hence you need some external USB card to do more than use the stereo digital audio part of the HDMI output.

Simon

pd and pd-gui are two separate processes and they run on separate cores, if there is more than one.

That makes me wonder now hoe the [pd~] separates the processes between 2 cores then. I really thought PD was a single core processor that you could only split with [pd~].

cheers

2013/2/1 Roman Haefeli reduzent@gmail.com

On Fre, 2013-02-01 at 04:14 -0200, Alexandre Torres Porres wrote:

...

by the way, since I've been abusing so much of the GUI to build me fancy interfaces, I can see how easily it can choke on its update and flood the CPU.

Now, could an external video card help on that? What is the issue?

Since the GUI process (called pd-gui) is computed on the CPU, a dedicated video card wouldn't help.

...

I'm using a macbook air and I see that the video processor is built-in the CPU or something like that, hence no proper videocard. I figure the RPI has gotta be something like that.

So, does it mean a macbook pro or other laptop / computer with some badass video processor or card can run the patches better even with same clock?

No, as nothing of current Pd can be hardware accelerated. A fast GPU would only provide a benefit for Gem. However, pd and pd-gui are two separate processes and they run on separate cores, if there is more than one. Most laptops (netbooks not included) have at least two cores already. This means on many laptops, pd-gui does not eat away CPU power from the pd process.

Beware that there is also another bottleneck: the communication between pd and pd-gui. A possible cause for audio-dropouts is the saturation of the TCP link between the two as it might be triggered by frequent array updates or similar. If you experience troubles, try to make pd and pd-gui talk less to each other (for instance, by reducing the update rates for your GUIs).

Roman

---

Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list