Hello, I noticed a change between pd-extended and pd 0.49 on OSX regarding midi out.
My test patch is a 20ms metro that send 512 random midi control change. This patch works perfectly under linux, with all pd version. (The usb-midi device can handle it without any problem.) On osX, I don't have any problem with pd extended. With pd 0.49, I can't send that much data. pd is buffering the midi message : when I stop the metro it keep sending data for few seconds or more before the buffer is empty.
So, is there anything I can do to make 0.49 faster? Should I fill a bug report?
(if you are wondering why I need to send so much midi data, it's because I use a MIDI to DMX device to control a huge amount of led. I use this interface for 15 years under linux, so I know there are no problem on this side.)
cheers Cyrille
Hi Cyrille,
Yes there's an open issue for it and it is for macOS & Windows : https://github.com/pure-data/pure-data/issues/554
Would be great if it gets fixed.
May be you spot the bug in the code?
?
:)
Mensaje telepatico asistido por maquinas.
On 4/9/2019 3:29 PM, cyrille henry wrote:
Hello, I noticed a change between pd-extended and pd 0.49 on OSX regarding midi out.
My test patch is a 20ms metro that send 512 random midi control change. This patch works perfectly under linux, with all pd version. (The usb-midi device can handle it without any problem.) On osX, I don't have any problem with pd extended. With pd 0.49, I can't send that much data. pd is buffering the midi message : when I stop the metro it keep sending data for few seconds or more before the buffer is empty.
So, is there anything I can do to make 0.49 faster? Should I fill a bug report?
(if you are wondering why I need to send so much midi data, it's because I use a MIDI to DMX device to control a huge amount of led. I use this interface for 15 years under linux, so I know there are no problem on this side.)
cheers Cyrille
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
Hello, My problem is very different than this issue. It's about sending data, not receiving. The problem is that the speed is not fast enough: latency can occur with large amount of data, but the midi port is not stuck.
I did not look at the code: I don't have a mac, I can just occasionally test on this platform.
cheers c
Le 09/04/2019 à 22:55, Lucas Cordiviola a écrit :
Hi Cyrille,
Yes there's an open issue for it and it is for macOS & Windows : https://github.com/pure-data/pure-data/issues/554
Would be great if it gets fixed.
May be you spot the bug in the code?
?
:)
Mensaje telepatico asistido por maquinas.
On 4/9/2019 3:29 PM, cyrille henry wrote:
Hello, I noticed a change between pd-extended and pd 0.49 on OSX regarding midi out.
My test patch is a 20ms metro that send 512 random midi control change. This patch works perfectly under linux, with all pd version. (The usb-midi device can handle it without any problem.) On osX, I don't have any problem with pd extended. With pd 0.49, I can't send that much data. pd is buffering the midi message : when I stop the metro it keep sending data for few seconds or more before the buffer is empty.
So, is there anything I can do to make 0.49 faster? Should I fill a bug report?
(if you are wondering why I need to send so much midi data, it's because I use a MIDI to DMX device to control a huge amount of led. I use this interface for 15 years under linux, so I know there are no problem on this side.)
cheers Cyrille
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
On 4/9/2019 6:41 PM, cyrille henry wrote:
My problem is very different than this issue. It's about sending data, not receiving.
Sorry I was a bit "general".
I meant Portmidi (mac & windows) has some input/output issues but Alsamidi (linux only) is Ok.
The output thing was discovered by you.
:)
Mensaje telepatico asistido por maquinas.
ok, thanks, I just fill a bug report on github. cheers C
Le 10/04/2019 à 01:22, Lucas Cordiviola a écrit :
On 4/9/2019 6:41 PM, cyrille henry wrote:
My problem is very different than this issue. It's about sending data, not receiving.
Sorry I was a bit "general".
I meant Portmidi (mac & windows) has some input/output issues but Alsamidi (linux only) is Ok.
The output thing was discovered by you.
:)
Mensaje telepatico asistido por maquinas.
Kingston University is offering 10 fully funded and further partially funded PhD studentships. If you are interested or know of other who might be interested contact Dr. Oded Ben-Tal Post Graduate Research students coordinator for music. The deadline for application is May 19th, further details at https://www.kingston.ac.uk/research/research-degrees/funding/phd-studentship... Music is part of the Kingston School of Arts. We would particularly welcome practice-based research projects which can be focused on composition, improvisation, production, or interdisciplinary projects enhancing our links with other artistic practices in the School of Arts.
Oded
-
cyrille henry -
Lucas Cordiviola -
oded＠ccrma.Stanford.EDU