[LAD] MIDI jitter - was: automation on Linux (modular approach)

Ralf Mardorf ralf.mardorf at alice-dsl.net
Thu Mar 25 13:55:41 UTC 2010


David Olofson wrote:
> On Thursday 25 March 2010, at 13.32.43, David Olofson <david at olofson.net> 
> wrote:
>   
>> On Thursday 25 March 2010, at 12.49.31, Ralf Mardorf <ralf.mardorf at alice-
>> dsl.net> wrote:
>> [...]
>>
>>     
>>> Btw. the
>>> graphics has access to the main memory, unfortunately it's a shared RAM,
>>> OTOH I used HPET so unwanted interrupts because of a shared RAM
>>> shouldn't be the cause, if I do understand the workings of HR timers
>>> correctly.
>>>       
> [...interrupts, DMA etc...]
>
> BTW, the most common problem with graphics and realtime systems seems to be 
> drivers abusing PCI port blocking as a performance hack. When the command 
> buffer on the video card is full, the PCI bus blocks the CPU (completely - no 
> IRQs, no nothing), instead of the driver going to sleep and waiting for an IRQ 
> or some other proper solution. Might improve the 3D framerates slightly, but 
> kills lowlatency audio...
>   

Mobo: M2A-VM HDMI
Graphics: ATI Radeon X1250-based graphics, *onboard*
Slot for another graphics is an PCI Express slot, no AGP etc.

The HDMI card is demounted from the PCI Express slot and HDMI is 
disabled by the BIOS.
I'm not using a proprietary 3D driver and I wasn't able to test one, 
because it doesn't work using Linux, even though there is a 3D driver 
for this card and Linux, perhaps the problem is "...-based".

The USB MIDI interface is a 1 in and 1 out, very cheap swissonic.

I suspect the graphics and the USB device causing to much MIDI jitter.

I need to search for an old test, but have no time to do it right now, 
but I do remember that Windows without HPET was better, than Linux with 
HPET as sequencer timer source, while all other Linux sequencer timer 
sources were much more out of time.

Btw. I don't think that Windows would solve this issue for me, the 
jitter for my current system is bad with both OSs. I know people where 
Linux is bad and Windows does solve this issue ... perhaps Windows will 
cause other issues for them ;).



More information about the Linux-audio-dev mailing list