From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Hofman Subject: Re: Impact of SW on USB Audio? Date: Wed, 11 Jun 2008 23:27:51 +0200 Message-ID: <48504357.7080601@insite.cz> References: <4847F425.7040402@insite.cz> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mailserver.bobrnet.net (bobrnet.cust.inethome.cz [88.146.180.6]) by alsa0.perex.cz (Postfix) with ESMTP id 3472324906 for ; Wed, 11 Jun 2008 23:27:57 +0200 (CEST) In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: alsa-devel-bounces@alsa-project.org Errors-To: alsa-devel-bounces@alsa-project.org To: Takashi Iwai Cc: alsa-devel@alsa-project.org List-Id: alsa-devel@alsa-project.org Takashi Iwai napsal(a): > At Thu, 05 Jun 2008 16:11:49 +0200, > Pavel Hofman wrote: >> Hi, >> >> Please excuse my non-devel question but the most knowledgeable people >> read this mailing list. >> >> Can user-space software (regular players) directly influence quality of >> USB adaptive audio output? IOW, how sensitive is adaptive USB to the >> rhythm of the data infeed to the USB controller? >> >> I understand for standard PCI soundcards it is all only about timely >> delivery of enough data to DMA memory,the card takes care of the rest >> itself. Do USB controllers shield applications from the rest in the same >> manner? >> >> The adaptive mode is sensitive to jitter of the data stream from the USB >> controller to the receiver. Can SW players influence the jitter, e.g. by >> moving the wav files from HDD to RAM disk etc.? I assume a lightly >> loaded system with no major data delays. >> >> Thanks a lot for any insight, I could not google-out any relevant >> information. > > The usb-audio uses isochronous transport, i.e. you'll send packets of > data at a certain period. But, it's a driver implementation issue and > from the user-space POV, there is no big difference. The URB-complete > callback corresponds to the DMA IRQ of PCI devices. > > (Of course, if you want a very short period in a real-time manner, the > transfer efficiency matters largely, though. But it's a different > story than the load you described above.) Takashi, Thanks a lot for your reply which has narrowed down the area I will have to look at detail. Regards, Pavel.