Date: Wed, 8 Dec 1999 12:49:01 -0800 (PST) From: The Utz Family <utz@serv.net> To: "Zach N. Heilig" <znh@thequest.net> Cc: Jay Sachs <jsachs@iclick.com>, freebsd-multimedia@FreeBSD.ORG Subject: Re: es1370 strangeness: sample leftovers Message-ID: <Pine.BSF.4.02.9912081238370.15757-100000@itchy.serv.net> In-Reply-To: <263122781FEB1EEF6D5C32D3@mail.uffdaonline.net>
next in thread | previous in thread | raw e-mail | index | archive | help
OTOH this sounds like the buffer isnt gettting free()'ed and malloc()'ed for the new sound. this is a reasonable speedup. having to redo the allocation everytime you want to emit a sample woudl be sub-optimal. so, then , assuming it's *supposed* to work that way then it seems likely that the maintainer has either not added an 'end of sample' token to read up to, or they are failing to recognize the token when they get to it. they may also be doing bulk reads that read the buffer in whole increments and the last read is picking up more data then was actually in the current sound, thus you get the longer remainder. this still pertains to finding and responding appropriately to an 'end of sample token' all of this rumination is assuming that the later sound is shorter than the previous sound, if that isnt the case then my entire theory is hosed. :-) On Wed, 8 Dec 1999, Zach N. Heilig wrote: > On Tue, Dec 07, 1999 at 10:17:39PM -0500, Jay Sachs wrote: > > Ever since I got any sound out of my SBPCI128 (thank you!), up through > > -STABLE as of last night, if I play an audio file (eg .au, .wav), via > > any means (e.g. wmsound, cat file.au > /dev/audio) there appears to be > > about a 1/4 second "leftover" of digitized sound that comes in right > > after the audio sample is done. It's definitely recognizable as leftover > > from a recently played sample. Any clues? > > I have one if these, but I use -current instead of -stable. I have noticed > this problem, I worked around it by adding one second of silence before > and after a sound clip (what gets played extra at the end seems to be > exactly what is missing at the beginning without padding with silence). > > Also, when recording from the sound card, there will be a random length > of silence before the actual recording starts. hmm. no inspiration here, other than it's a ring buffer and the offset is screwed up, but i am just blathering here...... > > My kernel config contains > > > > device pcm0 at isa? port ? tty irq 11 drq 1 flags 0x0 > > For -current, you could get away with just: > > device pcm0 > > I'm not sure if that is relavent for -stable. > > > dmesg yields > > > > es1: <AudioPCI ES1370> rev 0x01 int a irq 11 on pci0.16.0 > > pcm1: using I/O space register mapping at 0xef00 > > pcm0 not found > > I get: > pcm0: <AudioPCI ES1370> irq 11 at device 9.0 on pci0 > > (but then, I'm using -current). > > -- > Zach Heilig <znh@thequest.net> > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > with "unsubscribe freebsd-multimedia" in the body of the message > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-multimedia" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.02.9912081238370.15757-100000>