<div>Seeing that I don't have the issue with the higher bandwith DVB-C card, somethin is obviously wrong with the pchdtv driver or conflicting interrupt on my system.</div>
<div>Because the driver is failry new, I would blame the driver first :)</div>
<div>Who supports the DVB driver, is it <a href="http://linuxtv.org">linuxtv.org</a> or pchdtv?</div>
<div><span class="gmail_quote">On 8/28/05, <b class="gmail_sendername">Peter Knaggs</b> <<a href="mailto:email@example.com">firstname.lastname@example.org</a>> wrote:</span>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">Hi Steve,<br><br>Thanks for your reply.<br>I'm not familiar with dvbsnoop,<br>sounds useful I'll have to give it a try.
<br>There's even a Debian package for it,<br>what do you know :)<br><br>> I get this only with pcdhtv HD-3000 card and<br>> not with my DVB-C card. There is<br>> something wrong with buffer management or<br>> interrupt in the cx88xx driver.
<br><br>I'm using <a href="http://18.104.22.168">22.214.171.124</a> here with the HD3000 and it<br>captures OK from OTA 8VSB ATSC, I don't have cable.<br>At least it captures always OK using the dvb-apps.<br>I'm still learning the DVB API to get pchdtv to
<br>work correctly, hence my questions on the list.<br><br>The interrupt troubles you're having sound similar<br>to those in this post:<br><a href="http://www.pchdtv.com/forum/viewtopic.php?t=933">http://www.pchdtv.com/forum/viewtopic.php?t=933
</a><br><br>> I'm no expert so I'm pulling this out of you know where...<br><br>Eh, me neither. I think I might have goofed up<br>the demux part of my patch. It might be that setting<br>DMX_IMMEDIATE_START like I've done, causes the
<br>demux buffer to start filling, and instead I should<br>do a DMX_START ioctl just before I start the<br>capture, instead. That might explain the EOVERFLOW<br>I'm getting when I read the first block.<br>I'll give it a try, back to the drawing board...
<br><br><br>> I'm usually recording full transport, so I need<br>> to try to do this on 1 pid<br>> only if this happens as well.<br><br>The HD3000 card would still need to transfer the<br>entire stream across the PCI bus, it doesn't do
<br>any hardware filtering afaik. So I don't see why<br>setting the PIDs would make any difference to how<br>it behaves.<br><br>Cheers,<br>Peter.<br></blockquote></div><br>