Skip Menu |

This queue is for tickets about the Linux-DVB-DVBT CPAN distribution.

Report information
The Basics
Id: 56362
Status: resolved
Priority: 0/
Queue: Linux-DVB-DVBT
People
Owner: Nobody in particular
Requestors: martin [...] gkc.org.uk
Cc:
AdminCc:
Bug Information
Severity: Important
Broken in: 2.03
Fixed in: (no value)
Attachments
dvb_tune.c
dvbt-record
Linux-DVB-DVBT-2.038.tar.gz
LOG
LOG-scantest
LOG2
LOG3
scantest.tar.gz

History Show all quoted text
  Wed Apr 07 15:07:55 2010 martin [...] gkc.org.uk - Ticket created
Subject: Does not work with DiBcom 3000P/M-C DVB-T
I have three tuners on my machine, all are detected with dvbt-devices. The Twinhan VP7045/46 USB DVB-T seems to work OK. The other two are AVerMedia AVerTV tuners, detected as "DiBcom 3000P/M-C DVB-T" and these do not work at all. dvbt-strength -a 1 gives: Unable to tune frontend at /usr/local/bin/dvbt-strength line 54 Note that dvbt-record does not accept the -a option to choose an adapter.
  Thu Apr 08 09:46:37 2010 cpan [...] sdprice.plus.com - Correspondence added
On Wed Apr 07 15:07:55 2010, MWARD wrote: Show quoted text
> I have three tuners on my machine, all are detected with dvbt-devices. > > The Twinhan VP7045/46 USB DVB-T seems to work OK. > > The other two are AVerMedia AVerTV tuners, detected as "DiBcom 3000P/M-C > DVB-T" and these do not work at all. > > dvbt-strength -a 1 > > gives: > > Unable to tune frontend at /usr/local/bin/dvbt-strength line 54 > > Note that dvbt-record does not accept the -a option to choose an adapter. >
Sorry to hear you are having problems. Could you run dvbt-strength with '-debug 15 -verbose 15' and attach the resulting output? It may well be that the default timeout is not sufficient for these cards. You can change the timeout in your own scripts by setting the 'timeout' field of the dvb object in milliseconds (the default is 900). In general, I've provided the installed scripts more as examples of how the module can be used, rather than as production scripts. I was expecting people to take copies and roll their own. Having said that, I'll modify the record script (and any other I've forgotten) to add the -a option!
  Thu Apr 08 09:46:39 2010 The RT System itself - Status changed from 'new' to 'open'
  Thu Apr 08 11:08:19 2010 martin [...] gkc.org.uk - Correspondence added
Subject: Re: [rt.cpan.org #56362] Does not work with DiBcom 3000P/M-C DVB-T
Date: Thu, 8 Apr 2010 16:07:58 +0100
To: bug-Linux-DVB-DVBT [...] rt.cpan.org
From: Martin Ward <martin [...] gkc.org.uk>
On Thursday 08 Apr 2010 14:46, Steve Price via RT wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=56362 > > > On Wed Apr 07 15:07:55 2010, MWARD wrote:
> > I have three tuners on my machine, all are detected with dvbt-devices. > > > > The Twinhan VP7045/46 USB DVB-T seems to work OK. > > > > The other two are AVerMedia AVerTV tuners, detected as "DiBcom 3000P/M-C > > DVB-T" and these do not work at all. > > > > dvbt-strength -a 1 > > > > gives: > > > > Unable to tune frontend at /usr/local/bin/dvbt-strength line 54 > > > > Note that dvbt-record does not accept the -a option to choose an adapter.
> > Sorry to hear you are having problems. Could you run dvbt-strength with > '-debug 15 -verbose 15' and attach the resulting output?
Thanks for the quick response. The output is attached. Show quoted text
> It may well be that the default timeout is not sufficient for these > cards. You can change the timeout in your own scripts by setting the > 'timeout' field of the dvb object in milliseconds (the default is 900).
I tried changing the timeout: my $dvb = Linux::DVB::DVBT->new('adapter_num' => $adapter, 'timeout' => 90000) ; but this made no difference: it is obviously not timing out. (See the attached LOG3 file). Show quoted text
> In general, I've provided the installed scripts more as examples of how > the module can be used, rather than as production scripts. I was > expecting people to take copies and roll their own. Having said that, > I'll modify the record script (and any other I've forgotten) to add the > -a option!
Thanks. Currently I use two different methods to record from my two types of tuner: The Twinhan uses: dvbstream -f freq-in-KHz 8192 to broadcast the whole transport stream to the multicast network (which is bound to the loopback device: I found that otherwise the data would spam my old wireless router!). Then individual programmes are recorded with dvb_dumprtp and dvb_ts_filter like this: dvb_dumprtp | dvb_ts_filter 201 401 >! test.mpg Since the data is on a multicast network, we can record several programmes on the same stream. This approach didn't work with the AVerTV tuners, for some reason, so to record programmes from these I run: tzap -a adapter_number -r "BBC ONE" in the background to tune in to the station, and then: setenv DEMUX /dev/dvb/adapter1/demux0 setenv DVR /dev/dvb/adapter1/dvr0 test_dvr file.mpg 0x0258 0x0259 to capture the output. The test_dvr source code is in linuxtv-dvb-apps-1.1.1, which I downloaded from here: http://www.linuxtv.org/downloads/ I wrote some perl scripts to handle the details. -- Martin STRL Reader in Software Engineering and Royal Society Industry Fellow martin@gkc.org.uk http://www.cse.dmu.ac.uk/~mward/ Erdos number: 4 G.K.Chesterton web site: http://www.cse.dmu.ac.uk/~mward/gkc/ Mirrors: http://www.gkc.org.uk and http://www.gkc.org.uk/gkc

Message body is not shown because sender requested not to inline it.

Message body is not shown because sender requested not to inline it.

  Fri Apr 09 08:56:29 2010 cpan [...] sdprice.plus.com - Correspondence added
That's very weird, usually I see at most a couple of iterations and then the tuner comes back with the status 0x1F (= everything's stable and locked). All you seem to get is signal presence (0x01). Is there any chance you can run the 'scan/dvbscan' utility on one of these cards? The reason for asking is that it's another program that I know of that uses similar code to me for setting the frontend. We both allow approx. 2 seconds for the tuner to become stable. On a related note, I've spotted that my code doesn't actually use the reported timeout while waiting for the frontend (the timeout value is actually used elsewhere). I'm just amending the code so that changing the timeout *will* actually affect this too!
  Fri Apr 09 10:02:45 2010 martin [...] gkc.org.uk - Correspondence added
Subject: Re: [rt.cpan.org #56362] Does not work with DiBcom 3000P/M-C DVB-T
Date: Fri, 9 Apr 2010 15:02:17 +0100
To: bug-Linux-DVB-DVBT [...] rt.cpan.org
From: Martin Ward <martin [...] gkc.org.uk>
On Friday 09 Apr 2010 13:56, Steve Price via RT wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=56362 > > > That's very weird, usually I see at most a couple of iterations and > then the tuner comes back with the status 0x1F (= everything's stable > and locked). All you seem to get is signal presence (0x01). > > Is there any chance you can run the 'scan/dvbscan' utility on one of > these cards? The reason for asking is that it's another program that I > know of that uses similar code to me for setting the frontend. We both > allow approx. 2 seconds for the tuner to become stable. > > On a related note, I've spotted that my code doesn't actually use the > reported timeout while waiting for the frontend (the timeout value is > actually used elsewhere). I'm just amending the code so that changing > the timeout *will* actually affect this too!
Attached is the output of: scandvb -a 1 /usr/share/dvb-apps/scan/dvb-t/uk-PontopPike scandvb is from dvb-apps-1.1.1-1mdv2007.0 LOG2 is the output of: scandvb -v -v -v -a 1 /usr/share/dvb-apps/scan/dvb-t/uk-PontopPike Is that what you wanted? /usr/share/dvb-apps/scan/dvb-t/uk-PontopPike contains this information: # Pontop Pike, UK # T freq bw fec_hi fec_lo mod transmission-mode guard-interval hierarchy T 690000000 8MHz 3/4 3/4 QAM16 2k 1/32 NONE -- Martin STRL Reader in Software Engineering and Royal Society Industry Fellow martin@gkc.org.uk http://www.cse.dmu.ac.uk/~mward/ Erdos number: 4 G.K.Chesterton web site: http://www.cse.dmu.ac.uk/~mward/gkc/ Mirrors: http://www.gkc.org.uk and http://www.gkc.org.uk/gkc

Message body is not shown because sender requested not to inline it.

Message body is not shown because sender requested not to inline it.

  Mon Apr 12 04:39:31 2010 cpan [...] sdprice.plus.com - Correspondence added
Thanks, that was exactly what I wanted. The only difference I can see is that my code accesses the hardware with a shorter period between reads. I can only think that the device needs longer between reads (?) I've amended the dvb_tune.c file to match the period used by 'scan'. Could you replace the existing build file (in dvb_lib) with this new one, then rebuild and retry? Other this, I'm at a bit of a loss as to what could be causing the problem! (I assume you know that you can run the new build without installing it, by running in the build directory: $ perl -Mblib script/dvbt-strength -a 1 )
Subject: dvb_tune.c

Message body is not shown because it is too large.

  Mon Apr 12 10:12:32 2010 martin [...] gkc.org.uk - Correspondence added
Subject: Re: [rt.cpan.org #56362] Does not work with DiBcom 3000P/M-C DVB-T
Date: Mon, 12 Apr 2010 15:12:14 +0100
To: bug-Linux-DVB-DVBT [...] rt.cpan.org
From: Martin Ward <martin [...] gkc.org.uk>
On Monday 12 Apr 2010 09:39, Steve Price via RT wrote: Show quoted text
> I've amended the dvb_tune.c file to match the period used by 'scan'. > Could you replace the existing build file (in dvb_lib) with this new > one, then rebuild and retry?
This version does not work with either tuner! % perl -Mblib script/dvbt-strength -v 1 [dvbt-strength (6609) 14:05:35 12/04/2010 DVB0] =============================================================== [dvbt-strength (6609) 14:05:35 12/04/2010 DVB0] dvbt-strength v2.001 [dvbt-strength (6609) 14:05:35 12/04/2010 DVB0] Linux::DVB::DVBT v2.03 [dvbt-strength (6609) 14:05:35 12/04/2010 DVB0] == Locked DVB0 == Unable to tune frontend at script/dvbt-strength line 51 If I run it twice in quick sucession, then it works the second time. The problem appears to be with the new timeout code: With USLEEP = 200 and timeout = 900, dvb_frontend_wait_lock will loop 4 times, waiting 200mS each time: which is only 800 mS. Changing USLEEP to 100 fixed it, but I suggest that the loop calculation should also round up: // timeout is in ms - convert to number of wait loops timeout = (timeout + USLEEP - 1) / USLEEP ; It still doesn't work with the AVerMedia AVerTV: time perl -Mblib script/dvbt-strength -a 1 -v 1 [dvbt-strength (7753) 14:18:12 12/04/2010 DVB1] =============================================================== [dvbt-strength (7753) 14:18:12 12/04/2010 DVB1] dvbt-strength v2.001 [dvbt-strength (7753) 14:18:12 12/04/2010 DVB1] Linux::DVB::DVBT v2.03 [dvbt-strength (7753) 14:18:12 12/04/2010 DVB1] == Locked DVB1 == Unable to tune frontend at script/dvbt-strength line 51 0.132u 0.012s 0:10.11 1.3% 0+0k 0+0io 0pf+0w The above was with the timeout value changed to 9000. -- Martin STRL Reader in Software Engineering and Royal Society Industry Fellow martin@gkc.org.uk http://www.cse.dmu.ac.uk/~mward/ Erdos number: 4 G.K.Chesterton web site: http://www.cse.dmu.ac.uk/~mward/gkc/ Mirrors: http://www.gkc.org.uk and http://www.gkc.org.uk/gkc
  Wed Apr 14 04:02:19 2010 cpan [...] sdprice.plus.com - Correspondence added
Hi Martin, Sorry about that - I knew that it required the changes in the other files, but thought that this file would be enough since it should have locked after 1 or 2 attempts. I've attached a hacked version of 'scan', that I've added debug timestamps and readback of the frontend settings. I've found this useful in the past for debugging other people's hardware. Would you mind building it (hopefully just running 'make' should do it), then run './scantest' with your freq file and post the results? You don't need to set the debug level, I've nailed it to "high". You probably want to set '-a 1' as usual though. I'm hoping this extra information will point me in the right direction (since it can only be (a) frontend settings and/or (b) the polling loop waiting for lock).
Subject: scantest.tar.gz
Download scantest.tar.gz
application/x-gzip 28k

Message body not shown because it is not plain text.

  Wed Apr 14 04:51:34 2010 martin [...] gkc.org.uk - Correspondence added
Subject: Re: [rt.cpan.org #56362] Does not work with DiBcom 3000P/M-C DVB-T
Date: Wed, 14 Apr 2010 09:50:25 +0100
To: bug-Linux-DVB-DVBT [...] rt.cpan.org
From: Martin Ward <martin [...] gkc.org.uk>
On Wednesday 14 Apr 2010 09:02, Steve Price via RT wrote: Show quoted text
> Sorry about that - I knew that it required the changes in the other > files, but thought that this file would be enough since it should have > locked after 1 or 2 attempts. > > I've attached a hacked version of 'scan', that I've added debug > timestamps and readback of the frontend settings. I've found this useful > in the past for debugging other people's hardware. Would you mind > building it (hopefully just running 'make' should do it), then run > './scantest' with your freq file and post the results? You don't need to > set the debug level, I've nailed it to "high". You probably want to set > '-a 1' as usual though.
Here is the output of ./scantest -a 1 /usr/share/dvb-apps/scan/dvb-t/uk-PontopPike -- Martin STRL Reader in Software Engineering and Royal Society Industry Fellow martin@gkc.org.uk http://www.cse.dmu.ac.uk/~mward/ Erdos number: 4 G.K.Chesterton web site: http://www.cse.dmu.ac.uk/~mward/gkc/ Mirrors: http://www.gkc.org.uk and http://www.gkc.org.uk/gkc

Message body is not shown because sender requested not to inline it.

  Wed Apr 14 08:34:32 2010 cpan [...] sdprice.plus.com - Correspondence added
Right, I think I've spotted the problem. I'm setting the spectral inversion to 0 (which has worked for all other cases); 'scan' sets it to AUTO & your tuner reads back with it set to 1. I was never very happy with the code associated with the inversion bit, all those years ago when I first ported it, but I assumed it was based on some hidden knowledge. I now know a little more about the standards, and I'm happier with what I've changed the code to! I've attached a new debug build of my module. See if that works for you. If so, please also try out the other scripts and let me know if you have any other problems before I release it (I've been working on the epg & made various tweaks to the ffmpeg helper module).
Subject: Linux-DVB-DVBT-2.038.tar.gz
Download Linux-DVB-DVBT-2.038.tar.gz
application/x-gzip 179.5k

Message body not shown because it is not plain text.

  Wed Apr 14 18:15:54 2010 martin [...] gkc.org.uk - Correspondence added
Subject: Re: [rt.cpan.org #56362] Does not work with DiBcom 3000P/M-C DVB-T
Date: Wed, 14 Apr 2010 23:15:37 +0100
To: bug-Linux-DVB-DVBT [...] rt.cpan.org
From: Martin Ward <martin [...] gkc.org.uk>
On Wednesday 14 Apr 2010 13:34, Steve Price via RT wrote: Show quoted text
> I've attached a new debug build of my module. See if that works for you. > If so, please also try out the other scripts and let me know if you have > any other problems before I release it (I've been working on the epg & > made various tweaks to the ffmpeg helper module).
Looks good so far: except that dvbt-record doesn't have the -a option! I have tested dvbt-devices, dvbt-scan, dvbt-strength and dvbt-record -- Martin STRL Reader in Software Engineering and Royal Society Industry Fellow martin@gkc.org.uk http://www.cse.dmu.ac.uk/~mward/ Erdos number: 4 G.K.Chesterton web site: http://www.cse.dmu.ac.uk/~mward/gkc/ Mirrors: http://www.gkc.org.uk and http://www.gkc.org.uk/gkc
  Thu Apr 15 05:21:21 2010 martin [...] gkc.org.uk - Correspondence added
Subject: Re: [rt.cpan.org #56362] Does not work with DiBcom 3000P/M-C DVB-T
Date: Thu, 15 Apr 2010 10:21:07 +0100
To: bug-Linux-DVB-DVBT [...] rt.cpan.org
From: Martin Ward <martin [...] gkc.org.uk>
On Wednesday 14 Apr 2010 13:34, Steve Price via RT wrote: Show quoted text
> I've attached a new debug build of my module. See if that works for you. > If so, please also try out the other scripts and let me know if you have > any other problems before I release it (I've been working on the epg & > made various tweaks to the ffmpeg helper module).
Also tested dvbt-epg and dvbt-multirec, which both work. -- Martin STRL Reader in Software Engineering and Royal Society Industry Fellow martin@gkc.org.uk http://www.cse.dmu.ac.uk/~mward/ Erdos number: 4 G.K.Chesterton web site: http://www.cse.dmu.ac.uk/~mward/gkc/ Mirrors: http://www.gkc.org.uk and http://www.gkc.org.uk/gkc
  Thu Apr 15 07:08:33 2010 cpan [...] sdprice.plus.com - Correspondence added
Excellent news. Sorry about the script (it's typical, I went round them all adding the -a option, but in this particular script I forgot to update the GetOpts!). I've attached the changed file. I'll tidy up my files and release this version as 2.04. Thanks for your assistance in finding the root cause.
Subject: dvbt-record
Download dvbt-record
application/octet-stream 3.7k

Message body not shown because it is not plain text.

  Thu Apr 15 09:27:30 2010 martin [...] gkc.org.uk - Correspondence added
Subject: Re: [rt.cpan.org #56362] Does not work with DiBcom 3000P/M-C DVB-T
Date: Thu, 15 Apr 2010 14:27:15 +0100
To: bug-Linux-DVB-DVBT [...] rt.cpan.org
From: Martin Ward <martin [...] gkc.org.uk>
On Thursday 15 Apr 2010 12:08, Steve Price via RT wrote: Show quoted text
> Excellent news. Sorry about the script (it's typical, I went round them > all adding the -a option, but in this particular script I forgot to > update the GetOpts!). I've attached the changed file. > > I'll tidy up my files and release this version as 2.04. Thanks for your > assistance in finding the root cause.
You're welcome :-) -- Martin STRL Reader in Software Engineering and Royal Society Industry Fellow martin@gkc.org.uk http://www.cse.dmu.ac.uk/~mward/ Erdos number: 4 G.K.Chesterton web site: http://www.cse.dmu.ac.uk/~mward/gkc/ Mirrors: http://www.gkc.org.uk and http://www.gkc.org.uk/gkc
  Fri Apr 16 03:48:26 2010 cpan [...] sdprice.plus.com - Correspondence added
Hi Martin, Just released 2.04 on to CPAN. Let me know if you see any further issues.
  Fri Apr 16 03:48:55 2010 cpan [...] sdprice.plus.com - Correspondence added
Resolved in version 2.04
  Fri Apr 16 03:48:56 2010 cpan [...] sdprice.plus.com - Status changed from 'open' to 'resolved'