DVB Owners Discussion Forum - dvbowners.com > BDA Plug-in for Web Scheduler

Full Version: BDA Plug-in for Web Scheduler

DVB Owners Discussion Forum - dvbowners.com > Technical & Development Forums > WebScheduler Development

Pages: 1, 2, 3, 4

Spoonfed

May 6 2004, 07:07 PM

Null,

Seems your new version works better.

The brief tests ive done provide a file with actual data for all SD channels.

The channel scan also works,just doesn't provide the channel name.

So just to confirm this muxer will not handle HD or AC3?

Re the .bin file naming. I've acutally found an option in zoomplayer that allows me to playback captures while recording, even with .bin extention. However when WS has finished the capture the file stays as .bin, im guessing as the OS sees a file is reading at th time WS tries to rename.
No biggy i can it manually for now.

PS: DOH! I spoke to soon

While it did do a number of Nine captures even after other channel captures (tried one or two). I coppied the .dll (which i understand is the only difference from the non BDA version, maybe im wrong here) over the top of my normal WS version. It captured SBS fine and the others, but 2kb again for nine

Even going back to complete new 2.2.2 version downloaded just tonight 2kb still.

This is the output on nine at capture time.

(2004-05-06 19-12) test2 Nine.mpg
win32-DLL : Trying to Lock Channel F=191625 B=7 on card 1
Capture DLL lockChannel(191625) = 0 attempt:1
win32-DLL : Trying to Start Capture of Pid v = 513 a = 660 on card 1
win32-DLL-BDA : Changing Frequency 191625 bandwidth 7.
win32-DLL-BDA : Mapping Video PID 513
win32-DLL-BDA : Mapping Audio PID 660
win32-DLL-BDA : Signal Strength 83
win32-DLL-BDA : locked = 1 : Quality = 98

null_pointer

May 6 2004, 07:09 PM

Can you have a look at the output (command line) and see if the tuner is trying to tune more than once on channel nine.

No, the current setup will not do HD or AC3.

Spoonfed

May 6 2004, 07:43 PM

Not sure what u mean.

My previous post is the output in the Dos window.

Is it "Capture DLL lockChannel(191625) = 0 attempt:1"
where = 0 means? what 1 attempt?

Or if it is tuning more than once will the be mulitples of my complete "cut and paste" in previous post?

In which case, no, there is only ever one of these.

It seems to be OK with nine on the first start up of WS. Ie if you start it it seems now one can capture nine ok, then another channel, then nine again ok. Close WS and start it again and nine outputs 2kb files.

I just tested this again. Run it from .bat captured nine twice ok. Close WS and then run it as a service and nine capture 2kb file, ten is capturing fine as i type but.

I always run it as a service anyways, so only ever starts once on boot should shouldn't be a problem, but it is a bit of concern as it may hinder reliabity.

The only way to get nine working again is to reboot the PC and start WS again. I will let you know a little later if rebootng PC with WS as auto service allows it to work ok.

Also of note i have had WS throw up quick splash of text when starting it after closing it. Its like the normal DOS text when starting but i can see something written about execption or somethiing, it flashes up so fast and closes its impossible to read.
Next time it does it i'll manually open a DOS window and start it that way so if it does it i can give you the output.

EDIT: Of note, the BDA version seems to use noticable CPU.
Ie with the SDK version CPU was at Zero while capturing.
With the BDA java.exe (which is effectly the WS process) jumps up and down between 3 and 13 % cpu. No big deal as i can still play back files with FFDshow filtering chewing the CPU up, but it could be an issue for some right on the edge? Would it be the muxing causing this or the new drivers? (i haven't measured cpu with .ts output, which would i guess elliminate the muxing filter)

Spoonfed

May 6 2004, 08:48 PM

OK some more

I have found that this version of WS can loose control of its capture.

Ie a capture can be running, one can stop it, it appears stopped in the schedule list BUT the file keeps growing even after the scheduled end time.

The only way to stop this is to kill WS. I guess it would probably continue till the drive is full?

This is what brings about the WS start issue i mentioned early. Once the above happens (which i have yet to determine 100% why) on starting WS again the following happens

C:Program FilesWebScheduler>java -cp . -jar WebScheduler.jar
New DataStore Object created (2.2.2.0)
channels.obj found and loaded 5 channels
datachannelMap.store loaded.
Loading Users
1 Users added
Times.obj found and loaded.
Allowing 45 seconds for wakeup
win32-DLL : Process Attached
win32-DLL : Creating Waitable Timer
win32-DLL : Waitable Timer ID 0
win32-DLL : Creating Wakeup Watch Thread
win32-DLL : Wakeup Watch Thread Created
win32-DLL : Wakeup Thread Init
Setting wakeup time (2004 5 7 22 29 15)
win32-DLL : Setting Wake up Time (y2004 m5 d7 h22 m29 s15)
dataepg.store loaded.
Timer Thread: Created
Timer Thread: Started
Base server Exception: java.net.BindException: Address already in use: JVM_Bind
java.net.BindException: Address already in use: JVM_Bind
at java.net.PlainSocketImpl.socketBind(Native Method)
at java.net.PlainSocketImpl.bind(Unknown Source)
at java.net.ServerSocket.bind(Unknown Source)
at java.net.ServerSocket.<init>(Unknown Source)
at java.net.ServerSocket.<init>(Unknown Source)
at WebScheduler.main(WebScheduler.java:50)

The only way to the restart WS is a reboot. I guess if one could kill the java.exe process it may restart but Task Manager cannot kill it, so reboot is the only way i know of.

Any ideas what is going on here? Very strange behaviour, have never noticed such with the SDK WS versions.

EDIT:

I can confirm that nine is still very flaky. On fresh boot as service got ERROR in schedule window, 2nd attempt 2kb file. Then stop service run from DOS, captured fine. Try again as service, 2kb file. Stop service try from DOS, 2kb file grrrrrrrrrr

haha

Every time if 2kb file or actual capture same in DOS window
(2004-05-06 20-58) kjlj Nine.mpg
win32-DLL : Trying to Lock Channel F=191625 B=7 on card 1
Capture DLL lockChannel(191625) = 0 attempt:1
win32-DLL : Trying to Start Capture of Pid v = 513 a = 660 on card 1
win32-DLL-BDA : Changing Frequency 191625 bandwidth 7.
win32-DLL-BDA : Mapping Video PID 513
win32-DLL-BDA : Mapping Audio PID 660
win32-DLL-BDA : Signal Strength 83
win32-DLL-BDA : locked = 1 : Quality = 98

where to me it has locked the channel obtaining both the quality and signal strength, which makes no sense as to why its not captureing.

null_pointer

May 6 2004, 09:49 PM

Ok, the fist thing is, dont run it as a service, this is still to early in development to try to use in an automated fashion.

When you get a schedule that is having problems (2kb) files open graphedit and connect to the running graph, then save the graph and send it to me. Also do this for a capture that is working. I will look at them.

I can not reproduce any of the problems you are seeing so ther eis not a lot I can do at the moment until I can find a way to reproduce the problems.

Elroy

May 7 2004, 01:23 AM

G'day Shaun,

Just tested the latest version. There were a few things I wanted to highlight.

1. I can confirm spoonfed's findings of WS losing control of a capture. The 'recording continuing after I stopped it via WS' thing happened to me last night. Furthermore, I had one capture tonight that stopped itself. So something wierd is definately going on here.

2. I noticed that once a capture starts, in the output window I get lines saying 'skipping 180, skipping 2' etc. The numbers always go in the order 180, 2, 30, 2, 2, 1, 30, and then start from 180 again. Should this be happening? I can't remember seeing this with the SDK version.

3. Like spoonfed, I am getting .bin files produced. However, this only occurs if I try to capture the same program on multiple occasions (i.e. capture for two minutes, stop capture, try to capture same program again). I am only doing this for testing purposes. I assume it is producing .bin files as it doesn't want to/can't overwrite the .mpg produced after the first capture? If so, this shouldn't be a problem for captures of entire shows.

4. I got pixellation/blocks/corruption during a couple of the captures. These occurred for only a split second, but they were noticable.

5. Channel searching went fine. Still no channel names after the search though, just numbers (I know you already know about this).

6. Capturing channel nine is fine for me - I haven't had the problem that spoonfed is having. I am not currently running it as a service though.

Could it be the mux filter that is causing problems? I know this is probably a stupid suggestion, but I was not having any problems with the version that output .ts files. Is it worth trying the ShowShifter mux filter, or are we just going to have the same problems?

Cheers,

Elroy.

null_pointer

May 7 2004, 08:32 AM

QUOTE

1. I can confirm spoonfed's findings of WS losing control of a capture. The 'recording continuing after I stopped it via WS' thing happened to me last night. Furthermore, I had one capture tonight that stopped itself. So something wierd is definately going on here.

I will try to reproduce this with my machine.

QUOTE

2. I noticed that once a capture starts, in the output window I get lines saying 'skipping 180, skipping 2' etc. The numbers always go in the order 180, 2, 30, 2, 2, 1, 30, and then start from 180 again. Should this be happening? I can't remember seeing this with the SDK version.

This is probably from the Mux filter, I have no idea what it means.

QUOTE

3. Like spoonfed, I am getting .bin files produced. However, this only occurs if I try to capture the same program on multiple occasions (i.e. capture for two minutes, stop capture, try to capture same program again). I am only doing this for testing purposes. I assume it is producing .bin files as it doesn't want to/can't overwrite the .mpg produced after the first capture? If so, this shouldn't be a problem for captures of entire shows.

when a cap is in progress it captures to a BIN file, once the capture finishes it should rename the file to what is set in the server properties file. default mpg, for a bin file to be left for some reason the capture is not exiting cleanly.

QUOTE

4. I got pixellation/blocks/corruption during a couple of the captures. These occurred for only a split second, but they were noticable.

Could be related to mux but more likly it is something else.

QUOTE

5. Channel searching went fine. Still no channel names after the search though, just numbers (I know you already know about this).

Channel/Program scanning is going to be completly revamped to the TIF filter instead of the PSI filter so there is no need to test this for now.

QUOTE

6. Capturing channel nine is fine for me - I haven't had the problem that spoonfed is having. I am not currently running it as a service though.

Could it be the mux filter that is causing problems? I know this is probably a stupid suggestion, but I was not having any problems with the version that output .ts files. Is it worth trying the ShowShifter mux filter, or are we just going to have the same problems?

Yes, it vcould be the mux, or it could be something else, until I can reproduce the problems and see what is going on I do not know what is causing the problems.

Elroy

May 7 2004, 12:21 PM

QUOTE

Yes, it vcould be the mux, or it could be something else, until I can reproduce the problems and see what is going on I do not know what is causing the problems.

Any way I can help out with this at all? Sending output files, graphedit graphs, etc?

null_pointer

May 7 2004, 12:49 PM

I will test a bit over the weekend, I will try to duplicate the problems you guys are seeing, as soon as I have something to try I will post here.

If you can duplicate the problem with WS not stopping the capture etc and list a set of steps that can cause the problem that would be a big help.

Also if you can duplicate this problem with the MUX version try doing the same with the TS version and see if you get the same result, this will effectivly tell us if it is the Mux or not.

Spoonfed

May 7 2004, 01:47 PM

The "loss of control' issue seems somewhat random. It seems to also mainly occur on Nine. When it happens sometimes the "schedule" page does not update. Ie i will have "starting action" just sitting there and the file will be "growing" or in nines case 2kb. If i then end the action it usuall stops but sometims the action gets "stuck". Ie cannot edit, delete or anything to this action even on reboot. Even when WS is killed the .bin file created during the "loss of control" cannot be deleted until reboot.

I'll try and see what exactly causes all the issues, i'll test again in my other PC also, maybe i'll be more lucky

I don' think the 2kb nine issue is related to the muxer as i get it on both versions. Could it be something weird like nines frequence at 191625 where others are all ***500?

The "loss of control" issue may well be the muxer issue, but its very random and sometimes hard to reproduce, yet other times easy.

Spectrum

May 7 2004, 01:56 PM

Could you guys having problems with Nine try recording 9HD and 9 Guide instead of the main program for a while - I'm wondering if there's some problem with all of that additional interactive sports crap that Nine's got in the main program. Those extra components may be confusing the Network Provider filter.

Spectrum

Elroy

May 7 2004, 02:23 PM

QUOTE

Can only record SD stuff with the Mux filter, no HD or AC3. I don't think we get the 9 guide over here in Perth either (I've never seen it anyway) so I can't help with that sorry.

Like spoonfed said, could the nine frequnecy be the problem? Is their freq 191625 Oz wide?

Spoonfed

May 7 2004, 02:25 PM

I did try the guide once or twice failed also. The guide has a different VPID here but same APID.

Champion_R

May 7 2004, 04:57 PM

QUOTE

Be thankful we don't get the crappy guide. I'd prefer 14mbit HD and no guide as opposed to 12mbit HD and a crappy guide.

Spoonfed

May 7 2004, 05:02 PM

more info:
I've had 9 start twice now for me but the file size stopped at 140kb and the other test 4956kb. The caputre kept going as normal and the file was renamed as .mpg.

null_pointer

May 7 2004, 05:30 PM

For the file to be renamed to .mpg the capture thread exit notmally. I will start testing soon when I get home.

null_pointer

May 7 2004, 06:21 PM

Ok I can have reproduced the problem, that is a good start :-)

I think it is driver related, I have the following code:

CODE

HRESULT hr2;

    CComPtr <IBDA_Topology> bdaNetTop;

    hr2 = m_pTunerDevice.QueryInterface(&bdaNetTop);

if (FAILED(hr2))

{

printf("win32-DLL-BDA : Cannot Find IID_IBDA_Topology %xn", hr2);

return hr2;

    }

    printf("win32-DLL-BDA : Found IBDA_Topology interface %xn", hr2);

    CComPtr <IUnknown> unknown;

    hr2 = bdaNetTop->GetControlNode(0, 1, 1, &unknown);

if (FAILED(hr2))

{

printf("win32-DLL-BDA : Cannot GetControlNode %xn", hr2);

return hr2;

    }

    printf("win32-DLL-BDA : Got Control Node 1 %xn", hr2);

    CComPtr <IBDA_SignalStatistics> pSigStats;

    hr2 = unknown.QueryInterface(&pSigStats);

if (FAILED(hr2))

{

printf("win32-DLL-BDA : Cannot get IBDA_SignalStatistics %xn", hr2);

return hr2;

    }

    printf("win32-DLL-BDA : Found IBDA_SignalStatistics interface %xn", hr2);

    long quality = -1;

    hr2 = pSigStats->get_SignalQuality(&quality);

if (FAILED(hr2))

{

printf("win32-DLL-BDA : Error getting get_SignalQuality %xn", hr2);

return hr2;

    }

    printf("win32-DLL-BDA : Got Signal Quality : %dn", quality);

    BOOLEAN locked = FALSE;

    hr2 = pSigStats->get_SignalLocked(&locked);

if (FAILED(hr2))

{

printf("win32-DLL-BDA : Error getting get_SignalLocked %xn", hr2);

return hr2;

    }

    printf("win32-DLL-BDA : Got Signal Locked : %dn", locked);

    if(locked == FALSE)

    {

printf("win32-DLL-BDA : Signal Not Lockedn");

return -1;

    }

    printf("win32-DLL-BDA : About to leave RunGraph with code %xn", hr);

And normally it produces this when it is working:

CODE

win32-DLL : Trying to Lock Channel F=219500 B=7 on card 1

Capture DLL lockChannel(219500) = 0 attempt:1

About to start capture

win32-DLL : Trying to Start Capture of Pid v = 512 a = 650 on card 1

win32-DLL-BDA : Changing Frequency 219500 bandwidth 7.

win32-DLL-BDA : Mapping Video PID 512

win32-DLL-BDA : Mapping Audio PID 650

win32-DLL-BDA : Signal Strength 84

win32-DLL-BDA : Found IBDA_Topology interface 0

win32-DLL-BDA : Got Control Node 1 0

win32-DLL-BDA : Found IBDA_SignalStatistics interface 0

win32-DLL-BDA : Got Signal Quality : 98

win32-DLL-BDA : Got Signal Locked : 1

win32-DLL-BDA : About to leave RunGraph with code 0

Capture started with code : 0

But for some reason the call to get Quality locks up some times and I get the following:

CODE

At this point the java.exe is locked up, and I can not kill it off that is why I think this is a driver issue, it is a call the the driver that is locking up and windows can not kill it off.

What do you think Spectrum?

bionicdonkey

May 7 2004, 06:34 PM

null_pointer, i have the same problems. Digitalwatch crashes and can't be killed.

Spoonfed

May 7 2004, 06:34 PM

So im not just a complete spastic

hehe

I could image how hard it is to sort such issues without seeing it first hand.

Maybe the new bug fixed driver release spectrum will unleash soon will sort these issues (thats if its driver related).

I can confirm that the java.exe is freezing, and is unkillable. I've sometimes noticed the entry system get a bit flaky at this point with desktop icons disappearing and and generally unstable for a period.

Spectrum

May 8 2004, 01:13 PM

There was a loop bug in the code arrgh :x

http://home.swiftdsl.com.au/~spectrum/drivers.zip

Version 1.4.2 Fixes looping bug.

Spectrum

null_pointer

May 8 2004, 01:14 PM

OK, I have tested the new driver and it seems to have fixed the problem, try the new BDA driver (1.4.1) and the new version of WS (2.2.3.0) and see if it has fixed your lockup problem.

If it still locks up post the command line output of the last 30 or 40 line of output.

Spectrum

May 8 2004, 01:17 PM

I'd like to thank all you guys for being my guinea pigs. I'm just sorry the driver code isn't up to scratch.

Spectrum

null_pointer

May 8 2004, 01:18 PM

More than half the fun is finding the problems/bugs :-)

Spectrum

May 8 2004, 01:19 PM

Sure is - so everyone please be sure to upgrade to driver version 1.4.2

Spectrum

Spoonfed

May 8 2004, 01:44 PM

Ok WS 2.2.3 and new drivers

Still nine is a no go for me. (2kb file still)

Output on failed capture
(2004-05-08 13-26) nine Nine.mpg
Setting wakeup time (2014 5 8 13 25 35)
win32-DLL : Setting Wake up Time (y2014 m5 d8 h13 m25 s35)
win32-DLL : Trying to Lock Channel F=191625 B=7 on card 1
Capture DLL lockChannel(191625) = 0 attempt:1
About to start capture
win32-DLL : Trying to Start Capture of Pid v = 513 a = 660 on card 1
win32-DLL-BDA : Changing Frequency 191625 bandwidth 7.
win32-DLL-BDA : Mapping Video PID 513
win32-DLL-BDA : Mapping Audio PID 660
win32-DLL-BDA : Signal Strength 82
win32-DLL-BDA : Found IBDA_Topology interface 0
win32-DLL-BDA : Got Control Node 1 0
win32-DLL-BDA : Found IBDA_SignalStatistics interface 0
win32-DLL-BDA : Got Signal Quality : 98
win32-DLL-BDA : Got Signal Locked : 1
win32-DLL-BDA : About to leave RunGraph with code 0
Capture started with code : 0

Same for "nineguide" station. All other channels work fine.

As requested the graph in action for each. As you can see they are identical, same for the above outputed WS info.

http://members.optusnet.com.au/spoonfed/NI...AILED%202KB.pdf
http://members.optusnet.com.au/spoonfed/SB...BS%20WORKED.pdf

Maybe it just don't like me. I think im going have to move the card into the other PC again. Any ideas what may be causing an issue in the current config, ie conflicts. I would think this would apply in general but, not just to one channel/network?

Null, do you think having a certain "other" card :wink: :wink: in the same system could cause an issue? I was going to pull it out anyways as i need a damn PCI bay plate thing (computer is completely full haha).

http://members.optusnet.com.au/spoonfed/HPIM0256.JPG

Spectrum

May 8 2004, 02:05 PM

I can assure you that if you have a lock and the signal is present that the driver is delivering data. I have seen exactly the same thing on my HTPC and I think it's either a problem with nine's stream or one of the Microsoft filters. I had a good run last weekend when Nine was not showing the demo HD loop during the AFL. Later when the HD demo loop was running again the problem came back.

There are probably many variables, but my guess is something is confusing the Network Provider, or the MPEG-2 Demultiplexer, or the TIF filter or ... well it's anyone's guess.

Does this problem occur when you capture the full TS or only the remuxed Video/Audio?

Spectrum

JoeyBloggs

May 8 2004, 02:16 PM

Hey, I think you've done quite well for a first release. They installed and worked on a variety of hardware and o/s's

I'm sure between us we will quickly get the bugs worked out.

Regards
Joey

null_pointer

May 8 2004, 03:19 PM

I was mainly conserned about the lockup problem, now it looks like that is fixed with the new driver we can start to look at the nine problem.

I need the actual GRF files, when you connect to the running gaph with graphedit just select save as and save the GRF file for a working channel nine capture and a none working one.

This will save all the mapped pid info from the mpegDemux filter as well and I can see if the correct pids are being mapped etc. Just email then to me.

Spoonfed

May 8 2004, 03:25 PM

Null,

Ok, i just got home and have to go out again, so, errrm maybe late tonight.

The nine issue does occur with the .ts version of WS. I haven't tested the .ts version with the latest drivers yet but.

Champion_R

May 8 2004, 03:44 PM

I've just installed the Spectrum BDA driver 1.42. Graphedit correctly displayed Channel Ten. I then installed Webscheduler BDA 2.2.0.0 TS version. I've sucessfully recorded all networks. Nine HD and SD both work. I am yet to experience this "Nine" problem.

Edit: Seven SD and ED tested and work, plus Ten SD and HD.

Can't wait for Bionicdonkey to release a build of DW that works with the BDA drivers. They are so much better.

captainatomic

May 8 2004, 05:23 PM

I have just had a go at WS with BDA. Seemed well until this:

CODE

An unexpected exception has been detected in native code outside the VM.

Unexpected Signal : EXCEPTION_ACCESS_VIOLATION (0xc0000005) occurred at PC=0x2C6

44EA

Function=getFreeSpace+0x2CDA

Library=C:DVBWebSchedulerCaptureEngine.dll

Current Java thread:

at DllWrapper.startCapture(Native Method)

at CaptureThread.run(CaptureThread.java:141)

at java.lang.Thread.run(Unknown Source)

Dynamic libraries:

0x00400000 - 0x00406000 C:WINDOWSsystem32java.exe

0x77F50000 - 0x77FF7000 C:WINDOWSSystem32ntdll.dll

0x77E60000 - 0x77F46000 C:WINDOWSsystem32kernel32.dll

0x77DD0000 - 0x77E5D000 C:WINDOWSsystem32ADVAPI32.dll

0x78000000 - 0x78087000 C:WINDOWSsystem32RPCRT4.dll

0x77C10000 - 0x77C63000 C:WINDOWSsystem32MSVCRT.dll

0x08000000 - 0x08136000 C:Program FilesJavaj2re1.4.2_01binclientjv

m.dll

0x77D40000 - 0x77DCC000 C:WINDOWSsystem32USER32.dll

0x7E090000 - 0x7E0D1000 C:WINDOWSsystem32GDI32.dll

0x76B40000 - 0x76B6C000 C:WINDOWSsystem32WINMM.dll

0x10000000 - 0x10007000 C:Program FilesJavaj2re1.4.2_01binhpi.dll

0x00390000 - 0x0039E000 C:Program FilesJavaj2re1.4.2_01binverify.dl

l

0x003A0000 - 0x003B8000 C:Program FilesJavaj2re1.4.2_01binjava.dll

0x003C0000 - 0x003CD000 C:Program FilesJavaj2re1.4.2_01binzip.dll

0x02C60000 - 0x02C78000 C:DVBWebSchedulerCaptureEngine.dll

0x771B0000 - 0x772D4000 C:WINDOWSsystem32ole32.dll

0x77120000 - 0x771AB000 C:WINDOWSsystem32OLEAUT32.dll

0x02D00000 - 0x02D0F000 C:Program FilesJavaj2re1.4.2_01binnet.dll

0x71AB0000 - 0x71AC5000 C:WINDOWSsystem32WS2_32.dll

0x71AA0000 - 0x71AA8000 C:WINDOWSsystem32WS2HELP.dll

0x71A50000 - 0x71A8B000 C:WINDOWSsystem32mswsock.dll

0x71A90000 - 0x71A98000 C:WINDOWSSystem32wshtcpip.dll

0x7C890000 - 0x7C911000 C:WINDOWSsystem32CLBCATQ.DLL

0x77050000 - 0x77115000 C:WINDOWSsystem32COMRes.dll

0x77C00000 - 0x77C07000 C:WINDOWSsystem32VERSION.dll

0x35500000 - 0x3565F000 C:WINDOWSsystem32quartz.dll

0x5AD70000 - 0x5ADA4000 C:WINDOWSSystem32uxtheme.dll

0x35800000 - 0x35951000 C:WINDOWSsystem32msvidctl.dll

0x03010000 - 0x0301F000 C:WINDOWSSystem32MsDvbNp.ax

0x35680000 - 0x356A3000 C:WINDOWSSystem32devenum.dll

0x76670000 - 0x76757000 C:WINDOWSsystem32setupapi.dll

0x03060000 - 0x03067000 C:WINDOWSsystem32msdmo.dll

0x60F00000 - 0x60F20000 C:WINDOWSsystem32ksproxy.ax

0x5EF80000 - 0x5EF84000 C:WINDOWSsystem32ksuser.dll

0x03070000 - 0x03077000 C:WINDOWSSystem32BdaPlgin.ax

0x1FF00000 - 0x1FF24000 C:Program FilesDigivision HTPCFiltersmpg2spl

t.ax

0x1C200000 - 0x1C241000 C:WINDOWSsystem32qcap.dll

0x73BD0000 - 0x73BF0000 C:WINDOWSsystem32MSVFW32.dll

0x773D0000 - 0x77BC2000 C:WINDOWSsystem32SHELL32.dll

0x70A70000 - 0x70AD5000 C:WINDOWSsystem32SHLWAPI.dll

0x77340000 - 0x773CB000 C:WINDOWSsystem32COMCTL32.dll

0x71950000 - 0x71A34000 C:WINDOWSWinSxSx86_Microsoft.Windows.Common-C

ontrols_6595b64144ccf1df_6.0.0.0_x-ww_1382d70acomctl32.dll

0x03180000 - 0x0318A000 C:WINDOWSSystem32PsisRndr.ax

0x031A0000 - 0x031FB000 C:WINDOWSSystem32PsisDecd.dll

0x1C400000 - 0x1C426000 C:Program FilesV-Stream MultimediaDVB PlusDT

VRMpgMux.ax

0x03350000 - 0x0336B000 C:DVBWebSchedulerfiltersdump.ax

0x76C90000 - 0x76CB2000 C:WINDOWSsystem32imagehlp.dll

0x6D510000 - 0x6D58D000 C:WINDOWSsystem32DBGHELP.dll

0x76BF0000 - 0x76BFB000 C:WINDOWSsystem32PSAPI.DLL

Heap at VM Abort:

Heap

def new generation total 576K, used 419K [0x10010000, 0x100b0000, 0x104f0000)

eden space 512K, 78% used [0x10010000, 0x100745d8, 0x10090000)

from space 64K, 27% used [0x10090000, 0x10094650, 0x100a0000)

to space 64K, 0% used [0x100a0000, 0x100a0000, 0x100b0000)

tenured generation total 1408K, used 290K [0x104f0000, 0x10650000, 0x14010000

)

the space 1408K, 20% used [0x104f0000, 0x10538868, 0x10538a00, 0x10650000)

compacting perm gen total 4096K, used 2153K [0x14010000, 0x14410000, 0x1801000

0)

the space 4096K, 52% used [0x14010000, 0x1422a408, 0x1422a600, 0x14410000)

Local Time = Sat May 08 17:20:03 2004

Elapsed Time = 934

#

# The exception above was detected in native code outside the VM

#

# Java VM: Java HotSpot(TM) Client VM (1.4.2_01-b06 mixed mode)

#

# An error report file has been saved as hs_err_pid3372.log.

# Please refer to the file for further information.

#

What does that mean? Is this of any use to shaun?

I kept the log file if ya'll need it for some de-bugging

null_pointer

May 8 2004, 05:35 PM

Hmm, thats interesting, the Function getFreeSpace has nothing to do with the new version, this function has not changes in months.

Elroy

May 8 2004, 07:14 PM

G'day Shaun,

Just did a short test with the latest version of both the drivers and WS. I got good recordings on all channels tested. I had one recording where I got blockiness/corruption like I did previously. Again, it was only for a second or two. Other than that, all went well. I've set up a schedule to record Pulp Fiction tonight, will report back tomorrow on how it goes.

Champion_R, are you sure you recorded Ten HD? Shaun has said previously that the mux filter he is using doesn't support AC3 audio or HD stuff.

Cheers.

Champion_R

May 8 2004, 07:40 PM

QUOTE

Champion_R, are you sure you recorded Ten HD? Shaun has said previously that the mux filter he is using doesn't support AC3 audio or HD stuff.

Cheers.

I'm using the version which records to non compliant TS files. It's best for me.

null_pointer

May 8 2004, 08:43 PM

New Version : 2.2.4.0

New version that uses the TIF filter to scan for programs (gives program names etc) the PSI filter is no longer needed and has been removed from the zip.

Still can not detect AC3 but working on it.

Spectrum

May 8 2004, 09:54 PM

Wow, it's really working great now.

Spectrum

captainatomic

May 8 2004, 11:01 PM

How can I select which card to use? I have 2 BDA cards installed and it always records from the first card - how do I select the second?

I have it set to ' NumberOfCards = 2' but cant see an option to select which card to record from.

If this isnt already a feature can you please intergrate it?

Thanks,

Jeremy Barr

null_pointer

May 8 2004, 11:17 PM

At the moment there has been no multi card support implementation or testing done at all, the aim is to get it working with one BDA card and then move on to multiple. With the limited man power that is currently assigned to this project i.e. me I can only do so much at one time.

Multi card support will be looked at later, when I do not know exactly, but it will be look at.

EDIT:

Currently the code should at least try to use or create 2 or more capture Graphs (depending on what you have "number of cards" set to). This means that currently it may attempt and may work with multiple cards, what happens if you have two schedules at the same time etc do you get an error?

Spoonfed

May 8 2004, 11:21 PM

geeeee doesn't the the "Null" support "hypethreading" or "dual cpu", time for an upgrade perhaps?

haha

very understandable, no doubt this "project" takes a fair chunk of your spare time up as it is. shoot some code this way and i'll do a "google" for a java to engrish translator and get to work

haha

I tried two cards and the first capture worked on the first BDA card but the 2nd capture gave "ERROR" on the schedule page trying to use the second card

captainatomic

May 8 2004, 11:28 PM

Spoonfed - We have something in common :wink:

I dont want 'multi cards' ie recording from 2 tuners - I just want the option to select which card it records from. Is that hard to implement?

kenneth

May 8 2004, 11:49 PM

Ignore this Post.

nate

May 9 2004, 09:18 AM

After doing a few captures with WS-BDA last night and having it work perfectly i sheduled a few to run overnight. They all worked fine, except one. It was a 15 minute shedule that only made a capture of about 1 and a half minutes.

There's no errors in any of the logs so i don't think it's WebScheduler's fault. The resulting file has heaps of glitches in it, but the log has the same signal strength (78) and quality (98) as every other capture i've made that only have about 1 glitch per minute.

I'm guessing that even though the signal quality and strength were reported as fine that there was a problem with reception that the mux filter couldn't cope with.

I think what spoonfed suggested in the Other thread would be useful to determine situations like this.

Also, null while we're still testing i'm wondering if it might be useful to have WS record two files. one from the mux filter, and the other as the non-standard TS stream. That way we can determine for sure if the mux filter is causing any problems.

null_pointer

May 9 2004, 09:35 AM

QUOTE

I think what spoonfed suggested in the Other thread would be useful to determine situations like this.

I am not sure if polling the tuner is going to help, if there is a glitch that is causing the demux problems then I am not sure a 10 or 15 second poll of the signal lock is going to help. Also what do you do if you get a signal drop, stop the capture? there is not a lot that can be done in this situation.

QUOTE

Also, null while we're still testing i'm wondering if it might be useful to have WS record two files. one from the mux filter, and the other as the non-standard TS stream. That way we can determine for sure if the mux filter is causing any problems.

Yes, that way we can work out if it is the demux of not, I will add this later today and upload the test.

null_pointer

May 9 2004, 09:40 AM

Spoonfed, I could not send you mail :-(

QUOTE

<XXXXXX@optusnet.com.au>: host mail.optusnet.com.au[211.29.132.250]said: 554
5.7.1 Rejected 203.113.207.159 is a dynamic IP (in reply to RCPT TO command)

I use ADSL.

So here is my reply:

I can not load the GRF files, I get an error "Device not ready" I can
create them here and load them but for some reason can not load yours.
Can you load them? If you can can you look in the MS Mpeg2Demux filter
and on the properties page have a look at the mappings and see if the
correct pids are mapped etc, that is what I wanted to see.

Does the output you included in the email include the time it lost
control?

When WS loses control is java.exe locking up? it is not that WS is
losing control it is that it is crashing and locking, well that is
what I was seeing with the 1.4 driver anyway. Can you confirm that
Java.exe is locked and can not be killed. The best way to do this is
try to view the EPG when it gets into this state.

What driver are you using? you should be using BDA 1.4.3

nate

May 9 2004, 09:41 AM

QUOTE

It's not going to give a better capture, but at least then you could log the drop in signal, and maybe even detect if the capture file doesn't grow for 10 seconds or somthing then stop and restart.

null_pointer

May 9 2004, 10:33 AM

There is a test CaptureEngine.dll available that will capture the complete TS stream as well as the remuxed pids you want etc.

Just drop it in the WS directory over the top of the one that is there.

http://www.digtv.ws/files/dvb/BDA/CaptureEngine.dll

It will capture the TS to a file with the same name as the mpg but with .ts appended to the end.

Spoonfed

May 9 2004, 04:50 PM

Nate,

I have also experienced the issue of nine stopping capture. Well the capture continues but the file stops "growing". It is at this time where usually if i try to "stop" this capture i am left with scheduled task "stuck" in the schedule page, unable to "stop" or "delete". Yet i can start a new working schedule immediately after and it works fine (as long as its not nine).

It seems that it is possible for most of us to have the "nine issue" the level of the issue seems to vary. Ie being stupid its worse for me

Well usually i only get 2kb file, sometimes it goes till i stop it (havent' tried for a extend period but) and sometimes the file "growth" simply stops on its own (as nate mentioned). The variables that change these occurances could be what? Location (im Brisbane)? PC harware? Level of XP (ie i run no Service Pack, scared of breaking it). Though all this does not really make sense given the other channels/networks work fine, which makes on think nine is doing something different. Which is NO supprise as their 2ch AC3 on the HD channel (fine on SD) cannot be decoded by my receiver yet ALL other AC3 can. Maybe nine are trying to be like Sony

Null, i have emailed you PDF's of my graphs and a "screen cap" of the PID info i think you are chasing. (different email address so you should be able to reply)

Re the reply email issue. It seems like you run your own mail server and Optarse thinks its a dynamic IP or something and hence blocks it. Stupid Optarse, they make Bigpond look good, worst mail restrictions of ANY ISP i know.

PS: I just did a 355mb capture from nine (10min) and it worked fine, tried again straight after, 2kb.

null_pointer

May 9 2004, 05:36 PM

Did you try the above test dll, does the TS get captures ok when a nine capture is having problems?

Spoonfed

May 9 2004, 06:01 PM

Ok, some progress.

My first capture of nine output both "growing" data to the .bin (mpg) and .ts files. However i could not stop this capture (as i mentioned in email) using WS "stop" and had to close WS.

I then try SBS and the capture worked fine with both the .bin and .ts "growing" with data (gotta love my technical terms)

Trying nine again, the .bin file stayed at 2kb and the .ts "grew" as it should. Again i had to close WS to stop the capture.

I then tested on all other networks. All but nine captured both the .ts .bin streams.

HOWEVER interestingly ABC has the same control issue as nine, where i have to close WS to "stop" the capture and the entry is then "stuck" in the schedule list.

So it seems you theory is correct. There is something with the nine stream that the demuxing does not like, yet the data is "flowing" as the complete network can be dumped to the drive (boy does that file grow fast

) Is there a newer demuxing filter (what is the current on in use? the Microsoft one, perhaps its a similar issue to DW's "pausing frame" issue?) Is there a newer on available? Perhaps service pack 1 installs a newer filter that can handle what ever the issue is better? (i have no service pack) Or am i just dribbling again

OR quite possibliy nine brisbane is a little different and introducing more issues due to this? (anyone else testing in brisbane?)

The "control" issue is something different again i think. I seems more prevelant in 2.2.4 than the previous versions if that helps.

null_pointer

May 9 2004, 06:12 PM

Have a look in graphedit for the following DirectShow filter:

MPEG-2 Demultiplexor

In graphedit expand the filter tree entry and look for the filename it should point to a file in the following location, make sure it does.

C:WINDOWSsystem32mpg2splt.ax

This is the Mpeg2Demux filter, my version of the file (right click on the file and look in version) is 6.5.1.900 what is yours?

This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.