[Support] Linuxserver.io - TVHeadend

[DZMM]

34 minutes ago, saarg said:

You can see the bitrate of the stream in the status view of tvheadend so you can compare it to the 4k streams you have tried.

 

Have you tried using spmc instead of kodi?

bitrates shouldn't be a problem as some of my 4k content was 4/5x what I was seeing in TVH.

 

I'd never come across spmc before.  I installed that on my mi box 3 and it works perfectly with default settings, just like on the windows machine.  I've seen some posts about v17 TV problems and I think I've got the same problems.  I'll report back if I work out how to fix.

Thanks for being so patient.

[saarg]

Kodi doesn't have an android developer now, so support isn't the best. The SPMC guy used to do it, but left Kodi for various reasons. SPMC is Android only. 

[DZMM]

2 hours ago, saarg said:

Kodi doesn't have an android developer now, so support isn't the best. The SPMC guy used to do it, but left Kodi for various reasons. SPMC is Android only. 

Totally correct - my problems were all Android.  From SBMC:

 

https://github.com/koying/SPMC/issues/754#issuecomment-281215588

 

Quote

the new video player in 17 is ok in windows but horrible on android. if you watch live tv "mpeg/.ts" it constantly stutters and crashes.

but SPMC based on 16 is rock solid with no issues.

SPMC working brilliantly - if anyone else is having android client problems, this is probably the answer.  It also has some neat ANdroid TV support (recommendations, now playing) that Kryton doesn't.  

[s.Oliver]

hey guys!

 

back here with another oscam/usb-reader/docker/passthrough-question.

 

for a test-installation i used a pc which did exactly the same (= same hardware). software is newer alas latest unRAID 6.3.2 with current oscam docker.

passing through an omnikey 3132 usb-reader to oscam-docker with --device=/dev/bus/usb/00x/00x (x=actual numbers from host). when trying to use the reader, oscam segfaults, oscam log shows Omnikey [pcsc] Restarting reader / Omnikey [pcsc] creating thread for device 0 / Omnikey [pcsc] Cannot open device: 0

 

now when i compare the setup to a working machine with unRAID 6.2.4 and the same reader-hardware, i do see in the docker log as last line (and guessing here) that the reader got recognized = passed through successfully; it reads: [HID Global] HID Global OMNIKEY CCID driver x86_64 v4.2.8 [email protected]

 

this line is missing from docker log on the new machine setup.

 

i tried various usb-ports (even exact the same one, that once 6.2.4 ran on this same machine and same reader) worked beautifully.

it seems, that the extra line parameter has no effect. on the other hand, oscam shows under Readers / Scan USB that device with the correct numbers (as they are used for the passthrough).

 

well, has anything changed with 6.3.x versions of unRAID and usb-passthrough to dockers? i tried to find glues for it, but haven't.

 

thx. and hope you're all doin' fine. 

Edited March 28, 2017 by s.Oliver

[saarg]

I'll check here if it works on 6.3.x.

The only thing I can think of is incompatibility with the unraid kernel. 

There are no newer drivers for your reader. 

[s.Oliver]

45 minutes ago, saarg said:

I'll check here if it works on 6.3.x.

The only thing I can think of is incompatibility with the unraid kernel. 

There are no newer drivers for your reader. 

hmm... what else i could test? are there any unraid shell commands we could use to "talk" to the usb-reader and verify it isn't an incompatibility?

 

ok, does that last line in the docker log (when it has started up) which indicates it found a passed through device show up on every-bodies log as well?

 

thx. saarg, for being quick here.

Edited March 28, 2017 by s.Oliver

[s.Oliver]

which usb-readers for smartcards (encrypted ones for cable, satellite) do you guys successfully use on 6.3 onwards?

maybe (if it is really an kernel incompatibility) i could simply get another one, which is verified working by someone here.

 

thx.

[s.Oliver]

ok, i might found something...

 

despite having ordered the same model, despite it looking absolutely identical it seems to be different.

i could do a crosscheck a few minutes ago and discovered, that the 1st (which works on 6.2.4) also works on 6.3.2. the 2nd (new one) doesn't work on 6.2.4. so it's not a docker-/passthrough bug/problem.

 

they both have same name, only a slightly different part no. and the 1st is rev. A and the 2nd (new) is rev. B.

 

@saarg: but you checked/said there are no newer drivers out there... bummer.

 

how can they fuxor up this bad with a rev. B model?

 

any good advice on usb-smartcard readers for pay-tv cards from you guys?

Edited March 28, 2017 by s.Oliver

[s.Oliver]

1 hour ago, s.Oliver said:

how can they fuxor up this bad with a rev. B model?

1 hour ago, s.Oliver said:

ok, i might found something...

how can they fuxor up this bad with a rev. B model?

 

 

 

...from my retailer i got the following response (freely translated to english):

 

the new models of omnikey-readers have a new usb identifier (PID/VID).

for those models you'll need a recent driver. old versions (v3.9.0 and earlier) won't recognize those identifiers.

 

customers need to download a recent driver from our website, for linux these are 4.0.5.5. if linux user want to use CCID drivers, they also need to update. the new versions of the CCID drivers should recognize the new PID/VID from the readers firmware 2.04.

 

if there are no newer CCID drivers for your linux variant, which recognize these new PID/VID, it is possible to manually add them to the config.

you need to edit the file /etc/libccid_info.plist and add the new PID/VID to three corresponding places and add also the name of the reader.

 

we are encouraging you to try out our new driver version 4.0.5.5; get them here:

https://www.hidglobal.com/drivers/21278

for 64bit version (replace the last '8' with '7' to get the 32bit version)

 

@saarg: the link provides a file named 'ifdokccid_linux_x86_64-v4.0.5.5.tar.gz' (402,5 KB). but it looks like this driver is way older than what you implemented? there's a v4.2.8 listed from 08.06.2016

 

hmm… comparing what i see under Tools/System Devices:

 

Rev. A (which works):

ID 076b:3022 OmniKey AG CardMan 3021 (even though it is a 3121)

 

Rev. B (doesn't work):

ID 076b:3031 OmniKey AG

 

@saarg: could that help a bit to get that newer one also working?

 

thx. for any help. that machine is for a friend and it would be too bad to not get it working, just because of a cheap usb-reader. i already convinced him to get an unRAID system built, but now we need the final touches working.

[saarg]

3 hours ago, s.Oliver said:

 

 

...from my retailer i got the following response (freely translated to english):

 

the new models of omnikey-readers have a new usb identifier (PID/VID).

for those models you'll need a recent driver. old versions (v3.9.0 and earlier) won't recognize those identifiers.

 

customers need to download a recent driver from our website, for linux these are 4.0.5.5. if linux user want to use CCID drivers, they also need to update. the new versions of the CCID drivers should recognize the new PID/VID from the readers firmware 2.04.

 

if there are no newer CCID drivers for your linux variant, which recognize these new PID/VID, it is possible to manually add them to the config.

you need to edit the file /etc/libccid_info.plist and add the new PID/VID to three corresponding places and add also the name of the reader.

 

we are encouraging you to try out our new driver version 4.0.5.5; get them here:

https://www.hidglobal.com/drivers/21278

for 64bit version (replace the last '8' with '7' to get the 32bit version)

 

@saarg: the link provides a file named 'ifdokccid_linux_x86_64-v4.0.5.5.tar.gz' (402,5 KB). but it looks like this driver is way older than what you implemented? there's a v4.2.8 listed from 08.06.2016

 

hmm… comparing what i see under Tools/System Devices:

 

Rev. A (which works):

ID 076b:3022 OmniKey AG CardMan 3021 (even though it is a 3121)

 

Rev. B (doesn't work):

ID 076b:3031 OmniKey AG

 

@saarg: could that help a bit to get that newer one also working?

 

thx. for any help. that machine is for a friend and it would be too bad to not get it working, just because of a cheap usb-reader. i already convinced him to get an unRAID system built, but now we need the final touches working.

 

I'll patch the file to add the new reader. Until I get time to do it you can edit the file yourself to get it working.

To edit the file do the below. Each line is one command.

 

docker exec -it oscam bash
apk --update add nano
nano /usr/lib/pcsc/drivers/ifdokccid_linux_x86_64-v4.2.8.bundle/Contents/Info.plist

Add the reader to the list toghether with the other readers. Save the changes and restart the oscam container.

 

By the way, nice monologue you had 

[s.Oliver]

6 hours ago, saarg said:

 

I'll patch the file to add the new reader. Until I get time to do it you can edit the file yourself to get it working.

To edit the file do the below. Each line is one command.

 

docker exec -it oscam bash
apk --update add nano
nano /usr/lib/pcsc/drivers/ifdokccid_linux_x86_64-v4.2.8.bundle/Contents/Info.plist

Add the reader to the list toghether with the other readers. Save the changes and restart the oscam container.

 

By the way, nice monologue you had 

yah, thanks for that 

 

well, i tried adding it, but somewhere there seems to be still a small kind of problem.

 

starting up the docker after the change it's log shows one 'good' line as expected:

[HID Global] HID Global OMNIKEY CCID driver x86_64 v4.2.8 [email protected]

 

but then these errors:

00000000 readerfactory.c:1110:RFInitializeReader() Open Port 0x200000 Failed (usb:076b/3031:libudev:0:/dev/bus/usb/001/006)

00000008 readerfactory.c:375:RFAddReader() OMNIKEY CardMan (076B:3031) 3121 [OMNIKEY 3x21 Smart Card Reader] init failed.

00000009 hotplug_libudev.c:520:HPAddDevice() Failed adding USB device: OMNIKEY CardMan (076B:3031) 3121

 

so the changes made seem to get picked up, but... (inserted a good reason here) it won't accept/activate it properly.

 

got an idea?

 

Edited March 29, 2017 by s.Oliver

[saarg]

Sorry, but I got no other idea. Looks like hid need to update their driver. You could try contacting them and ask.

[s.Oliver]

1 hour ago, saarg said:

Sorry, but I got no other idea. Looks like hid need to update their driver. You could try contacting them and ask.

 

my translated text earlier, seemed to be an answer my retailer got from HID.

there the version 4.0.5.5 was recommended as the current latest driver (non-beta anyway), which he used/tested with debian wheezy (64bit) and confirms it working/recognizing the reader out-of-the-box.

 

so, i would like to ask, if you could use that version 4.0.5.5 (the link earlier posted) and incorporate it, instead of the one actually being included (beta 4.2.8).

maybe they screwed up the list of PID/VIDs in the beta driver.

 

thx. saarg

Edited March 29, 2017 by s.Oliver

[s.Oliver]

2 hours ago, s.Oliver said:

 

my translated text earlier, seemed to be an answer my retailer got from HID.

there the version 4.0.5.5 was recommended as the current latest driver (non-beta anyway), which he used/tested with debian wheezy (64bit) and confirms it working/recognizing the reader out-of-the-box.

 

so, i would like to ask, if you could use that version 4.0.5.5 (the link earlier posted) and incorporate it, instead of the one actually being included (beta 4.2.8).

maybe they screwed up the list of PID/VIDs in the beta driver.

 

thx. saarg

 

just had a look into the 4.0.5.5 files and couldn't even find that VID in there. so i guess they fuxored somewhere, something badly up.

i sent them an email and so we'll see.

 

thx. saarg for helping out this far.

[saarg]

You could try the non-beta version yourself by deleting the files from the beta driver and installing the non-beta version. There is an install script for it. It might be as easy as replacing the .so file. Not at home, so can't check until the evening. 

[s.Oliver]

28 minutes ago, saarg said:

You could try the non-beta version yourself by deleting the files from the beta driver and installing the non-beta version. There is an install script for it. It might be as easy as replacing the .so file. Not at home, so can't check until the evening. 

 

no problem, i can live with your time schedule and appreciating your effort.

thx. man.

 

do you just tune into FTA channels?

[saarg]

10 minutes ago, s.Oliver said:

 

no problem, i can live with your time schedule and appreciating your effort.

thx. man.

 

do you just tune into FTA channels?

 

I have a pcsc reader that doesn't need the hid driver . It's sim card size reader as that is what my provider use. 

[s.Oliver]

1 minute ago, saarg said:

 

I have a pcsc reader that doesn't need the hid driver . It's sim card size reader as that is what my provider use. 

 

i see. do maybe different models of your reader exist? (smartcard size)

[saarg]

6 minutes ago, s.Oliver said:

 

i see. do maybe different models of your reader exist? (smartcard size)

 

I don't know. I'm using Omnikey CardMan 6121. It doesn't look like there are a normal size version. 

 

If you are looking for another reader I would suggest you do not look for a pcsc reader. Instead go for a smartreader+, infinity USB smart or similar that doesn't need drivers. 

Edited March 29, 2017 by saarg

[s.Oliver]

2 hours ago, saarg said:

 

I don't know. I'm using Omnikey CardMan 6121. It doesn't look like there are a normal size version. 

 

If you are looking for another reader I would suggest you do not look for a pcsc reader. Instead go for a smartreader+, infinity USB smart or similar that doesn't need drivers. 

 

yah, that would have been my next question (hey, how do you pulled that one – how far do brain waves travel?) 

ok, will probably look for one of those mentioned ones. might ask you then, if you think/know if it's supported.

[s.Oliver]

found this one: "Smartmouse / Easymouse 2 USB Premium"

all talk is that most linux'es support it directly (has an FTDI chipset). protocols supported: smartmouse / phoenix – does smartmouse here mean smartreader (or smargo) in oscam?

 

what do you think of that one?

[saarg]

6 hours ago, s.Oliver said:

found this one: "Smartmouse / Easymouse 2 USB Premium"

all talk is that most linux'es support it directly (has an FTDI chipset). protocols supported: smartmouse / phoenix – does smartmouse here mean smartreader (or smargo) in oscam?

 

what do you think of that one?

 

Don't know about that one. The only ones I know about is the ones I mentioned.

Smartmouse/phoenix protocol is the mouse one in oscam if I remember correctly.

 

Regarding trying the old driver to see if it works, you can just copy the driver to the appdata folder of oscam and unpack it, then execute into the container and run the install file. I think it should delete the beta version. Then you have to add your reader ID to the plist file again. I don't think it will work though, but worth a try.

[s.Oliver]

3 hours ago, saarg said:

 

Don't know about that one. The only ones I know about is the ones I mentioned.

Smartmouse/phoenix protocol is the mouse one in oscam if I remember correctly.

 

Regarding trying the old driver to see if it works, you can just copy the driver to the appdata folder of oscam and unpack it, then execute into the container and run the install file. I think it should delete the beta version. Then you have to add your reader ID to the plist file again. I don't think it will work though, but worth a try.

 

alright thx. i'll try that one (could get it faster and cheaper), but if it doesn't work will try the infinity usb smart one.

[Zan]

On 3/12/2017 at 10:29 AM, saarg said:

 

I tested comskip and mkv and it's not working for some reason. I think it might be related to ffmpeg being updated when we moved the base image from Alpine 3.4 to 3.5.

I'll see if I manage to compile ffmpeg instead of using the alpine version. 

 

Hey Saarg, I've found that tvheadend with the default recording cache scheme (System) is grabbing memory and not releasing it until I assume when the recording ends. This has been causing my tvheadend docker to re-start. I've since change the cache scheme to Sync + Don't Keep. I was thinking this might be a good default option for you to set in the docker. What do you think?

 

That seems to be one of the reasons why my mkvs werent "seekable" in vlc (the mkvs were abruptly cut off when tvheadend crashes) - another issue seems to be the duration of the mkv but I'm not sure if it's then vlc that isn't good with large mkvs or maybe a tvheadend mkv corruption issue... Will keep investigating.

[saarg]

5 hours ago, Zan said:

 

Hey Saarg, I've found that tvheadend with the default recording cache scheme (System) is grabbing memory and not releasing it until I assume when the recording ends. This has been causing my tvheadend docker to re-start. I've since change the cache scheme to Sync + Don't Keep. I was thinking this might be a good default option for you to set in the docker. What do you think?

 

That seems to be one of the reasons why my mkvs werent "seekable" in vlc (the mkvs were abruptly cut off when tvheadend crashes) - another issue seems to be the duration of the mkv but I'm not sure if it's then vlc that isn't good with large mkvs or maybe a tvheadend mkv corruption issue... Will keep investigating.

 

I want as little to do with the config as possible and I'm not changing it for one person. If you don't have enough memory, you should set the correct option for your use case.