Comments for www.KernelLabs.com http://www.kernellabs.com/blog ... it's not just another Linux blog Sat, 12 May 2012 14:48:42 +0000 hourly 1 http://wordpress.org/?v=3.3.2 Comment on HVR-1800/1850 Analog support by Steven http://www.kernellabs.com/blog/?p=1636&cpage=1#comment-2513 Steven Sat, 12 May 2012 14:48:42 +0000 http://www.kernellabs.com/blog/?p=1636#comment-2513 tekdoc: Thanks for the feedback. Devin plans to look at my original patches. He picked up my HVR1850 sample hardware last week so, time permitting, we should get back to fixes in the near future. Stay tuned. tekdoc: Thanks for the feedback.

Devin plans to look at my original patches. He picked up my HVR1850 sample hardware last week so, time permitting, we should get back to fixes in the near future.

Stay tuned.

]]> Comment on HVR-1800/1850 Analog support by tekdoc http://www.kernellabs.com/blog/?p=1636&cpage=1#comment-2512 tekdoc Sat, 12 May 2012 14:36:07 +0000 http://www.kernellabs.com/blog/?p=1636#comment-2512 Just an update on my HVR-1800 analog support testing. Please note, I am primarily concerned with using my HVR-1800 with MythTv so please keep that in mind as you follow along. Earlier this week Devin and I discussed how in MythTv, the default image size setting of 480x480 (under recording profiles) triggers a driver bug that causes distorted colors. Changing the image size to 720x480 resolves this problem. This testing was done WITHOUT the January 2012 cx23885/cx25840 patches applied to the 3.2 and 3.3 kernels. I subsequently tested WITH the January 2012 cx23885/cx25840 patches applied to the 3.2 and 3.3 kernels and found several problems. First and foremost, one of the patches causes color distortion in ANY program that can only be reversed by reverting to an earlier version of the driver without the patches applied. Next, one of the patches prevents MythTv from accessing the mpeg device (/dev/video1 on my system) so live TV cannot be displayed at all. I am still working on determining exactly which patches trigger these two bugs. Lastly, the patch "[media] cx25840 / cx23885: Fixing audio/volume regression" causes NO AUDIO with the HVR-1800. With this patch REVERTED, the volume is so loud that the audio is actually clipping. I have been able to work around both issues by modifying the driver to force the volume register to a specific, fixed value. Interestingly, I found that increasing this value decreases the volume and vice versa. Since I control the volume of my system with my receiver, I am OK with having the volume level fixed. In order to have an adjustable volume level, someone would have to figure out the correct usable range for this value and modify the driver to reflect this. So in summary, with some modifications, the HVR-1800 can be used today with analog in MythTv. But If the latest patches are important for providing access to HVR-1800 features or performance improvements, we need to keep debugging these drivers. Just an update on my HVR-1800 analog support testing. Please note, I am primarily concerned with using my HVR-1800 with MythTv so please keep that in mind as you follow along.

Earlier this week Devin and I discussed how in MythTv, the default image size setting of 480×480 (under recording profiles) triggers a driver bug that causes distorted colors. Changing the image size to 720×480 resolves this problem. This testing was done WITHOUT the January 2012 cx23885/cx25840 patches applied to the 3.2 and 3.3 kernels.

I subsequently tested WITH the January 2012 cx23885/cx25840 patches applied to the 3.2 and 3.3 kernels and found several problems. First and foremost, one of the patches causes color distortion in ANY program that can only be reversed by reverting to an earlier version of the driver without the patches applied. Next, one of the patches prevents MythTv from accessing the mpeg device (/dev/video1 on my system) so live TV cannot be displayed at all. I am still working on determining exactly which patches trigger these two bugs.

Lastly, the patch “[media] cx25840 / cx23885: Fixing audio/volume regression” causes NO AUDIO with the HVR-1800. With this patch REVERTED, the volume is so loud that the audio is actually clipping. I have been able to work around both issues by modifying the driver to force the volume register to a specific, fixed value. Interestingly, I found that increasing this value decreases the volume and vice versa. Since I control the volume of my system with my receiver, I am OK with having the volume level fixed. In order to have an adjustable volume level, someone would have to figure out the correct usable range for this value and modify the driver to reflect this.

So in summary, with some modifications, the HVR-1800 can be used today with analog in MythTv. But If the latest patches are important for providing access to HVR-1800 features or performance improvements, we need to keep debugging these drivers.

]]> Comment on 4th of July: PCTV 80e Support by labarowski http://www.kernellabs.com/blog/?p=1435&cpage=1#comment-2511 labarowski Tue, 08 May 2012 00:06:15 +0000 http://www.kernellabs.com/blog/?p=1435#comment-2511 Thanks for your work on all these articles! I found that link very informative. I also looked at your HVR-1600 Performance and Reverse Engineering article and found it very interesting. I have a somewhat limited understanding of hardware from classes and my personal studies but I'm afraid drivers are a bit beyond me. I have enough trouble getting pre-made programs to compile under Linux, so I should probably start small. Nevertheless, great work on all of this and thanks for putting the information out there for others. I'll definitely keep an eye on this blog as I get time. Thanks for your work on all these articles! I found that link very informative. I also looked at your HVR-1600 Performance and Reverse Engineering article and found it very interesting. I have a somewhat limited understanding of hardware from classes and my personal studies but I’m afraid drivers are a bit beyond me. I have enough trouble getting pre-made programs to compile under Linux, so I should probably start small. Nevertheless, great work on all of this and thanks for putting the information out there for others. I’ll definitely keep an eye on this blog as I get time.

]]> Comment on 4th of July: PCTV 80e Support by Devin Heitmueller http://www.kernellabs.com/blog/?p=1435&cpage=1#comment-2510 Devin Heitmueller Mon, 07 May 2012 20:40:59 +0000 http://www.kernellabs.com/blog/?p=1435#comment-2510 Hi Labarowski, Yeah, different USB bridges support different functionality depending on the use case. The following article describes the details of how these devices are setup in more detail: http://www.kernellabs.com/blog/?p=1045 Regarding supporting your device, the core elements for supporting such a device are present, but the work tends to be more than just a USB VID/PID. You would need to know or be able to determine the GPIO layout, identify the correct IF and other tuning parameters related to the chip setup, and there's the possibility that the version of the driver for the drx-j doesn't actually match the silicon that you're using. Devin Hi Labarowski,

Yeah, different USB bridges support different functionality depending on the use case. The following article describes the details of how these devices are setup in more detail:

http://www.kernellabs.com/blog/?p=1045

Regarding supporting your device, the core elements for supporting such a device are present, but the work tends to be more than just a USB VID/PID. You would need to know or be able to determine the GPIO layout, identify the correct IF and other tuning parameters related to the chip setup, and there’s the possibility that the version of the driver for the drx-j doesn’t actually match the silicon that you’re using.

Devin

]]> Comment on 4th of July: PCTV 80e Support by labarowski http://www.kernellabs.com/blog/?p=1435&cpage=1#comment-2509 labarowski Mon, 07 May 2012 20:15:25 +0000 http://www.kernellabs.com/blog/?p=1435#comment-2509 Thanks Devin. I was under the impression that the usb bridge was a more passive device that just passed on whatever information was given it. It's not terribly surprising that this isn't the case though. This explains why the company didn't bother to implement fm mode in this version. Out of curiosity's sake more than anything, what would be necessary to get the ATSC mode working using these drivers? Would it be as simple as changing a vendor or product id somewhere in the driver source or would much more in depth changes be necessary? Thanks Devin. I was under the impression that the usb bridge was a more passive device that just passed on whatever information was given it. It’s not terribly surprising that this isn’t the case though. This explains why the company didn’t bother to implement fm mode in this version. Out of curiosity’s sake more than anything, what would be necessary to get the ATSC mode working using these drivers? Would it be as simple as changing a vendor or product id somewhere in the driver source or would much more in depth changes be necessary?

]]> Comment on 4th of July: PCTV 80e Support by Devin Heitmueller http://www.kernellabs.com/blog/?p=1435&cpage=1#comment-2508 Devin Heitmueller Mon, 07 May 2012 15:57:32 +0000 http://www.kernellabs.com/blog/?p=1435#comment-2508 Hello Labarowski, That device will not support FM. The em2874 bridge chip only supports MPEG transport streams and has no functionality for uncompressed audio. Indeed the 18271 supports FM, as well as some variants of the drx-j chip. However you need all three links in the chain to support it (the tuner, demodulator, and bridge) in order for it to work. Devin Hello Labarowski,

That device will not support FM. The em2874 bridge chip only supports MPEG transport streams and has no functionality for uncompressed audio. Indeed the 18271 supports FM, as well as some variants of the drx-j chip. However you need all three links in the chain to support it (the tuner, demodulator, and bridge) in order for it to work.

Devin

]]> Comment on 4th of July: PCTV 80e Support by labarowski http://www.kernellabs.com/blog/?p=1435&cpage=1#comment-2507 labarowski Sat, 05 May 2012 21:57:39 +0000 http://www.kernellabs.com/blog/?p=1435#comment-2507 http://www.reddit.com/r/RTLSDR/comments/r5ilx/does_the_atsc_version_of_the_ezcap_tuner_have_the/ So that's the thread that I was referring to. Guess I can't post links html style. http://www.reddit.com/r/RTLSDR/comments/r5ilx/does_the_atsc_version_of_the_ezcap_tuner_have_the/

So that’s the thread that I was referring to. Guess I can’t post links html style.

]]> Comment on 4th of July: PCTV 80e Support by labarowski http://www.kernellabs.com/blog/?p=1435&cpage=1#comment-2506 labarowski Sat, 05 May 2012 21:55:59 +0000 http://www.kernellabs.com/blog/?p=1435#comment-2506 Devin, Thanks for your hard work on this, It's sad to see so many people who think this was posted just to gloat. To cut to the chase, I'm really not very advanced with linux but I did manage to get these drivers to compile. However, they don't seem to work with my particular Trident based card. I've put some information which I've found on the card below. I suspect that the FM function of this card may pass I/Q samples directly making it a good little Software Defined Radio device. Check this thread out as well as it's parent reddit for more info. If you have any interest, I'm sure your help would be appreciated. ezcap EzTV306_1.2 Trident DRX3933J_B2 (demodulator) NXP TDA18271HDC2 Tuner Empia em2874B usb bridge https://www.box.com/s/cf09c4a217274d663127 WHat would I need to do, beyond the work already done on these drivers, to get this device running, especially in fm mode? Thanks! Devin,
Thanks for your hard work on this, It’s sad to see so many people who think this was posted just to gloat. To cut to the chase, I’m really not very advanced with linux but I did manage to get these drivers to compile. However, they don’t seem to work with my particular Trident based card. I’ve put some information which I’ve found on the card below. I suspect that the FM function of this card may pass I/Q samples directly making it a good little Software Defined Radio device. Check this thread out as well as it’s parent reddit for more info. If you have any interest, I’m sure your help would be appreciated.

ezcap EzTV306_1.2
Trident DRX3933J_B2 (demodulator)
NXP TDA18271HDC2 Tuner
Empia em2874B usb bridge

https://www.box.com/s/cf09c4a217274d663127

WHat would I need to do, beyond the work already done on these drivers, to get this device running, especially in fm mode? Thanks!

]]> Comment on HVR-1800/1850 Analog support by Britney http://www.kernellabs.com/blog/?p=1636&cpage=1#comment-2504 Britney Tue, 17 Apr 2012 15:50:32 +0000 http://www.kernellabs.com/blog/?p=1636#comment-2504 I have been unable to get analog to work on my HVR-1800. I built linux-media about 2 weeks ago. dmesg shows the firmware is loaded and the mpeg encoder is at /dev/video3: [ 5.849706] cx23885 driver version 0.0.3 loaded [ 5.849787] cx23885 0000:02:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 [ 5.849932] CORE cx23885[0]: subsystem: 0070:7809, board: Hauppauge WinTV-HVR1800 [card=2,autodetected] [ 6.239386] cx23885[0]: hauppauge eeprom: model=78631 [ 6.289124] cx25840 8-0044: cx23887 A/V decoder found @ 0x88 (cx23885[0]) [ 6.983004] cx25840 8-0044: loaded v4l-cx23885-avcore-01.fw firmware (16382 bytes) [ 9.670505] cx23885[0]: registered device video2 [v4l2] [ 9.670527] cx23885[0]: registered device vbi2 [ 9.670749] cx23885[0]: registered ALSA audio device [ 10.605682] cx23885[0]: registered device video3 [mpeg] I then set the channel with ivtv-tune and used mplayer. This is what I get with cat /dev/video3: http://dl.dropbox.com/u/2429209/1800analog.mpg Am I doing something wrong? Thanks, Britney I have been unable to get analog to work on my HVR-1800. I built linux-media about 2 weeks ago.

dmesg shows the firmware is loaded and the mpeg encoder is at /dev/video3:
[ 5.849706] cx23885 driver version 0.0.3 loaded
[ 5.849787] cx23885 0000:02:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
[ 5.849932] CORE cx23885[0]: subsystem: 0070:7809, board: Hauppauge WinTV-HVR1800 [card=2,autodetected]
[ 6.239386] cx23885[0]: hauppauge eeprom: model=78631
[ 6.289124] cx25840 8-0044: cx23887 A/V decoder found @ 0×88 (cx23885[0])
[ 6.983004] cx25840 8-0044: loaded v4l-cx23885-avcore-01.fw firmware (16382 bytes)
[ 9.670505] cx23885[0]: registered device video2 [v4l2]
[ 9.670527] cx23885[0]: registered device vbi2
[ 9.670749] cx23885[0]: registered ALSA audio device
[ 10.605682] cx23885[0]: registered device video3 [mpeg]

I then set the channel with ivtv-tune and used mplayer. This is what I get with cat /dev/video3: http://dl.dropbox.com/u/2429209/1800analog.mpg

Am I doing something wrong?

Thanks,
Britney

]]> Comment on Call for testers: PCTV 74e Support by Devin Heitmueller http://www.kernellabs.com/blog/?p=1378&cpage=1#comment-2503 Devin Heitmueller Mon, 16 Apr 2012 07:29:31 +0000 http://www.kernellabs.com/blog/?p=1378#comment-2503 Hello Saulus, I haven't done any testing with USB3 (as I actually don't have a USB3 host controller). I would suggest though making sure that the device works under USB 3 under Windows (to ensure it isn't some sort of hardware quirk). Beyond that, all I can suggest is bringing it to the attention of the linux-media mailing list to see if anybody has any ideas. Cheers, Devin Hello Saulus,

I haven’t done any testing with USB3 (as I actually don’t have a USB3 host controller). I would suggest though making sure that the device works under USB 3 under Windows (to ensure it isn’t some sort of hardware quirk). Beyond that, all I can suggest is bringing it to the attention of the linux-media mailing list to see if anybody has any ideas.

Cheers,

Devin

]]>