176 replies to this topic

[athoik]

Hello,


Here is a log using media player of Merlin3 and OpenPLi 4.

 OpenPLi4_Merlin3_AC3PLus.zip   153.26KB   47 downloads

In Merlin3 logs i can see the following (when i play a file with AC3Plus sound):

write(2, "AUDIO_SET_BYPASS(7) - ", 22AUDIO_SET_BYPASS(7) - )  = 22
ioctl(54, AUDIO_SET_BYPASS_MODE, 0x7)   = 0

In OpenPLi there is no audio set bypass 0x7.

Here is the sample i used for testing.

video-HD1.ts (43.9 MB)

[pieterg]

tough luck then, OpenPLi enigma2 uses 0x22 for AC3+

[athoik]

tough luck then, OpenPLi enigma2 uses 0x22 for AC3+

Can you show us the file that 0x22 is defined?

Something like the following is a possible solution?

#ifdef dreambox 
 0x7
#else
 0x22
#endif

Or it can be done in runtime reading the /proc/stb/info/model and so on?

Edited by athoik, 16 October 2013 - 19:49.

[betacentauri]

http://sourceforge.n...dvb/decoder.cpp
Lines 34-100

[WanWizard]

We already added AC3+ support on 2011-04-09, long before DMM.

 

Since they seem to be very good at reading our code at times, this leaves you wondering why they decided to use 0x7 and break compatibility...

[athoik]

Their (closed) code, their rules!

Something like the following is acceptable solution?

if (::access("/proc/stb/info/model", R_OK) >= 0)
   bypass = 7
else
   bypass = 0x22

Edit1. But a customer might choose another STB brand next time..

Edited by athoik, 16 October 2013 - 20:21.

[WanWizard]

True. But then don't expect us to bend to their rules.

 

It may be an indication they don't want OpenPLi on DMM boxes. If so, so be it then. We don't loose money when nobody buys them anymore...

[Meega]

Edit1. But a customer might choose another STB brand next time..

Yep ,me!!

[DM800SE]

And the developers of OpenPLi will implement the codec AC3+ in these images?

 

Thanks.

[WanWizard]

As I wrote, OpenPLi images already support AC3+ since April 2011.

[DM800SE]

As I wrote, OpenPLi images already support AC3+ since April 2011.

 

But OpenPLI includes drivers in your images?

 

Thanks.

[hemertje]

Ask DMM to change 0x7 to 0x22...

[athoik]

Ask DMM to change 0x7 to 0x22...

In the meanwhile the following patch uses 0x7 when box is DMM.

 0001-Use-different-bypass-for-DD-when-box-is-DMM.patch.txt   675bytes   64 downloads

[jpuigs]

Ask DMM to change 0x7 to 0x22...

 

By reading this answer, I understand that Openpli won't change this Little thing to give DMM boxes AC3+ audio.

If it's true, DMM users will have to use other images.....

[DM800SE]

Ask DMM to change 0x7 to 0x22...

 

By reading this answer, I understand that Openpli won't change this Little thing to give DMM boxes AC3+ audio.

If it's true, DMM users will have to use other images.....

 

I think the same like you...

[DM800SE]

As I wrote, OpenPLi images already support AC3+ since April 2011.

 

This is not a support, is a patch. If you don't put OFF the ac3 downmix, the AC3+ don't hear... In other images (OoZooN, Merlin, Newnigma2, etc) AC3+ sound is heard without putting the ac3+ downmix OFF...

Edited by DM800SE, 16 October 2013 - 23:20.

[WanWizard]

The images you mention are DMM images. Using their closes source binary. We don't use DMM code, we use OpenPLi code. Only the binary platform drivers are supplied by the hardware vendor.

 

The more they put in their closed source binary, and the more non-standard stuff they put in their drivers, the more difficult it will be for us to keep supporting Dreamboxes. At the moment we're the only open source team that still does. And we don't like to add hacks to our code to work around idiotic descisions made by vendors.

 

If a vendor wants OpenPLi to support their hardware (and most do because it sells), then they have to put the effort in. We are not in the business to make them sell boxes. We don't make a penny if DMM sells more. If you want OpenPLi, don't buy a Dreambox.

 

If this goes on, I would not be suprised that in the future there will not be an OpenPLi for a Dreambox. We already don't support the v2 box series, and of the ones we do support, one is stuck at 2.1.

[littlesat]

And if it should be patched we do not want code such as if DMM box then in our code... Then it should be done approx like we did for the remote issue on the VU....

If you want OpenPLi, don't buy a Dreambox.

This is indeed a good recommendation especially when you want to buy a box with open source software... and in addition it seems you save money

 

But still it is extreme wierd that DMM starts to support DD+ suddenly after years of not doing this and then they do not follow the standard..???

Edited by littlesat, 17 October 2013 - 08:23.

[athoik]

Dreambox Ticket System : http://tickets.dream...trs/customer.pl

Go Ahead, start opening tickets requesting from DMM to change the ioctrl from 0x7 to 0x22 for DD+ (bypass = 0x22)

By the way there is no standard for 0x22, It's just ET that implement this first and used 0x22.

Standards require an organazation to standarize it... (eg http://www.dvb.org/standards)

Edited by athoik, 17 October 2013 - 11:54.

[athoik]

Hello,

This is my second patch for DD+. It uses compiler preprocessor, so there is no need to check if a file exists using access.

Subject: [PATCH] dvb/decoder.cpp: workaround for Dreamboxes DD+

Dreamboxes added after many years DD+ support, but they don't like
to follow "standards", breaking existing Open Source implementations.

This patch will use bypass 0x7 for Dreamboxes (only) instead of the well known 0x22.

diff --git a/lib/dvb/decoder.cpp b/lib/dvb/decoder.cpp
index 174d6cf..e242f11 100644
--- a/lib/dvb/decoder.cpp
+++ b/lib/dvb/decoder.cpp
@@ -82,7 +82,11 @@ int eDVBAudio::startPid(int pid, int type)
 		bypass = 0x10;
 		break;
 	case aDDP:
+#ifdef DREAMBOX
+		bypass = 7;
+#else
 		bypass = 0x22;
+#endif
 		break;
 	}
 
-- 
1.7.10.4

Attached Files

•  0001-dvb-decoder.cpp-workaround-for-Dreamboxes-DD.patch.txt   779bytes   71 downloads