XBMC for Linux port to SageTV HD Extender (based on a Sigma Designs chipset)?
#1
Question 
Just a dirty copy/paste from the mythtv mailing-list:
Quote:Hello,

I have been emailing SageTV about thier HD extender to get the source code since mid December. I figured I would never get a response but after 5 emails reminding them Sage has delivered!

http://download.sage.tv/pubcode/Pullreader.tar.gz

http://download.sage.tv/pubcode/freetype-2.3.5.tar.bz2

http://download.sage.tv/pubcode/linux-2....est.tar.gz

http://download.sage.tv/pubcode/busybox-1.7.2.tar.gz

http://download.sage.tv/pubcode/jpegsrc.v6b.tar.gz

http://download.sage.tv/pubcode/tremor.tar.gz

http://download.sage.tv/pubcode/ffmpeg.tar.gz

http://download.sage.tv/pubcode/libpng-1.2.20.tar.bz2

http://download.sage.tv/pubcode/zlib-1.2.3.tar.gz

I am not really sure what I have here but I think things look good to run a full MythFE on this. If this is possible it would be the ultimate addition to MythTV, low power, silent, small form factor (well second to Hauppauge Component capture card).

Would anyone with the skill to reverse engineer this box be willing to step up and own this project? I am sure it would take much to get the community to purchase a box for you!

Let the discussion begin!

Mitchell

Could this be a candidate for XBMC? Seeing as it's a HD-Extender that supposedly handles h264/1080p, could there be some hw-acceleration in there?

-C-
Reply
#2
Thumbs Down 
Please read the replies to these two topic-threads:
http://forum.xbmc.org/showthread.php?tid=30032
http://forum.xbmc.org/showthread.php?tid=30207

The same applies for the "SageTV HD Extender" box Nod
("SageTV HD Extender" is based on a Sigma Designs chip)
Always read the XBMC online-manual, FAQ and search the forum before posting.
Do not e-mail XBMC-Team members directly asking for support. Read/follow the forum rules.
For troubleshooting and bug reporting please make sure you read this first.
Reply
#3
Gamester17 Wrote:Please read the replies to these two topic-threads:
http://forum.xbmc.org/showthread.php?tid=30032
http://forum.xbmc.org/showthread.php?tid=30207

The same applies for the "SageTV HD Extender" box Nod
("SageTV HD Extender" is based on a Sigma Designs chip)

Actually, this is incorrect. The Sage Mdia Extender supports both OpenGL acceleration and hardware MPEG-2 and MPEG-4 decoding. This is a requirement because SageTV uses OpenGL to render its UI and SageTV uses more CPU than XBMC does.

The real question is does the released code contain enough information to utilize the hardware chips.
Reply
#4
The Sigma Designs chips have only 2d acceleration.

http://www.sigmadesigns.com/public/Produ...guide.html
Reply
#5
althekiller Wrote:The Sigma Designs chips have only 2d acceleration.

http://www.sigmadesigns.com/public/Produ...guide.html

Interesting, they must be using software GL then.

All I know is if it can run SageTV it can surely run XBMC, after all SageTV is even running in a JVM.
Reply
#6
althekiller Wrote:The Sigma Designs chips have only 2d acceleration.

http://www.sigmadesigns.com/public/Produ...guide.html

This is correct, however the Sigma chips also can have another processor act as the primary processor while they act as a slave and only handle video decoding.

This could mean that a hardware OpenGL 2.0 solution could be implemented

However, from reading the specs on the Myth wiki that Gamester posted, it looks like the Sage extender uses the exact same hardware as the Negear digital entertainer HD box, and does not use an external processor.

Therefore, no hardware OpenGL 2.0 and only a 200Mhz ARM = no go. Sad
Reply
#7
waldo22 Wrote:This is correct, however the Sigma chips also can have another processor act as the primary processor while they act as a slave and only handle video decoding.

This could mean that a hardware OpenGL 2.0 solution could be implemented

However, from reading the specs on the Myth wiki that Gamester posted, it looks like the Sage extender uses the exact same hardware as the Negear digital entertainer HD box, and does not use an external processor.

Therefore, no hardware OpenGL 2.0 and only a 200Mhz ARM = no go. Sad

Again - this makes absolutely no sense when you consider the facts

- The SageTV UI is runs in OpenGL
- The default SageTV UI is arguably much more graphically intensive than most XBMC themes
- SageTV itself is not even native code, it is written in Java
- SageTV runs fine on this hardware platform

Considering all the above, the statement that the box hardware not physically handle the XBMC UI, makes no sense whatsoever. There would for sure be other porting barriers, but I wish people would stop dismissing it outright for being underpowered, because it simply can't be, it would make zero sense.
Reply
#8
Anything is possible, and your patches to make XBMC work on sigma hardware would definitely be accepted in SVN. Until some enterprising dude sits down and adds support, XBMC for linux only works on intel linux, with opengl 2.0 graphics hardware, and needs a fast CPU.
Reply
#9
brunes Wrote:- The SageTV UI is runs in OpenGL
Clearly only 2D extensions, read the SD datasheets
Quote:- The default SageTV UI is arguably much more graphically intensive than most XBMC themes
From the screenshots on their site, the only response I have to that is "HA!"
Quote:- SageTV itself is not even native code, it is written in Java
Not necessarily true, many ARM cores support a feature called "Jazelle" Java accelerator in which case the bytecode is executed on hardware not a VM.
Quote:- SageTV runs fine on this hardware platform
I would certainly hope it runs fine if they expect people to pay $200 for it. This doesn't make XBMC equivalent or even comparable to SageTV.
Quote:Considering all the above, the statement that the box hardware not physically handle the XBMC UI, makes no sense whatsoever. There would for sure be other porting barriers, but I wish people would stop dismissing it outright for being underpowered, because it simply can't be, it would make zero sense.
You've clearly displayed a lack of either the understanding of what is involved here or effort in demonstrating your knowlege on the subject. That said, no one currently on the team is interested. Patches will most certainly be considered, but are in no way _guaranteed_ acceptance into SVN as rodalpho implies.
Reply
#10
So what about the HD200 box, would that be workable for XMBC?
Reply
#11
No, same deal, slightly fast SoC. If you can't deduce these facts yourself, please at least link to the technical specs.
Reply
#12
althekiller Wrote:No, same deal, slightly fast SoC.
And "slightly fast SoC" means what exactly. Did you mean to say "slightly faster" perhaps?
Quote:If you can't deduce these facts yourself, please at least link to the technical specs.
Sorry, didn't mean to make you do anything here - just looking for some input...

It uses Sigma chip SMP8635LF More info for comparison to view here

HOST CPU is 300Mhz MIPS
512MB DDR400 RAM

Still 2D which probably is what would hold xbmc back I guess. Too bad - it's a very capable machine for SageTV.
Reply
#13
Yes "faster". I'd have fixed it but I figured those dots were easy enough to connect. The first hurdle is the unsupported CPU architecture, then the no GL, then fighting the DSP codecs.
Reply
#14
althekiller Wrote:Yes "faster". I'd have fixed it but I figured those dots were easy enough to connect. The first hurdle is the unsupported CPU architecture, then the no GL, then fighting the DSP codecs.

What about the PCH - are either versions of that a possibility? I'll add links to the technical specs if you're unfamiliar with that one as well - just let me know.
Reply
#15
PCH been discussed already, it too is a no go but I think someone mentioned an XBMC look alike skin for it...
Openelec Gotham, MCE remote(s), Intel i3 NUC, DVDs fed from unRAID cataloged by DVD Profiler. HD-DVD encoded with Handbrake to x.264. Yamaha receiver(s)
Reply

Logout Mark Read Team Forum Stats Members Help
XBMC for Linux port to SageTV HD Extender (based on a Sigma Designs chipset)?0