Weblog

After a nice weekend

We promised ourselves to start or end a good day with Champagne or a Cava ;-)

We are again busy with firmware(s) … it never stops ;-)

People do not know what ROT13 is .. I wrote somewhere a sentence, but no-one did notice it yet .. It has a easter-egg / feature / functionality hint in it ..

Testing some features/ functions of the CFULLHDMA after some feedback users

I’ve received some new files which are not playable … it’s good to get a lot content to see if it’s playable yes or no ..

Let’s make it available to some people .. ;-)

Oh Oh .. server probs

I had some severe issues with my server this evening and I had to restart it. Did not want to wait .. so any pending ftp upload: sorry ..

b.t.w. if you upload content for analyse with CFULLHDMA please include your name and/or e-mail or notify me by e-mail cause I see duplicate info on the server and I do not know why this is ..

Beta firmware CFULLHDMA 1.09.11

v1.09.11
1. speed up switch embedded MKV subtitles
2. fixed: MKV subtitle control code issue
3. fixed: Some mp4 could not be played.
4. Improved: some MKV files with artefact issue

You need to be a beta tester to get this beta-firmware. Put a comment under this entry with your real e-mail address and I’ll put you on the beta-testers list. note for GMAIL & HOTMAIL users: maybe my mail will arrive in your SPAM inbox.. check it !

CFULLHDMA: Trying to understand decoding H.264

During the development of the CFULLHDMA I am always trying the best out of things. (= chipset + firmware). Knowing that not everything can be supported it is always a challenge to get the best.

Therefore my search for answers ..

Challenging the Sigma Designs 8635 chipset is one of the main things I like, especially as we know that the SDK of Sigma Designs is far from perfect.

SD 8635 does not support MKV, so this is added by a lot of manufactures themselves. The chip is strong in TS and MP4 .. 

We did some tests and got feedback from users: some certain files could not be played correct: during playback a lot of artefacts (squares) could be seen. In search for ansers why this happened a lot of debugging and research has been done.

In general this is what is happening: the filestream is forced to be played at H.264 High Profile @ L4.1 but the stream is a High Profile @ L5.x one and this is causing the artefacts.

The H.264 High Profile L5.x does need much more memory than the profile L4.1 (Blueray and HDDVD do use L4.1) ..  Also the ref frames have a role in this part: the more ref. frames, the more issues can be expected.

So in search for solutions it’s being checked if enough power/memory can be made available to decode the H.264 High profile L5.x

So my first suggestion is to remux the stream from L5.x to L4.x and check the amount of ref. frames.

It’s also wise to check some forums. I think the L4.1 is the profile to choose.

note: I am not H.264 expert, so if someone has more info about difference between L4.x and L5.x: please give me some hints