Page 1 of 1
Posted: Mon Feb 20, 2012 7:44 pm
I was wondering if there's a well documented video decoder chip to interface with our SAM3U4E. I know the SAM9's have enough muscle to tackle decoding via software, but I would like to make a video project around the SAM3U.
The video playback target:
- at least 640x480
-audio would be nice too, but not necessary
-MPG 2 or newer formats
I doubt I will be the first to interface with a video decoder, but did not find any projects online!
I'm hoping one of the members saved a link to share with us.
Re: Video decoder
Posted: Mon Feb 20, 2012 10:51 pm
marioR wrote:I doubt I will be the first to interface with a video decoder, but did not find any projects online!
Because normally people pick a chip with sufficient horse power to do the decoding, I don't see you being able to do this in software on a Cortex-M3, and the hardware solutions tend not to publish a lot of technical detail, or require NDA's.
Also licencing MPEG, or other video formats is a nightmare.
You also need a lot of memory, and a high-bandwidth data source, again something I'm not seeing here. Also something to drive an LCD Panel.
ATMEL has ARM9 parts running at 400 MHz, attached to DDR SDRAM, and including video decoding. See the AT91SAM9M10.
Re: Video decoder
Posted: Tue Feb 21, 2012 12:43 am
As if the lack of a FREE competent compiler wasn't bad enough!
It does seem like I'm dealing with a different type of animal all together. The lack of an open online video project did seem to suggest something was amiss. But, like you mention, nothing can get done if manufacturers don't release appropriate documentation, or if they require their customers to sign NDAs.
May I ask (without, hopefully, touching on any one's sensibilities) which compiler is the most "open" for the SAMx platform (meaning, more avr-gcc like in terms of popularity)?
What would be the cost of entry for the aforementioned compiler?
I would like to research more on the AT91SAM9M10. I really would like to get this video playback project going, but if I did, it does seem like I might get into trouble for publishing the source code on the internet for such project.
Last thing I would like to mention: MPG2 has been around for a really really long time. I'm sure all related patents have expired by now, so why such a tight grip on the market?
Please post comments/suggestions to my questions if anybody has any.