Nicholas A. Sinnott-Armstrong wrote:
On Nov 11, 2007, at 12:20 PM, James Richard Tyrer wrote:
STM claims 1080i H.264 HiProfile:
http://www.st.com/stonline/products/literature/bd/11102/sti7100.pdf
This isn't a full data sheet but just a product blurb.
It does have USB 2.0 and SATA built in. Sorry no Ethernet. It might be able to use an external
Ethernet chip. It has UARTs but don't know how fast they are.
On page 4 there is a picture of an ethernet controller hanging off the EMI --
looks promising.
As for the UARTs, I imagine the one for IR is 40Khz and the one for RS232 is
115200, but those
are just guesses -- they could be general purpose and go to much higher speeds.
There are a total
of 4.
No way to tell if they are fast enough for 10 Mb/s. IIUC, faster
Ethernet requires a parallel date bus.
As for the rest of it, it looks really nice! If we can get it working with all
those features, I would
be happy. SATA is a nice bonus feature -- just plug it in and there you go. Is
there a cost estimate?
Now we get to the bad parts:
1) Only decodes Mpeg-2 and AVC. I imagine there would be a larger demand if we
supported Mpeg-4
and DV -- the rest of the codecs, while nice to support, are not necessary.
Page 6 says: "compatible with all popular audio standards". This
probably means that this isn't final yet. Some chips have an audio
accelerator but require software for audio CODECs. Don't know how this
chip will handle it. IAC, MCUs are now fast enough to do audio decode.
2) It has HDCP. While this is "good," since we don't have to worry about it and
we get support for
some larger TVs and higher resolutions, it has severe legal problems (are
we even allowed to buy
HDCP-enabled chips in the first place, let alone sell them?) IIUC.
We probably need a license to purchase. But with a single chip system
where all of the decoding is firmware based, it should be possible to
obtain a license. The only unfortunate thing here is that user update
of the Flash ROM wouldn't be possible.
The requirements of the license are that if we decode video media and
output in what is called High Resolution that it can only go out the
HDMI with HDCP and that there is no way to modify the hardware (epoxy
potting seems to be sufficient).
Note as usual that I hate DRM and I hate the license, but our purpose
here is to sell a commercially viable product. Yes Faust got screwed
and wouldn't have been OK if God hadn't intervened. I hope that we have
some 'gods' to intervene if we need them (like the EFF).
Then there is Transport Stream Input. See #1.1.2. I am not 100% clear
on what these are, but block diagrams show video input connected to them
and it says: "1 bidirectional interface" which must mean (A/V ?) output.
3) No updating anything. This is a first-gen piece of silicon, so we should
face the facts -- there will
be problems. If we use a DSP and FPGA, just load another updated program
and be done.
Most of the diagrams show a Flash chip. It isn't clear how much of the
software would be in the Flash chip and how much is built in. If they
are smart they would put most of it in the Flash chip except for the
hardware acceleration of decoding.
4) No A/V inputs! Fine, it has 2 channel PCM in, but nothing worth noting. We
could add some external
device to the EMI, but that would likely increase the cost to above the
other options.
Yes, but there is an External Memory Interface and DMA. I don't see
that we need A/V inputs except for what comes over the interconnect
cable (UCB, Ethernet, 1395) or from the optional HD. Also it has some
of those I2C interfaces which can be used for control (to tell it what
to do).
Also, this is just plain speculation
Absolutely! Until we see a data sheet, everything is speculation.
<SNIP>
--
JRT
_______________________________________________
Open-hardware-ethervideo mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-hardware-ethervideo
