[opendtv] Re: Pros/Cons of DSP processor based MPEG software decoder or hardware based MPEG decoder
- From: "Keith Jack" <kjack2@xxxxxxxxxxxxxxxxxxxxx>
- To: <opendtv@xxxxxxxxxxxxx>
- Date: Tue, 8 Feb 2005 14:29:25 -0800
> 4) The notion of dynamically swappable SW video decoders may=20
> work on PCs but it won't work very well in STBs and DTV sets.=20
> The possibility of doing a channel change between, say,=20
> H.264/AVC and VC-1 and not having a long burp in the video is=20
> remote without a lot of very careful systems work (which=20
> hasn't been done yet by any standards body).
As long as the decoder can decode a frame of video and swap codec =
operation
in less than 1/2 video frame, you could do it with proper buffer =
management.
=20
> -----Original Message-----
> From: opendtv-bounce@xxxxxxxxxxxxx=20
> [mailto:opendtv-bounce@xxxxxxxxxxxxx] On Behalf Of Donald Koeleman
> Sent: Tuesday, February 08, 2005 2:15 PM
> To: opendtv@xxxxxxxxxxxxx
> Subject: [opendtv] Re: Pros/Cons of DSP processor based MPEG=20
> software decoder or hardware based MPEG decoder
>=20
>=20
>=20
> ----- Original Message -----=20
> From: "Tom McMahon" <TLM@xxxxxxxxxx>
> To: <opendtv@xxxxxxxxxxxxx>
> Sent: Tuesday, February 08, 2005 9:15 PM
> Subject: [opendtv] Re: Pros/Cons of DSP processor based MPEG=20
> software decoder or hardware based MPEG decoder
>=20
>=20
> A couple of points:
>=20
> 1) There are now a number of hardware H.264/AVC chips and=20
> also a number of licensable H.264/AVC hardware (Verilog/VHDL)=20
> cores. I can send you a list if you're interested in that=20
> level of detail.
>=20
> 2) H.264/AV was developed jointly by MPEG (ISO) and VCEG=20
> (ITU) in a cooperative embodiment known at the JVT (Joint=20
> Video Team). So MPEG versus MS isn't quite an accurate=20
> characterization. Also, MS is very much an H.264/AVC patent=20
> holder as well, so saying MS versus any of the above doesn't=20
> quite work. Weird situation this one.
>=20
> Two groups of products, Windows Media 9 and derivatives, and=20
> ISO MPEG 4 AVC/ITU H.264.
>=20
> 3) My sources tell me that once a video decoder design is=20
> implemented, debugged, and optimized for hardware, it=20
> occupies such a small piece of real estate on typical STB=20
> chips that it is easy to put more than one such decoder on=20
> the chip. They also suggest that this notion of recycling=20
> certain math/function blocks across two different decoders=20
> doesn't work out very well. Even codecs which have some=20
> degree of similarity at some level (H.264/AVC and VC-1 some=20
> would say) are very very different at the hardware=20
> implementation level.
>=20
> Indeed, the problem would propably not lie with footprint,=20
> but with processing and memory demands.
>=20
> In case of multiple codecs, licensing may indeed bcome a cost factor.
>=20
> 4) The notion of dynamically swappable SW video decoders may=20
> work on PCs but it won't work very well in STBs and DTV sets.=20
> The possibility of doing a channel change between, say,=20
> H.264/AVC and VC-1 and not having a long burp in the video is=20
> remote without a lot of very careful systems work (which=20
> hasn't been done yet by any standards body).
>=20
> Don't think some-one is actually proposing to do so. The=20
> programmable chips are only designed to allow for upgrading=20
> to better codecs, not to be used for switching between=20
> multiple codecs on a continuous basis.
>=20
> 5) As far as I know, the following have announced decisions to go with
> H.264/AVC: DIRECTV, Echostar, DVB mandated for HDTV=20
> applications, DVB optional for SDTV, Euro1080, French DTV (or=20
> some subset), Japan for mobile.
>=20
> Let's not forget Premiere wich has announced a November 1=20
> launch of its three channel HD bouquet.
>=20
> French DVB-T only for the HD channels and new services=20
> (mandated), the regular channels are still to start=20
> broadcasting in MPEG 2 MP@ML this spring. The mandated use of=20
> H.264 met with resistence from the broadcasters, so TF1 made=20
> a jump forward and proposed to use H.264 for the regular=20
> broadcasts from the start.
>=20
> Donald
> -----Original Message-----
> From: opendtv-bounce@xxxxxxxxxxxxx=20
> [mailto:opendtv-bounce@xxxxxxxxxxxxx] On Behalf Of Donald Koeleman
> Sent: Tuesday, February 08, 2005 5:35 AM
> To: opendtv@xxxxxxxxxxxxx
> Subject: [opendtv] Re: Pros/Cons of DSP processor based MPEG=20
> software decoder or hardware based MPEG decoder
>=20
> MPEG 2 MP@ML decoder chips are generaly all hardware, the=20
> technology on these chips has matured over the past decade=20
> and there is no real need to update the decoding capabilities=20
> after the box leaves the factory. MPEG 4 Main profile=20
> standard definition decoders are available in hardware,=20
> however anything newer or higher spec, like H.264 or WM9/VC1=20
> uses software codecs. Several reasons, largely to do with=20
> economies of scale, as numbers are still relatively small.=20
> Codecs are still under development, so generations of=20
> capability sets replace eachother quite rapidly, so no time=20
> to design and build a hardware implementation and recoup=20
> cost. The market is still split between MPEG and Microsoft,=20
> and large operators are still making up their minds as wich=20
> platform to choose and current service providers do not want=20
> to be locked in right at the start of the development of this=20
> technology.
>=20
> Donald
> ----- Original Message -----
> From: "Prashant Desai" <prashant.desai@xxxxxxxxxx>
> To: <opendtv@xxxxxxxxxxxxx>
> Sent: Tuesday, February 08, 2005 12:25 PM
> Subject: [opendtv] Pros/Cons of DSP processor based MPEG=20
> software decoder or hardware based MPEG decoder
>=20
>=20
> Hello All
> i am new to this digital content would , i would like to=20
> know and understand the Pros/Cons of different methods of=20
> implementation of the MPEG -4 , MPEG-2 decoder is better=20
> compared to the other .
>=20
> Based on my current understanding there are two ways by=20
> which the MPEG-4 /MPEG-2 decoder's can be implemented the=20
> first method is using a combo of DSP processor and software=20
> implementation of MPEG-4/2 decoder whereas the
> other is to implement the full hardware based MPEG-4/2=20
> decoder , i think
> it's always better to have 100% hardware based
> decoder from performance point of view , i dont know what are=20
> the advantages of using a software based MPEG-4/2 decoder on=20
> the top of DSP processor ..........
>=20
> i would be really grateful if you guys can provide your=20
> opinions and imputs in this regard
>=20
> warm regards,
> Prashant Desai
>=20
>=20
>=20
>=20
> ----------------------------------------------------------------------
> You can UNSUBSCRIBE from the OpenDTV list in two ways:
>=20
> - Using the UNSUBSCRIBE command in your user configuration=20
> settings at FreeLists.org
>=20
> - By sending a message to: opendtv-request@xxxxxxxxxxxxx with=20
> the word unsubscribe in the subject line.
>=20
>=20
>=20
>=20
>=20
> ----------------------------------------------------------------------
> You can UNSUBSCRIBE from the OpenDTV list in two ways:
>=20
> - Using the UNSUBSCRIBE command in your user configuration=20
> settings at FreeLists.org
>=20
> - By sending a message to: opendtv-request@xxxxxxxxxxxxx with=20
> the word unsubscribe in the subject line.
>=20
>=20
> =20
> =20
> ----------------------------------------------------------------------
> You can UNSUBSCRIBE from the OpenDTV list in two ways:
>=20
> - Using the UNSUBSCRIBE command in your user configuration=20
> settings at FreeLists.org=20
>=20
> - By sending a message to: opendtv-request@xxxxxxxxxxxxx with=20
> the word unsubscribe in the subject line.
>=20
>=20
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts:
