NEW Graphics PowerVR SGX Accelerator?

[CornyAxim]

i2700g can hardware decode h264 (not all, 100%, but lots)
hardware h264 in i2700g:
- H.264 AVC (MPEG4 Part 10) Baseline Profile (no cabac)
- Inverse zig-zag scan with programmable scan ordering
- Deblocking filter acceleration
- Inverse H.264 intra prediction
- Full bi-directional prediction
- Half, quarter and eighth-pel motion compensation h264

from powervr:
"The reduction in CPU load achieved is typically in excess of 80% for H.264, thereby allowing the system designer to target lower cost and lower power systems. This is of increased importance when performing quarter-pel motion compensation"

Only good coder is required with lots of free time and i think -> typical movie: 640x272 h264 600kbps-1mbps -> 25 fps smooth on x50v

Why can the 2700g decode "decode h264 (not all, 100%, but lots)" How can it decode some and not all?

As well what does "(MPEG4 Part 10) Baseline Profile (no cabac)" and "Inverse H.264 intra prediction" mean? This is all very techy!! Can anyone interpret what that means?

[irwin2]

"(MPEG4 Part 10) Baseline Profile (no cabac)"
Looking from the MPEG side the standard is called MPEG-4 Part 10 (ISO 14496-10), looking from the ITU side, it is called H.264.
AVC/H.264 defines four different Profiles: Baseline, Main, Extended and High Profile

i2700g helps decode baseline h.264. A full rate H.264 Baseline profile stream decode can be achieved with the core running at less than 50MHz (CIF resolution: 352x288 pixels, 30 frames per second). 640x480 30fps required 100MHz.

"CABAC"
AVC/H.264 defines two, more advanced tools for entropy coding of the bitstream syntax (macroblock-type, motionvectors + reference-index...) than MPEG-4 ASP: Context-Adaptive Variable Length Coding (CAVLC) and Context-Adaptive Binary Arithmetic Coding (CABAC) CABAC, compared to CAVLC (aka UVLC) which is the default method in AVC/H.264, is a more powerful compression method, being said to bring down the bitrate additonally by about 10-15% (especially on high bitrates). CABAC (as CAVLC) is a lossless method and therefore will never hurt the quality, but will slow down encoding and decoding.

i2700g helps with CAVLC.

"Inverse H.264 intra prediction"
read here:
http://www.multimedia.edu.pl/publica..._ECCTD2005.pdf

i2700g helps in this cpu-eating process.

In Dell x50v i2700g (mvda2 video core) running at 75MHz. This means h264 640x272 or 352 in 24-25 fps is real on x50v.
That means: i2700g can helps decode h.264.

[axzilla]

In the interests of science, this is the same clip ripped from DVD.

It's a 1:29:00 movie compressed down to 200MB with audio set at 128Kb/s 44.1KHz stereo. 320x240 frame size.

One is MPEG-4 with AAC audio and the other is H.264 with AAC audio. So the only differences between the two is the image quality encoder. The audio is the same. Both files are 200MB.

Notice how one is clearer than the other? These are native resolution, not enlarged.



So for the same file size H.264 gives a noticably better picture. Even better than MPEG-4. Imagine the benefits over MPEG-2! I don't have an MPEG-2 encoder handy to demonstrate....

your video looks blurry to me on both I guess its because of your super low bitrate my videos are crystal clear but I encode mine at about 125Mb per 30 mins

[CornyAxim]

Most interesting! Wow.