In Sunday's New York Times, an article about Time-Warner explains that part of their movie studio's success comes from thwarting online movie rip-offs, using tactics that included flooding movie-sharing web sites with fake versions of "The Dark Knight" to frustrate would-be illegal downloaders. (And the studio takes pride in the fact that a real illegal copy didn't appear online until a full 38-hours after its debut.)
On the political front last week we saw the debut of "Paris Hilton for President" -- a mock ad, responding to a McCain attack ad against Obama -- noteworthy here because it was released only on the Internet, not a dime spent on TV time, yet if you watch news in the U.S. you've probably seen it.
Four years ago, in the previous U.S. elections / Summer Olympics cycle, Internet video was still an oddity. Two years ago, this curious new development lost one U.S. Senator his job (after his ethnic slur was posted on YouTube), but Internet video was still a new thing the kids were doing. Now, as reflected by recent headlines, Internet video is mainstream, and as a threat it's taken quite seriously.
Comment on this blog entry
August 06, 2008
Codecs: 24 design ideas for video and imaging
By
Cliff
Roth
Codecs are essential to video and image processing. Here are twenty-four codec-related design articles.
Massively parallel processing arrays (MPPAs) for embedded HD video and imaging (Part 2)
On2's TrueMotion VP7 video codec and "golden frames"
Video codecs, part 1: Intraframe coding and bitrates
Video codecs, part 2: Interframe coding, MPEG-2 & MPEG-4
Video codecs, part 3: H.264 & video over networks
Video codecs, part 4: Wavelet codecs
Introduction to video transcoding for CE
System and Memory Architectures for High-Definition Consumer Electronics
MPEG-4 system layer and A/V containers for wireless video
Multi-core MPEG-4 video encode partitioning
Video compression artifacts and MPEG noise reduction
Introduction to video compression
Video codecs tutorial: Trade-offs with H.264, VC-1 and other advanced codecs
Encoding JPEG2000 video by combining DSP and FPGA
H.264 encoder design using Application Engine Synthesis
Image compression software with region-of-interest (ROI) codec reduces bandwidth
Video quality assessment with motion and temporal artifacts considered
Video quality analysis with automated, scripted testing
Data compression for high-speed DSP, part 2
How to use field-programmable object arrays (FPOAs) in image processing
Bandlet de-interlace and up-conversion for HDTV via "super-resolution" framework
Surveillance IP camera design with intelligent encoding for reduced bandwidth, higher quality
Streaming video with "TimeSlice" multicore-friendly processing eliminates dropped frames
Encoding JPEG2000 video by combining DSP and FPGA
Comment on this blog entry
July 30, 2008
3-chip vs. 1 image sensor
By
Cliff
Roth
During a briefing on Kodak's new high-end 50-megapixel image sensor -- costing a few thousand dollars each -- I finally got the answer to a question that has puzzled me ever since multi-megapixel imaging became commonplace over a decade ago. The question: Why do high-end pro video cameras continue to use three separate image sensors -- for red, green and blue -- while pro-quality digital still cameras (DSCs) use a single high-megapixel image sensor, such as the new "world's most megapixels" Kodak sensor (see
Kodak delivers next-generation 50-MP CCD image sensor.)
Kodak's new CCD image sensor boasts 50-megapixels
In the video world, the 3-chip imaging system is widely considered better for color accuracy, improved contrast, and more "film-like" images. It requires more optics however -- a prism to split the image three-ways.
The answer, explained Kodak, has to do with artifacts that are created by having red, green and blue-sensing pixels located right next to each other, rather than on top of each other, in the single-chip color image sensor. The processing circuitry must, in essence, create a pseudo red, green, and blue image by calculating in-between pixels. As the image sensor's resolution goes higher and higher, into the multi-megapixel and tens of megapixels zone, these artifact errors become slighter and slighter.
But for the comparatively coarse resolution of standard def video (720 x 480 = 350k pixels, or just 0.35 megapixels) -- and even HD video (which at 1920 x 1080 works out to 2-megapixels) -- the artifact errors are more significant. Hence, 3-chip imaging continues to rule for pro video, but not for pro still cameras.
Comment on this blog entry
