The real truth about megapixels & HD

Learn more and bust through the myths


So, although H.264 needs less bandwidth and storage capacity, it does bring some negatives. Some are in the process of being overcome. For instance, PC graphics cards are moving quickly to the H.264 standard and video management software vendors are making vast improvements in H.264 processing so that, eventually, the industry will overcome this particular drawback...just not yet! Today, when viewing multiple cameras simultaneously, over-configure the servers in the control room to cope with the more intensive processing requirements. Or, remember, although M-JPEG does not yield the bandwidth and storage savings when compared to H.264, M-JPEG does not require as powerful PCs or processors.

Another issue is standards. Too often, H.264 is implemented in a proprietary manner by camera manufacturers. Video Management Software (VMS) vendors simply can't support all the 'flavors' of H.264. Be sure that the cameras selected are fully supported by the VMS vendor and that they support the full range of features provided by the camera."

Alan Schwartz-Product Manager-EverFocus, Duarte, Calif.

"Terminology is SO important, as I'm sure the folks at Coke, Kleenex and Frigidaire would agree. For years, when we said 'CCTV,' it did not matter if you thought of it as 'video' 'composite' or 'NTSC/PAL,' we knew what we meant. Then, with the advent of local and wide area networking, we found black boxes that convert the 'analog' video signals into digital data and transmit that data over local and wide area networks to other black boxes which convert the digital data back to its original 'analog' form for display and/or recording on CCTV equipment.

It has been transparent to the average dealer/installer or end-user, but for many years now 90 percent of the video within an 'analog' camera-from the CCD sensor, through the signal processing and controls/menu system-is all digital data. Only at the very end of the process is the digital data converted to NTSC/PAL 'analog video' and delivered at the BNC connector on the camera.

In parallel with this, technology developed sensors capable of capturing images with much higher resolution, much greater data content, than 'standard definition NTSC/PAL' video. Why not put these sensors into our CCTV cameras and improve the pictures? Those international standards for the transmission of video (NTSC & PAL) were defined to handle only a specific, limited amount of bandwidth, which translates to the maximum amount of data/information they can carry. The information content in a high resolution image is more than 'standard analog' video is designed to handle.

What if we don't plan to convert the signal to NTSC/PAL? We're no longer bound by those bandwidth and resolution limitations: we can support any resolution sensor that the data transport medium can handle. IP (typically TCP/IP Ethernet) networks run at least 10Mb/s, frequently 100Mb/s or more.

As a benchmark, a nominal NTSC picture can be represented on a computer monitor in a digital image comprised of 345,600 dots or pixels (Picture Elements) arranged in 480 rows of 720 pixels per row. PAL is 768x564 pixels.

Free of the bandwidth restrictions of NTSC/PAL, we can utilize sensors with more pixels to achieve finer detail and higher resolution. As we saw early on with consumer digital cameras, resolutions of 1280x1024 are readily available: 1,310,720 pixels, or 1.3 megapixels. Enter the megapixel camera. How do we transport that amount of data? At the advent of this technology, IP networks were the only available medium, so there were only IP megapixel cameras. While an IP camera need not be a megapixel camera, there was a point in time where a megapixel camera was automatically an IP camera. Needed megapixel? Had to go IP!

With the advent of HDTV (1080p translates to 1920 x 1080 pixels, or 2,073,600 pixels - about 2 megapixels; 720p translates to 1280 x 720 = 921,600 pixels or ~1 megapixel), the broadcast industry developed protocols to transmit HDTV over coax, called high definition-serial digital interface (HD-SDI), standardized in SMPTE 292M (Society of Motion Picture and Television Engineers); this describes a nominal data rate of 1.485 Gbit/s over coax, terminated with BNC connectors.

So, with a protocol defined to transmit high definition (1 and 2 megapixel) signals over coax (of which coincidentally there are millions of installed runs with BNC connectors at both ends, a standard definition camera at one end and a recorder at the other end) enter an alternative protocol for transmission of megapixel video images: HD-CCTV. EverFocus saw the potential for HD-CCTV and supported the initiative by becoming a founding member of the HD-CCTV Alliance.