Fallen Heroes Fund

 

A Discussion On HDTV Internal Workings

First let's start with a block diagram of a typical digital TV:

The main differences from brand to brand will be in the technology of the video processing, deinterlacers, noise reduction, scalers, etc., but the basic block diagram will be the same. (I didn't draw this diagram, I lifted it from cirrius.com.)

Ok, now as you can see there are two areas where video processing occurs. In the case of SD signals there is the magic block called the video digitizer. Here the signals from the analog tuner and the analog inputs are fed to it. In that magic box is where the signal is digitized and converted to a memory bit array and then processed. For the purpose of this discussion we will say a 640x480 pixel memory array. All of the deinterlacing and other video processing has been done by the time the signal is stored in this video array output. they process the SD material at this point because it takes less firepower to do sophisticated processing at 640x480 than it would at a HD resolution, thus less lag time. The output of this video digitizer is then sent to the Video Up Scaler where a video pixel array output is built for the resolution of the display being driven. In the case of a CRT display this would be a 1920x1080 pixel array. In the case of a 1280x720 fixed pixel display, that would be the output size.

Now it is a pretty simple matter to output interlaced video to the CRT tube by outputting the odd lines on one scan of the CRT followed by the even lines on the next pass. If driving a fixed pixel display the lines are transferred line by line.

Now the beauty of digitizing the analog signals allow for digital processing which can allow the processor to look at neighboring pixels in the arrays to achieve a better smoothness when de-interlacing video and to allow for the 3:2 reverse pull down to occur without getting the display out of sync with the 60 fields per second refresh of a CRT or 60 frame per second of a fixed pixel display.

Remember when the 3:2 reverse pull down occurs (the processor operating in the film mode) the effective frame rate out of the SD video digitizer is 24 frames per second. All this gets worked out by just transferring the 24fps video data into the Video Up-Scaler at 24fps and letting the signal be sent out to the display at 60 fields (or frames) per second. By allowing the display output pixel array to be accessed at 60fps the picture will change whenever the next 24fps video arrives, since in the digital realm the input and output to the HD pixel array can asynchronously to each other, the 3:2 cadence is not necessary as it would be if we were dealing with an analog only system where the timing had to be at 60fps throughout.

Notice the HDMI, or DVI if so equipped, is just received and passed to the Video Up-Scaler because there would be no pre-processing required being digital video already. This block diagram does not show HD component inputs, but presumably they would go through a digitizer and then fed to the Video Up-Scaler in a similar path to the HDMI.

To my knowledge, no manufacturer has divulged the inner workings of their video processors in great detail, but you can get an idea by checking out this link: What Is Video Processing?

Also you may want to check out:
http://www.cirrus.com/en/pubs/appNo...CS493xx_DTV.pdf

Add Comment...