Archived Tech-Notes
Published by: Larry Bloomfield & Jim Mendrala The following are our current e-mail addresses:
E-mail = hdtvguy@garlic.com or J.Mendrala@ieee.org
We have copied the original Tech-Notes below as it was sent out. Some of the information may be out of date.
**********************************************DTV Tech Notes
% Larry Bloomfield & Jim Mendrala
(541) 385-9115 or (805) 294-1049
E-mail = larrybend@aol.com
J_Mendrala@compuserve.com
============
January 15, 1998
============
DTV Tech Note - 013
============
Sharing experiences, knowledge or anything else relating to DTV, HDTV etc. with your fellow engineers is what we are all about. For a copy of our policies and who we will make this available to, please e-mail us. Welcome to all the new subscribers. We now have over 150. We hope you will participate with question, answers and/or comments. To those of you who have been with us and wondered were we've been, read on. New things are in the works.
*************
Subj: Dolby Digital AC3
From: Larry Bloomfield
Audio problems not necessarily East of Denver.
Our prime purpose here in the DTV Tech Notes has always been to be a forum where we can present problems and help find solutions. There is a major problem looming in the not too distant future of HDTV. As we all know, the audio will be Dolby Digital AC3. This is an encoded bit stream which consists of 5.1 channels; specifically five full frequency range channels; Right front, Center front, Left front, Right surround (rear) and Left surround (rear) and one frequency limited channel of sub-bass material. It sounds great but when it is encoded as AC3 when it comes in on let say a network feed, it is not user friendly at all! Like it or not, full frequency or not, before the audio was encoded, that adds up to six channels of audio. There are still many small market stations that have problems dealing with "stereo" (2 channels) audio much less six.
Bring in the encoded AC3 on the network and sending it, unmolested on through at the local station level, shouldn't be a problem. Our question is what happens when it is necessary to do a voice over, a promo over the close of a network show or even more important when it is necessary to do an EAS? As we see it, the local station will have to decode the AC3 back into six channels add what they need to add and re-encode it.
Our questions are: Is there a simpler way of doing it? How can we do our local audio business such as inserts without screwing up the incoming signal? If you think about it, we're sure you can thing of several more.
Larry
**************
We're not Alone!
The governments of Canada and South Korea have recently announced formal decisions to adopt the ATSC Digital Television (DTV) Standard for digital terrestrial television broadcasts in those countries, thus joining the United States in mandating the ATSC Standard for such broadcasts.
Other counties are expected to join soon. For more info, see:
http://www.atsc.org/Presshtml/PR-Canada&SKorea_Adopt_ATSC.html
*******
Subj: Fwd: Faroudja Inc. To Deploy Prototype Upconverters Next...
From: MEDIACREAT@aol.com
Someone has gotta come out and say whether the following is the "real thing" or just an HDTV technological "cheat." To sum up what Faroudja is doing... it's like the propeller manufacturers saying how they have designed an even better prop for the B-36, when the B-47 jet was the obvious choice.
Maybe Faroudja's next advanced technological step is to buy CBS's color wheel patents, or an Edsel, i.e., "NTSC" dealership! I hear Chrysler's auto turbine technology is on the block too! (Ed Note: Don't sell your CBS color wheel stock. Texas Instruments is using a color wheel with their DMD projections system.)
---------
Faroudja Inc. To Deploy Prototype Upconverters Next...
Faroudja Inc. To Deploy Prototype Upconverters Next Month; Revolutionary Delivery of HDTV-Quality Images From Conventional Broadcast Sources
(Ed note: When I saw this same press release, I contacted Faroudja and could not get them to comment. I did, however speak with Snell-Willcox who also has an Up Converter. The results of that interview is in the Jan '98 issue of Broadcast Engineering which will be out in a few days.
**********
Subj: Sony's E-Cinema
(The following came from Bill Hogan. It is about Larry Thorpe's talk at the STE. It originally was sent out to an internet forum similar to this one for Telecine interested folks. (STE is a Los Angeles based organization of Television Engineers. The organization has been around for over 50 years and has made many valuable contributions to the Television industry.) Since E-Cinema and HDTV have many thing in common, it is presented here for your information.)
Sony, trying to keep their foot in the door with tape machines,
(still there primary business) has announced (publicly stated) that at
NAB they will show a prototype of a new uncompressed HDTV VTR. This
replaces their open reel 1" machine that uses tape that costs
$1,000.00/HR. When the HDTV Standards were announced at this time a year
ago Sony found that they did not have a Modern VTR for Post Production.
Instead they had spent the effort on a 7 to 1 compressed format for
field recording that fits on the back of Camcorder sized like a
Digi-Beta.
This recorder will record the current HDTV digital standard of 1920 x
1080 at 60 (or 59.94) interlaced fields per second. Also it was
announced that it would record full bandwidth RGB (no transcoding) at
1920 x 1080 24 Frames per second PROGRESSIVE. This is because when you
do the math you find that the data rate is the same for the two
formats. (Ron Martin--are you listening?) Yes, Sony recording at 24
Frames progressive. Bet you it will also record 720 progressive too.
They also announced they are going to make a 24 Frame progressive Field
camera for "Cinema" shooting for going back to film for distribution.
The VTR is based on their "D2" transport---OH NO!!!!!!!
Only catch is that the VTR will not be available til "Some time in
1999" In the mean time D5 is gaining a strong foothold in the market.
Regards, Bill
Bill Hogan bhogan@sprocketdig.com v.818-566-7700
Sprocket Digital Burbank,CA f.818-566-4477
*************
Subj: Transform Any Receiver to Dolby Digital; Vantas...
(The following is presented as the Dolby Digital AC-3 will be used in conjunction with HDTV.)
Vantas, a Microtek company based in Redondo Beach, CA announced the introduction of the first digital surround processor/amplifiers that allow users to play back Dolby Digital (AC-3) with any receiver. These products were shown at the 1998 International Consumer Electronics Show in Las Vegas. The majority of the more than 29 million Dolby Surround systems sold to date can not play back 5.1-channel Dolby Digital. Dolby Digital is a universal standard found on DVD, laserdisc, and PC multimedia and will migrate to DSS, HDTV and digital cable.
They showed two models which will enable all Pro Logic and stereo home theater systems to play back 5.1 channels of Dolby Digital encoded sources.
For further information, contact Elliot Rubin, Director of Marketing, at (310) 297-5720 or rubin@vantas.com
*************
Panasonic Develops First Single-Chip Digital Television Video Decoder
Panasonic AVC American Laboratories, Inc. (PAVCAL) announced that it has
completed development of the world's first low-cost single-chip device which will switch between any of the eighteen different television formats, each suited to different purposes. It will also have the capability of addressing the different screen ratios (16:9 'wide-screen' or 4:3), numbers of horizontal and vertical lines of resolution, and scanning methods (either 'interlaced' scanning, like today's TV displays, or 'progressive' scanning, like computer monitors).
"This is the first single-chip device that can decode and display all of the digital TV signals that can be broadcast, using the new digital standard, in any of the different HDTV or standard definition formats," said Sai Naimpally, PAVCAL Vice President and leader of its DTV development team. "It processes the digital signals in two ways, both decoding them for display in their original format, and converting them for use in today's televisions."
The chip--technically termed a "Digital Television MPEG2 Main Profile at High Level Video Decoder" -- functions in both a 'full-spec' mode and a 'down-conversion' mode. In the full-spec mode, it decodes the compressed video signal from the broadcast and outputs the original format, that is, either HDTV (1080-lines interlaced or 720-lines progressive) or SDTV (480-lines interlaced or 480-lines progressive). Single chip operation is made possible by use of 500 MHz concurrent 16 Mbit Rambus(TM) DRAM's.
The 'down-conversion' mode converts all compressed video signals to 480-interlaced and 480-progressive formats. This is accomplished by a memory-efficient MPEG down-conversion algorithm developed by PAVCAL.
The operation of the decoder chip conforms to both the DTV Broadcast Standard adopted by the Federal Communications Commission and the more-detailed ATSC DTV Standard, drafted by the all-industry Advanced Television Systems Committee (ATSC).
(Ed note: Additional information and diagrams will also be in my column in Broadcast Engineering's Jan. '98 issue LB)
*************
Subj: Sampling - a refresher
From: James Mendrala
We've seen different sampling rates batted around. It wouldn't hurt to know the why and wherefore about them. First you have to know where 4:2:2 came from. As we all know, the frequency of the color subcarrier is 3,579,545.265 Hz. This is because the subcarrier has to be an odd harmonic of the 4.5 MHz offset of the aural carrier from the visual carrier. Otherwise severe morie between the two would look horrendous. In order to take a composite color signal and sample it digitally you need a little more than three samples per cycle of the 3.58 MHz to satisfy Niquest's rule on sampling. if you take 4 samples of the video you satisfy that requirement. So in order to do that you need a sampling frequency 4 times the subcarrier frequency or 14,318,181 Hz. Along comes component. Since the bandwidth of the color difference signals are half of the bandwidth of the luminance signal the term 4:2:2 comes into being. The Luminance is sampled at the 4 times subcarrier and the color difference is sampled at 7,159090.5 Hz. CCIR-601 standardized the frequency of the sample rate at 27 MHz for both PAL and NTSC. that gives us a bout 720 samples per line in both PAL and NTSC. (Since the frame rate in PAL is 25 fps the longer time gives more lines vertically.) MPEG main level@main profile calls for a 4:2:0 sampling.
what this means is that the color difference is sampled at half the bandwidth horizontally and half the number of scan lines vertically. MPEG-2 4:2:2 samples like the numbers show. Full bandwidth for the luminance and half the bandwidth for the two color difference signals. This MPEG-2 4:2:2 is new and will probably be the broadcasters SDTV format of choice for DTV. If the video is sampled at a higher frequency it is usually multiples of the 3.579545 MHz or there abouts. For example the telecine that Advanced Video Designs, the HiRes 1440 samples at 8:8:8 and outputs at 8:8:8, 8:4:4 or 8:2:2 so the video which is over sampled looks good when re-sampled at Rec-601 clocks. In theory a 8:8:8 video would have about 1440 lines of resolution double that of NTSC's 720 lines.
Jim
***********
Subj: On the move
By: Larry Bloomfield
Things change. Since our last issue, I am no longer with KTVZ. Although I have gotten to love Central Oregon, as soon as I find a new home for my talents, I anticipate moving.
Since our last issue, I was contacted by Broadcast Engineering and asked to write for them and get paid for it too. This will not be a full time job, however. To put your mind at ease and as you know, the DTV Tech Notes is a work of love and none of us get paid for what we do here, nor do we ask nothing from any of our subscribers except input. We want to keep it that way. I have promised the folks at BE that I will not print anything in here first if it is going to appear in their publication. That's fair. By the same token, I will not use anything in BE that you have provided us here, with out your permission. I don't think there well be a problem telling you about things I've run across that pertains to what we do here and refer you to it in BE. The section that I'll be writing with be entitled: "Behind the Headlines." It's been very exciting, so far. In addition to those things Jim and I hope you share with us for this publication, I hope you let me know if you have anything you think, especially beyond the areas of DTV, HDTV etc., that might be of interest to the BE readers as well. I am hoping that my new contacts will help provide information for this venue as well.
Larry
*************
(Ed note: Because of the similarities of HDTV and Electronic Cinema, as mentioned earlier, we have included this presentation for your information.)
Electronic Cinema
by Jim Mendrala
(c) 1995, 1996, 1997 All rights reserved.
Electronic Cinema is being discussed and worked on by several American and foreign companies. Theater goers, in the not to distant future, will see bright, high resolution images of at least 35mm quality, projected onto the big wide screen. This idea of converting movies into a high-resolution, digital bit stream (or packets of data) and delivering that bit stream to theaters with quality as good as the film itself has the major motion picture studios turning an eye towards the economic realities of electronic distribution - Electronic Cinema.
Electronic Cinema can bring a number of important advantages to the evolution of the film industry. First is the amount of time and money saved, an especially important consideration to the producer and/or distributor. Second, Electronic Cinema can equal or better the very best cinema of today with its 35mm or 70mm quality images. Technically, with today's technology, there isn't any reason why film images (limited only by the film itself) cannot be projected electronically onto the big screen.
Some believe that HDTV is the cinema of the future. But HDTV has some inherent problems as we shall see. Its resolution, though high, is only capable of super-16mm quality. The incompatible frame rate, inadequate bandwidth, inadequate number of scan lines, interlaced fields, and bandwidth limited color, all combine to stymie efforts at real qualitative improvements in image quality for the wide-screen cinema. The broadcast-imposed standards of the television industry have tended to thwart the application of a time- and money-saving video technology to a major field of application - namely Electronic Cinema.
Today, movies are printed and the prints are sent via land, sea, and air to the various theater chains. Even though security is tight, piracy of the print to be shown does happen. Prints for a typical movie are expensive. They can average as much as $2700 per print copy. Prints also show wear and tear. Prints get scratched and cinched. Prints break and have to be spliced back together while the audience sits and waits. Sometimes print reels inadvertently get shown out of sequence. Also, when switching from one reel to another, the film might be projected out of focus, sometimes for only a short time, sometimes for the length of the whole reel.
In the near future, a whole new way of delivering movies will evolve. Films today, with existing HDTV technology, can be transformed into a digital signal using either the NHK 1125/60 HDTV system or the European 1250/50 HDTV system. With the 1125/60 HDTV system, the frame rate is 30 frames per second (fps). This is not desirable, since film in the U.S. is shot and projected at 24 fps.
The 1125/60 HDTV telecine must convert the 24 fps to 30 fps using the 3:2 pulldown technique. Not a very good idea. Various digital compression schemes, such as MPEG, have ways to look only at the actual 24 fps, thus freeing up some of the time that would be wasted on compressing a video frame made up of one field of the previous film frame and one field of the next film frame. Sometimes the 3:2 pulldown detectors get fooled. This is another reason why Electronic Cinema cannot be led by conventional HDTV television technology that is being advocated in this country.
The American Society of Cinematographers (ASC) have insisted upon a 24 fps rate. With the European 1250/50 HDTV system, the frame rate is 25 fps. This is closer to the 24 frame rate used in the U.S. and the one the (ASC) is insisting on. Film in European and other 50 Hz countries is projected at 25 frames per second. The difference between 24 and 25 is 4%. With Electronic Cinema films can be shown at their original frame rates, be it 24 fps or 25 fps.
Both of the above HDTV systems use a 2:1 interlace, yet the film is scanned progressively and converted to interlaced scan. The main reason for using the 2:1 interlace was primarily to reduce flicker. Electronic Cinema progressively scans the film. One major service bureau presently scans the film progressively and generates a digital file for every frame, with a resolution equal to or better than the film itself. We have seen the results on the wide-screen in such films as Forest Gump, Apollo 13 and others. Those digital bit streams were put back on film with no apparent loss of resolution even though they were manipulated through various types of computers.
As you can see, capturing a film digitally with quality as high as the film itself is being done today.
Let's look at a typical film projector. The projector has either a two or three blade shutter that allows each frame to be displayed two or three times per frame. Thus, what is seen on the screen is either a 48 or 72 picture-per-second picture but at a 24 fps rate. Because of the light loss with a 3 blade shutter most theaters use a two blade shutter. This gives a picture a perceptible flicker that gets worse as the amount of light is increased. Flicker tends to disappear when the display rate approaches 60 fps. With today's digital technology, to display a picture 2 or 3 times between frames is not a problem. 24 fps can be displayed at 72 (3x24) picture-per-second and 25 fps can be displayed at 75 pictures-per-second, well above where flicker tends to be perceptible. MPEG compression, in a way, does this now when the 24 frame image is decompressed and it's output displayed at 30 fps.
Most compression schemes are upward scalable. This means that if MPEG, as an example, was used, it could be scaled to do wide-screen, Electronic Cinema including scope-type films with their 2.35:1 aspect ratio and beyond.
Today, when a producer or film studio transfers a film to video, the process is very lengthy. A colorist does a scene-by-scene color correction on the film. The film from the film lab, as good as it is, is not as color correct as is required in a HDTV viewing situation. After the scenes are color corrected, the film is transferred into a digital signal and recorded. Electronic Cinema would be no exception. A colorist would be required here also. As a matter of fact, the only difference is that the film would be observed on a large screen not some small CRT type of monitor. Because of the CRT's phosphors, CRT's cannot display as much color as the film image contains. New projectors can display as much color as the film has. HDTV has reduced resolution in the color. Equal resolution color is a must for big wide-screens, something HDTV cannot deliver.
Lately, a new person has been added to the list. A compressionist. Sometimes the colorist does both functions and is known as a Compression/Colorist. That person not only optimized the color but also the compression. The main reason for the need of the compressionist is that today films are destined for bandwidth-limited systems. A CD-ROM, for example, cannot support the high data rates necessary for wide-screen, high-resolution pictures. Even the new Digital Video Disk (DVD) is only up to standard broadcast quality. The so called "Sweet Spot" in MPEG encoding for professional broadcast quality is around 6 MHz. For HDTV and wide-screen Electronic Cinema, the data rates are much higher. Today, it is not only possible to record that high amount of data with existing technology, but it is possible to distribute that data by fiber optics or, more economically, by satellite.
Today's satellites are designed for the traditional data, communication, and television type of signals. Even with today's satellites though, the much better Electronic Cinema type of pictures are possible. In the near future, satellites dedicated to Electronic Cinema movie distribution will be in place to replace the current system of distribution.
Since the signal is digital, no loss of quality would be visible to the moviegoer. And because it is digital, various encryption schemes could be employed to protect the feature from unauthorized exhibition or piracy.
Once the digital bit stream reaches the theater, either via fiber optics or satellite, it finally needs to be displayed.
Projection systems today fall into three main categories, emissive, transmissive light valve, and reflected light valve. Emissive displays are based upon cathode ray tubes (CRT) or laser technology. CRT projectors today are relatively dim and limited to approximately 1 KW power input. The best CRT projectors give a maximum light output of approximately 1,000 lumens at peak white. CRT projectors can suffer poor resolution in the corners, spot growth at high beam currents, or visible line structure at low beam currents. Laser CRT projectors offer potentially higher power and more efficient operation but depend on very low operating temperatures and the efficiency of the blue lazing material is low. Laser projectors in general require mechanical scanning and high power demands and potentially high costs. Image speckle has been an obstacle in laser projectors even though techniques can minimize this problem.
Until the Liquid Crystal Display (LCD) projector, large screen projectors were based on electron beam addressed oil films to produce high brightness projected images. These projectors require continuous adjustment during their operating life.
LCD Projectors based upon transmissive active matrix light valves, however, have efficiency losses due to polarization and partial blocking of the light path by the active matrix. They also deliver pixelized images, with low resolution, which is inherent in their design. Image lag is also a problem on some displays.
Reflective LCD light valves deliver, a very bright, pixel-free image with more than 400% more contrast and much higher resolution. The reflective LCD is addressed with a low level infrared, high resolution CRT imaged onto a layer of liquid crystal one for each of the tri-stimulus colors, RGB).
Digital Micromirror Device (DMD) projectors are reflective also but unlike CRT's or reflective LCD's, use tiny micro-size mirrors to reflect the light through a lens to the screen. Pulse width modulation modulates the intensity of the light as seen on the screen by the eye. This produces a linear modulation or unity gamma (i.e., gamma = 1.0). A DMD device for each of the primary colors (RGB) is used and being solid state, like a CCD color camera, requires only a simple one-time registration. Interlaced pictures are not suited to DMD devices, however, as only half the maximum possible picture brightness would result. By using progressive scan, the vertical spatial bandwidth is increased by about 60%. DMDs are inherently low-flicker devices with no lag. The DMD type of projector produces a picture which is similar to projected film. The technology is scalable and can provide aspect ratios of the projected image to include all known aspect ratios, such as 2.35:1, 2.2:1, 1.85:1, 1.78:1 (HDTV format), 1.66:1, and 1.33:1.
Electronic Cinema being developed today will evolve into a more efficient way of delivering to the moviegoer a bright, sharp, excellent color, high resolution, in focus, movie feature presentation with digital clarity and digital CD quality multi-track surround sound audio.
The loss of prints en route to the theater will be eliminated. Film breaks will be eliminated. Scratched or cinch marked prints will be a thing of the past. Encrypting of the digital data will make piracy extremely difficult. Satellite delivery will make distribution costs plummet. All solid state DMD type of projectors will make digital display of the movie possible and relatively maintenance free, a real advantage to the cineplex theaters of tomorrow.
Jim
************
Check out these web sites:
http://drogo.cselt.stet.it/mpet/ and
http://wwwam.HHI.DE/mpeg-video/
************
Subj: Some Closing Thoughts
By: Larry Bloomfield
The technology is moving so fast that it is difficult, at best, to stay up with things. I was once told that higher education teaches one how to ask a better question. Working with today's technology certainly is an education and, like you, I've got lots of questions. With your help, we can find most of the answers.
Today's digital video (601) comes with its own set of unique problems. It would be great if, once encoded, we never had to convert back to analog until we got to our final destination, the kinescope (picture tube). How can we reasonably, and at the same time keep cost down, approach this goal? There will be a whole new generation of equipment on the market in the not too distant future. NAB will be, without question, an education this year.
Here's another area of interest. There are only a finite number of satellites out there. No doubt there are more in cue to be launched and positioned. It would be interesting to know what the demands will be in the future for channel space on those highflying communications devices. With compression we can get much more on a single bird, but with multicasting, Hi-Defination, etc., the need will gobble up bandwidth like a pack of sharks on a feeding frenzy. I often wonder if Dr. Harold Rosen or any of the rest of my associates with Project Syncom ever thought what we were doing back in 1963-64 would ever evolve into this? We sure thought it was hot when we were able to tell within three meters of where Syncom II & III were (at 22,300 miles) back in those days. I wonder how much more accurate they are today?
A great deal of what is aired today originates on film. A producer must look at how he is going to get his film turned into an electronic signal with the greatest degree of fidelity. Will the target display system be in a theater, someone's living room or on a computer screen? Will the material be stored on tape, DVD or some, as yet undeveloped media? The term telecine has almost totally disappeared at the local level in this day and age and this is where the transfer must start. There are several techniques of scanning the frames of the original. The concept of film-to-tape transfer is becoming obsolete. There is other storage media. What must take its place is the concept of film-to-data transfer. It is only logical that Hi-Definition television will demand that the thousands upon thousands of feet of filmed material that exists as NTSC on tape today must be retransferred to data with the highest possible rate of pixel capture our technology can provide. This transfer process should only have to be done once so that the any of several target destinations can be accommodated -- theater, home entertainment such as DVD, a person's computer or part of a broadcast.
The move to digital television is a given. The timetable has been set and now... the questions. These questions can be applied to both local and network television, in very large towns to the smallest of markets. Who is going to do what? The areas of primary concern are Program sources and material. What changes need we make in our approach to creative services in all areas, especially production of commercials and public service messages? Can sales convince the client the new way is better or necessary? What about News gathering and presentation? Then there are the technical means of getting all of this to the public. These are the most obvious issues. The trick is to find the not- so-obvious issues, address them and find solutions.
No one person or company has all the answers, but together we can move foreword like electronic Magellan's; circumnavigate this world of digital television and bring, what was once thought to be flat, into true global prospective. I would like to see the pages of this part of Broadcast Engineering used as the navigational charts, wherein we can share the hazards and routes to safely sail from one issue to the next. Therefore, your input is most essential. Let me know what you are up to and can share with the rest of us.
Larry
****************
The DTV Tech Notes are published for broadcast professionals who are interested in DTV, HDTV etc. by Larry Bloomfield and Jim Mendrala. We can be reached by either e-mail or land lines (541) 385-9115, (805) 294-1049 or fax at (805) 294-0705. News items, comments, opinions etc. are always welcome from our readers; letters may be edited for brevity.
>>>>> larrybend@aol.com --------- J_Mendrala@compuserve.com <<<<<
DTV Tech Note articles may be reproduced in any form provided they are unaltered and credit is given to the DTV Tech Notes and the originating authors, when named.
END
