poynton@vector.Sun.COM (Charles A. Poynton) (08/11/89)
Here's a contribution that may unravel some of the confusion surrounding HDTV, ATV. IDTV, EDTV, and so on. The most important point is that HDTV relates to production and exchange of programs, and has immediate relevance to computer graphics. Advanced television (ATV) concerns the delivery of entertainment programming to consumers, and has very little to do with computing. To me it is ludicrous to suggest that is strategically important for the U.S. to resurrect domestic colour television receiver manufacturing. Even more ridiculous is the suggestion (by George Gilder in Fortune) that the U.S. can, in a year or two, leapfrog current [Japanese] HDTV technology. Let's get on with exploiting HDTV technology in computing, where we've ALREADY got a vital and profitable industry! (Or should I say, WHILE we've still got it?) The preceeding was my opinion. What follows is fact. C. p.s. 1250/24, 2048-by-1152, 74.25 MHz. Dare to be square! ----- Charles A. Poynton Sun Microsystems Inc. <poynton@sun.com> 2550 Garcia Avenue, MS 8-04 415-336-7846 Mountain View, CA 94043 "Japan has no laws against damage to its flag, but it has strict laws forbidding the burning of foreign flags lest this give offense to the country in question." -- The Economist, July 1, 1989, p. 19. ----- High Definition Television (HDTV) and Advanced Television (ATV) Glossary Charles A. Poynton, Sun Microsystems, Inc. TN32 89/08/07 18:26 525-LINE, 625-LINE TELEVISION TERMS 525/59.94/2:1. A raster scanning standard used primarily in North America and Japan, having 525 total lines (of which approximately 483 contain picture information), a field rate of 59.94 Hz, and interlace. Without the "/2:1" notation, interlace is implicit. Colour in 525/59.94 systems is commonly encoded using the NTSC method. Often colloquially referred to as 525/60, and often incorrectly denoted NTSC. 625/50/2:1. A raster scanning standard used primarily in Europe and Asia, having 625 total lines (of which 575 contain picture information), a field rate of 50 Hz, and interlace. Without the "/2:1" notation, interlace is implicit. Colour in 625/50 systems is usually encoded using the PAL method (although France, the USSR, and certain other countries use SECAM). Often incorrectly denoted PAL or SECAM. NTSC, National Television Systems Committee. (1) The group which in 1953 established 525-line, 2:1 interlaced, 59.94 Hz field rate, composite colour television signals in the U.S. More properly referred to as NTSC-II. [The original NTSC, now properly referred to as NTSC-I, established 525-line, 2:1 interlaced, 60.00 Hz field rate monochrome television in the U.S. in 1943.] (2) A method of composite colour encoding based on quadrature modulation of U and V colour difference signals onto a colour subcarrier. Used only in 525/59.94 systems, with a subcarrier frequency of about 3.579455 MHz. PAL, Phase Alternate Line. A method of composite colour encoding similar to NTSC, except that the phase of the V-axis colour difference signal inverts at the horizontal line rate. Commonly used in 625/50 systems with a subcarrier frequency of about 4.433618 MHz, but also used with a subcarrier of about 3.579455 MHz in the PAL-N system (e.g. Argentina), and with 525/59.94 scanning and a subcarrier frequency of about 3.575611 MHz in the PAL-M system (e.g. Brazil). SECAM, Sequential Couleur avec Memoire. A method of composite colour encoding using line-alternate U and V colour difference signals, frequency modulated onto a subcarrier. Used only in 625/50 systems (e.g. France, USSR). Component. A video system which conveys colour using three separate signals. Examples are RGB, YUV, MAC. Composite. A video system which uses the spectral interleaving (or frequency interleaving) technique to encode (combine) luminance and colour information into a single signal. Examples are NTSC, PAL, SECAM. S-video, S-connector, YC3.58, YC4.43. An interface which conveys luminance, and quadrature modulated chrominance, as two separate signals on a specific 4-pin mini-DIN connector. There are only two types of S-video: YC3.58 which has a 525/59.94 raster and chrominance modulated as in NTSC, and YC4.43 which has a 625/50 raster and chrominance modulated as in PAL. S-video is a form of component video, in that the three components are completely separable. MAC, Multiplexed Analog Component. A video system which transmits three colour components, usually YUV, in time-compressed serial analog form. Interlace. A video signal in which alternate raster lines of a frame are separated into two fields displaced in time by half the frame time. Also called 2:1 Interlace. Examples are 525/59.94/2:1, 625/50/2:1, 1125/60.00/2:1. Progressive. A video signal in which all scan lines of a complete frame are closely related in time. Also called 1:1 Interlace, Sequential, Non-interlaced, Pro-scan. Examples are 525/59.94/1:1, 1250/24/1:1. IMPROVED, EXTENDED, ADVANCED, HIGH DEFINITION TELEVISION TERMS IDTV, Improved Definition Television. A television system which offers picture quality substantially improved over conventional receivers, for signals originated in standard 525-line or 625-line format, by processing which involves the use of field store and/or frame store (memory) techniques at the receiver. One example is the use of field or frame memory to implement de-interlacing at the receiver, to reduce inter-line twitter compared to that of an interlaced display. IDTV techniques are implemented entirely at the receiver, and involve no change to picture origination equipment and no change to emission standards. EDTV, Extended Definition Television. A television system which offers picture quality substantially improved over conventional 525-line or 625-line receivers, by employing techniques at the transmitter and at the receiver which are transparent to (and cause no visible quality degradation to) existing 525-line or 625-line receivers. Examples are improved luminance/colour separation made possible by pre-combing the transmitted signals such as has been suggested by Faroudja, Central Dynamics, and Dr William Glenn, in order to reduce or eliminate NTSC artifacts such as dot crawl and hanging dots. Another example is the use of progressive scanning at the camera, interlaced transmission, and reconstruction of a progressive display at the receiver to reduce or eliminate interlace artifacts, as in the Faroudja SuperNTSC system. EDTV systems require changes in picture origination equipment, but are completely compliant with current emission regulations. Picture Aspect Ratio. The ratio of picture width to picture height. Usually abbreviated to Aspect Ratio. Current 525-line and 625-line systems both have a picture aspect ratio of 4:3. Letter-box. A television system which limits the recording or transmission of useful picture information to about three-quarters of the available vertical picture height of the distribution format (e.g. 525-line), in order to offer program material which has a wide aspect ratio. Wide-screen. A television system which offers a picture aspect ratio substantially wider than 4:3, using the full vertical picture area afforded by the distribution signal format (e.g. 525-line). Emission regulation changes are required for wide-screen TV transmission. Wide- screen may or may not be combined with EDTV; wide-screen is inherent in ATV and HDTV. ATV, Advanced Television. A system which distributes wide-screen television signals with resolution substantially improved over 525-line and 625-line systems. Terrestrial ATV broadcasting (VHF/UHF) would require substantial changes to current emission regulations. There is general acknowledgement that any ATV distribution system should offer at least stereo (two channel) audio, of CD quality. HDTV, High Definition Television. A system which has approximately twice the horizontal and vertical resolution of current 525-line and 625-line television, a frame rate of at least 24 Hz, component colour coding (e.g. RGB or YUV), and a picture aspect ratio of 16:9. PRODUCTION, EXCHANGE, DISTRIBUTION TERMS HDTV Production. The original creation and editing of HDTV program material. HDTV Exchange. The interchange of HDTV program material among production, distribution, and transmission organizations. Editing at the exchange stage is limited to insertion of segments such as commercials, and lengthening or shortening the duration of program material up to 5%. ATV Distribution. The distribution of ATV program material to the ultimate viewing audience. Distribution may be by physical media such as videotape or videodisc, or by transmission (see below). ATV Transmission. The distribution of ATV program material to the ultimate viewing audience through RF media such as terrestrial VHF/UHF broadcasting, cable television (CATV), or direct broadcast satellite (DBS). ATV TRANSMISSION TERMS Channel-compatible ATV. A system for transmitting ATV through VHF/UHF or CATV media which has carrier-frequency assignments conformant to current 525-line or 625-line television transmission. According to FCC Docket 87-268, terrestrial VHF/UHF ATV transmissions in the U.S. are constrained to be channel-compatible with 6 MHz channels. Channel- compatibility does not necessarily require single-channel transmission. Receiver-compatible ATV. An ATV transmission system which provides ATV program material, possibly with reduced aspect ratio, to current 525-line or 625-line receivers. According to FCC Docket 87-268, ATV transmissions in the U.S. are constrained to be receiver-compatible. Receiver-compatibility can be accomplished by single-channel ATV, augmentation-channel ATV, or simulcasting. Single-channel ATV. An ATV transmission system which alters the 525-line or 625-line broadcasting standard by the addition of augmentation signals within the channel bandwidth of the current standard. Such systems may degrade reception of current signals. Sarnoff ACTV-I is an example of a single-channel ATV system. Augmentation-channel ATV. An ATV transmission system which transmits an augmentation signal associated with a main channel 525/59.94 or 625/50 signal, in order that signals from both channels can be combined in an ATV receiver to form an ATV signal. Augmentation-channel systems provide receiver-compatibility by default. Augmentation channels of 3 MHz and 6 MHz have been proposed for the U.S. Examples of augmentation-channel ATV systems are N.A. Philips HDS-NA, and NYIT VISTA. Sarnoff ACTV-II is an example of a hybrid single- channel/augmentation-channel system, because it transmits augmentation information in both the main NTSC channel and a separate augmentation channel. Simulcast ATV. A system which transmits a complete ATV signal within an RF channel of the same bandwidth as current 525-line or 625-line broadcasts, and which achieves receiver-compatibility through simulcast of the same program material, possibly with reduced aspect ratio, in a separate standard channel. Zenith SC-TV is an example of such a system, which in this case is optimized to exploit a currently-unused VHF/UHF taboo channel to convey an entire ATV signal. Sometimes confused with Incompatible ATV. Incompatible ATV. A system which transmits a complete ATV signal in a format not intimately related to existing broadcast standards. An example of an incompatible system is NHK MUSE-9. An incompatible ATV system is channel-compatible if it transmits the ATV signal within a 6 MHz channel. An incompatible ATV system is receiver-compatible if it employs simulcast of the same program material, possibly with reduced aspect ratio, in a separate NTSC channel. DIGITAL HDTV TERMS CIF, Common Image Format. The standardization of the structure of the samples which represent picture information in digital HDTV, independent of frame rate and sync/blanking structure. CDR, Common Data Rate. The standardization of a single data rate for digital HDTV, applicable to two or more different frame rates. CFR, Common Frame Rate. The standardization of a single frame rate for digital HDTV. Sample aspect ratio. The ratio of horizontal sample pitch to vertical sample pitch. A sample aspect ratio of unity achieves square pixels. Orthogonal sampling. Sampling a digital HDTV picture with an array of samples placed on a regular two-dimensional array. Offset sampling. Sampling a digital HDTV picture with an array of samples where alternate rows of samples are displaced by half of the pitch of the samples along that axis. Offset sampling may be applied to any combination of the horizontal/vertical/temporal sampling axes of a video signal. Also called Quincunx sampling. Offset sampling in the vertical/temporal plane is more commonly called interlace. -----
poynton@vector.Sun.COM (Charles A. Poynton) (08/12/89)
A correspondent writes: >> p.s. 1250/24, 2048-by-1152, 74.25 MHz. Dare to be square! > Why frame rate of 24? Given there is movement in the film industry to > move to 30, ... A few people in Hollywood proposed 30 Hz film, and SMPTE had a study group on it, but there was never any popular support behind the idea. Among other things, - 24 Hz is quite sufficient for motion rendition, - 30 Hz consumes more film stock (tied to the price of silver!), - international program exchange would suffer (3 G$ U.S. trade surplus in exported movies), - few commercial projectors are capable of 30 Hz without modification. All in all, just no good reason to do it. > ... why burden TV with a slower rate. Ah, wait a minute here, we want to burden TVs with a slower rate because we can't afford the bandwidth to raise it! Keep in mind that in the olden days one had to choose a frame rate which simultaneously satisfied motion rendition AND flicker constraints. The fact of living rooms being (on average) brighter than movie theatres forced television in 1941 to adopt a 30 Hz frame rate. With framestore technology, these issues can be separated. > Look to the future, not the past. The future is now. Sony and Hitachi are recording digital HDTV on one-inch magtape at 1.188 Gb/s, a full order of magnitude higher than the best available U.S. technology (the Ampex D-2 machine at 114 Mb/s). Sony are shipping 2k-by-2k Trinitrons when the best Zenith can do is 640-by-480. That's a factor of eight. Sure we could glibly standardize double the horizontal and vertical resolution of HDTV but that would be a pointless theoretical exercise unless we can build the stuff. The best way for us to get back into building the stuff is to exploit the commercially-available Japanese technology -- now. C. ----- Charles A. Poynton Sun Microsystems Inc. <poynton@sun.com> 2550 Garcia Avenue, MS 8-04 415-336-7846 Mountain View, CA 94043 "There's no offense where none is taken." -- Ancient Chinese proverb -----
poynton@vector.Sun.COM (Charles A. Poynton) (08/12/89)
A correspondent writes: >> 1250/24, 2048-by-1152, 74.25 MHz. Dare to be square! > I don't understand how 1250 becomes 2048-by-1152, and what the active > elements stuff is about and how 1250/24 becomes 74.25 MHz. Well, briefly, you get to pick five numbers and the rest fall out of the wash. The best numbers to work with are sample rate, total and picture samples per line, and total and picture lines per frame. The best numbers to publish comprise that set, except publish frame rate instead of total samples per line. Total samples per line needs to be roughly 15% more than picture samples to accommodate horizontal scanning, Total lines per frame needs to be roughly 4% more than picture lines per frame to accommodate vertical scanning. Frame rate just falls out as sample rate, divided by total samples per line, divided by total lines per frame. So in my proposal, total samples per line would be 2475, about 20% greater than 2048, but you can derive the 2475 from the five numbers I gave. You really need to see the trademark Poynton raster diagram to make this all clear. How much can I do with ASCII text ... 74.25 MHz 2475 +---------------------------------------+ | V blank | | +-------------------------------+ | | 2048 | | | | | | | 1250 | H | 1152 picture | | blank | | | | (16:9 aspect ratio) | | | | | | | +-------+-------------------------------+ Video monitors are usually specified by horizontal line rate (easily derived as sampling frequency divided by total samples per line), and frame rate. Unfortunately many frame buffers (whoops, graphics cards) specify some rounded versions of H and V rates, and it's a real nuisance to work backwards to try to figure what the real parameters are. My 1250/24, 2048-by-1152, 74.25 MHz HDTV proposal is, briefly, - use 1250 total lines and 1152 picture lines, as in current Eu95 proposals, to appeal to the Europeans; - use the proposed Japanese [and semi-recommended SMPTE 240M] sampling frequency of 74.25 MHz; - use 2048 samples per picture width and square pixels to appeal to the computer industry; - use 24 Hz to be fully conformable to film, for the Hollywood production community and to utilize current film libraries which will be the source of 90% of the initial ATV programming, with ABSOLUTELY NO temporal artifacts; - accomplish down-conversion to current broadcast standards in exactly the way it is done today from film, that is, 3-2 pulldown 0.1% slow for 525/59.94 and 2-2 pulldown 4% fast for 625/50, with artifacts identical to those seen in today's film transfers; and - display at 72 Hz to satisfy even the most flicker-sensitive, Euro-gonomic, high-ambient viewers, including computer workstation users. Of course a 2048-by-1152 monitor at 72 Hz is quite a way off, so this should mollify the U.S. broadcasters, who are concerned that an early entry into ATV could cost them a lot of money. Don't laugh about this being a political compromise. This is serious business! C. ----- Charles A. Poynton Sun Microsystems Inc. <poynton@sun.com> 2550 Garcia Avenue, MS 8-04 415-336-7846 Mountain View, CA 94043 "As at the ski resorts where girls go looking for husbands, and husbands go looking for girls, the situation is not as symmetrical as it might seem at first." -- attributed to Alan Kay -----
bill@bilver.UUCP (Bill Vermillion) (08/12/89)
In article <121076@sun.Eng.Sun.COM> poynton@vector.Sun.COM (Charles A. Poynton) writes: >A correspondent writes: > >> Why frame rate of 24? Given there is movement in the film industry to >> move to 30, ... > >A few people in Hollywood proposed 30 Hz film, and SMPTE had a study group >on it, but there was never any popular support behind the idea. Among >other things, A friend of mine does film production for video. The commercial industry is using 30 fps film for video transfer. Your other reasons (deleted for space considerations) are definately valid for theatrical release. There is (and has been demonstrated) a "Hi-Fi" film format (I forget the name) that uses FILM at 60 FPS. Results are said to be startling. I would like to see it, at least once. -- Bill Vermillion - UUCP: {uiucuxc,hoptoad,petsd}!peora!tarpit!bilver!bill : bill@bilver.UUCP
ggs@ulysses.homer.nj.att.com (Griff Smith) (08/14/89)
In article <121076@sun.Eng.Sun.COM>, poynton@vector.Sun.COM (Charles A. Poynton) writes: > A few people in Hollywood proposed 30 Hz film, and SMPTE had a study group > on it, but there was never any popular support behind the idea. Among > other things, > > - 24 Hz is quite sufficient for motion rendition, ... > All in all, just no good reason to do it. For years, I had wondered why motion looked `real' on television, and artificial on film. Then I began to see some film that had been done at 30 Hz, and realized that the faster frame rate made the motion much more realistic. Furthermore, to me there is an additional dramatic improvement in going from 30 Hz to 60 HZ. 50Hz television won't do, it flickers too much. > > ... why burden TV with a slower rate. > > Ah, wait a minute here, we want to burden TVs with a slower rate because > we can't afford the bandwidth to raise it! I assume this is part of the attempt to get rid of interlace so the computer graphics folks can avoid motion artifacts. I already have to put up with motion artifacts while watching film on television, why do I have to lose realistic motion on recorded television just so computer graphics can look better? What is so evil about interlace? > ----- > Charles A. Poynton Sun Microsystems Inc. > <poynton@sun.com> 2550 Garcia Avenue, MS 8-04 > 415-336-7846 Mountain View, CA 94043 -- Griff Smith AT&T (Bell Laboratories), Murray Hill Phone: 1-201-582-7736 UUCP: {most AT&T sites}!ulysses!ggs Internet: ggs@ulysses.att.com
twhlai@watdragon.waterloo.edu (Tony Lai) (08/14/89)
In article <120919@sun.Eng.Sun.COM> poynton@vector.Sun.COM (Charles A. Poynton) writes: >Incompatible ATV. A system which transmits a complete ATV signal in a >format not intimately related to existing broadcast standards. An >example of an incompatible system is NHK MUSE-9. This is rather trivial, but I think you are thinking of MUSE-E, which is the incompatible system that NHK will transmit. I think MUSE-9 is an augmentation-channel system.
brown@astroatc.UUCP (Vidiot) (08/14/89)
In article <12027@ulysses.homer.nj.att.com> ggs@ulysses.homer.nj.att.com (Griff Smith) writes:
<
<For years, I had wondered why motion looked `real' on television, and
<artificial on film. Then I began to see some film that had been done
<at 30 Hz, and realized that the faster frame rate made the motion much
<more realistic. Furthermore, to me there is an additional dramatic
<improvement in going from 30 Hz to 60 HZ. 50Hz television won't do, it
<flickers too much.
The problem with transferring 24 fps film to video is the 59.94 Hz field
rate. The result is known as the 3-2 pulldown method, whereby one frame
of film is scanned for three field of video and then the next frame of film
is scanned for two fields of video. The process is repeated. This results
in some of the jerkiness that you see.
You tend to see alot of videos shot using 30 fps and I've seen some TV
shows shot at 30 fps.
--
harvard\ att!nicmad\
Vidiot ucbvax!uwvax..........!astroatc!brown
rutgers/ decvax!nicmad/
ARPA/INTERNET: brown%astroatc.UUCP@spool.cs.wisc.eduphil@diablo.amd.com (Phil Ngai) (08/14/89)
In article <278@bilver.UUCP> bill@.UUCP (Bill Vermillion) writes: |There is (and has been demonstrated) a "Hi-Fi" film format (I forget the name) |that uses FILM at 60 FPS. Results are said to be startling. I would like to |see it, at least once. Much as I hate to disagree with Charles, I'd like to add that there have been experiments with faster frame rates. I personally saw something at the Vancouver World Expo (Showscan?) and was impressed. Does anyone know if IMAX or OMNIMAX use higher frame rates, or just more film? I would not disagree that it is quite questionable whether we can afford such high data rates, I'm just saying that I don't think they would be wasted if we could. -- Phil Ngai, phil@diablo.amd.com {uunet,decwrl,ucbvax}!amdcad!phil "Nonviolence works! India hasn't had to use their nuclear weapons yet."
root@conexch.UUCP (Larry Dighera) (08/15/89)
In article <278@bilver.UUCP> bill@.UUCP (Bill Vermillion) writes: > >There is (and has been demonstrated) a "Hi-Fi" film format (I forget the name) >that uses FILM at 60 FPS. Results are said to be startling. I would like to >see it, at least once. Showscan Film Corporation, 3939 Landmark Street, Culver City, CA 90232-2315, (213) 558-0150 owns the patent on 60 frame per second cinematography. They market 70mm theaters with computer controled, hydraulically actuated seating, as well as Imax theaters. The realism is awesome! I recently had the opportunity to experience a demonstration film of a sports car racing down hill in the rural European mountains. I nearly wet the seat. Larry Dighera -- USPS: The Consultants' Exchange, PO Box 12100, Santa Ana, CA 92712 TELE: (714) 842-6348: BBS (N81); (714) 842-5851: Xenix guest account (E71) UUCP: conexch Any ACU 2400 17148425851 ogin:-""-ogin:-""-ogin: nuucp UUCP: ...!uunet!spsd!conexch!root || ...!ccicpg!conexch!root
mostelle@nprdc.arpa (Jim Mosteller) (08/15/89)
In article <26754@amdcad.AMD.COM> phil@diablo.AMD.COM (Phil Ngai) writes: (text deleted) >Does anyone know if IMAX or OMNIMAX use higher frame rates, or just >more film? (text deleted) Both IMAX and OMNIMAX pass film at a "normal" 24 FPS, but strobe each frame twice, giving an effective 48 images per second -- smoother, less flicker to watch. As I remember it, Showscan not only passes film at 60 FPS, but also double-strobes for an amazing 120 images per second. At that rate, the eye never has to try to compensate for any flicker at all -- _very_ smooth; very impressive. Disclaimer: It's been a few years since seeing a Showscan demo; my memory may be off, or not reflect their current art. -- Jim Mosteller mostelle@nprdc.navy.mil
malloy@nprdc.arpa (Sean Malloy) (08/15/89)
In article <26754@amdcad.AMD.COM> phil@diablo.AMD.COM (Phil Ngai) writes: >Does anyone know if IMAX or OMNIMAX use higher frame rates, or just >more film? As I understand it from the information that the Fleet Space Theater here in San Diego hands out, the Imax cameras use one of the stock film sizes (70mm or 135mm), but with the frames turned sideways on the film (so that instead of the frames being aligned top-to-bottom, they are side-to-side), and each frame is several times the size of a normal 70mm frame. The camera uses the normal film rates (although they did have it cranked up for the ShowScan presentation some months ago); they wind up using more linear feet of film because of the larger frame size. Sean Malloy | "The proton absorbs a photon Navy Personnel Research & Development Center | and emits two morons, a San Diego, CA 92152-6800 | lepton, a boson, and a malloy@nprdc.navy.mil | boson's mate. Why did I ever | take high-energy physics?"
tuna@athena.mit.edu (Kirk 'UhOh' Johnson) (08/15/89)
In article <26754@amdcad.AMD.COM> phil@diablo.AMD.COM (Phil Ngai) writes: > >Does anyone know if IMAX or OMNIMAX use higher frame rates, or just >more film? in the one OMNIMAX film i've seen (at the museum of science, in boston), the temporal aliasing was easily noticable, so i would tend to doubt that it uses a higher frame rate. just a very wide projection angle. kirk
ksbooth@watcgl.waterloo.edu (Kelly Booth) (08/15/89)
In article <3272@arctic.nprdc.arpa> mostelle@nprdc.navy.mil (Jim Mosteller) writes: >Both IMAX and OMNIMAX pass film at a "normal" 24 FPS, but strobe each >frame twice, giving an effective 48 images per second -- smoother, >less flicker to watch. Normal film is shown like this. Sound film is 24 fps with each frame shown twice (for 48/s) and silent film is 16 fps with each frame shown thrice (for 48/s).
ingoldsb@ctycal.COM (Terry Ingoldsby) (08/16/89)
In article <12027@ulysses.homer.nj.att.com>, ggs@ulysses.homer.nj.att.com (Griff Smith) writes: > I assume this is part of the attempt to get rid of interlace so the > computer graphics folks can avoid motion artifacts. I already have to > put up with motion artifacts while watching film on television, why do > I have to lose realistic motion on recorded television just so computer > graphics can look better? What is so evil about interlace? > Interlace can be a real pain if you want to draw thin horizontal lines, or diagonal lines that cause only a single pixel to be illuminated on a scan line. In these cases the refresh rate is only the frame rate (eg. 30 Hz), not the field rate, and flicker becomes quite annoying. You can occasionally see this on poorly designed text overlaid during TV sports programs. The flicker can be very visible. -- Terry Ingoldsby ctycal!ingoldsb@calgary.UUCP Land Information Systems or The City of Calgary ...{alberta,ubc-cs,utai}!calgary!ctycal!ingoldsb
3ksnn64@pur-ee.UUCP (Joe Cychosz) (08/17/89)
In article <3273@skinner.nprdc.arpa> malloy@nprdc.arpa (Sean Malloy) writes: > >As I understand it from the information that the Fleet Space Theater >here in San Diego hands out, the Imax cameras use one of the stock >film sizes (70mm or 135mm), but with the frames turned sideways on the >film (so that instead of the frames being aligned top-to-bottom, they >are side-to-side), and each frame is several times the size of a >normal 70mm frame. Omnimax and Imax use 70mm film. And yes the image is allong the sprocket holes instead of across. The size for the frame is approximately 70mm by 50mm. I have seen a longer format in which a frame was about 2.5 times longer than the standard Omnimax frame. I believe that this was used in the GE pavilion at Epcot Center.
ggs@ulysses.homer.nj.att.com (Griff Smith) (08/17/89)
In article <428@ctycal.UUCP>, ingoldsb@ctycal.COM (Terry Ingoldsby) writes: > In article <12027@ulysses.homer.nj.att.com>, ggs@ulysses.homer.nj.att.com (Griff Smith) writes: > > I assume this is part of the attempt to get rid of interlace so the > > computer graphics folks can avoid motion artifacts. > > ... What is so evil about interlace? > > > Interlace can be a real pain if you want to draw thin horizontal lines, or > diagonal lines that cause only a single pixel to be illuminated on a scan > line. In these cases the refresh rate is only the frame rate (eg. 30 Hz), > not the field rate, and flicker becomes quite annoying. You can > occasionally see this on poorly designed text overlaid during TV sports > programs. The flicker can be very visible. > -- > Terry Ingoldsby ctycal!ingoldsb@calgary.UUCP > Land Information Systems or > The City of Calgary ...{alberta,ubc-cs,utai}!calgary!ctycal!ingoldsb I don't think I've ever noticed this. If this is the kind of argument that is being used, I think the industry is putting something over on us. I HAVE noticed the following: to 1) When watching film on TV, pans get doubled (probably also tripled) images because the stationary images get frozen on my retina while I try to follow the apparent motion. If a 24 fps HDTV standard is adopted, with triple scanning of frames, I'm going to see tripled images any time the image pans. 2) When watching film on film, pans are horribly blurred because the image DOESN'T get frozen on my retina while I follow the apparent motion. IMAX is just as bad as the others. Does anyone know of attempts to build strobed theater projection systems that avoid this problem? Interlace seems to be an excellent way to cut the bandwidth in half while avoiding flicker and motion artifacts. Other than pressure from the movie industry, what are the other arguments for eliminating interlace? -- Griff Smith AT&T (Bell Laboratories), Murray Hill Phone: 1-201-582-7736 UUCP: {most AT&T sites}!ulysses!ggs Internet: ggs@ulysses.att.com
dya@unccvax.UUCP (York David Anthony @ WKTD, Wilmington, NC) (08/18/89)
In article <12045@ulysses.homer.nj.att.com>, ggs@ulysses.homer.nj.att.com (Griff Smith) writes: > Interlace seems to be an excellent way to cut the bandwidth in half > while avoiding flicker and motion artifacts. Other than pressure > from the movie industry, what are the other arguments for eliminating > interlace? 1. Interlace causes a line crawl artifact which is extremely obvious and highly annoying. 2. Interlace lowers the vertical resolution for a given number of TV lines. This is principally due to intratarget leakage in tube type cameras, and the precharge/decay time characteristics of the CRT phosphour. ("Deinterlacing" by using a long persistance phosphour works, but screws up motion royally.) 3. Interlace requires the ability to retrigger the vertical oscillator with much greater precision than progressive scan, in order that the "odd" lines fall exactly inside the "even" lines; 4. Odd/even field housekeeping can sometimes be a pain in the butt when it comes to designing things like time base correctors and such. Why bother? Yes, interlacing does save approximately half the bandwidth, but it also cuts the information content in the diagonal and vertical domain. York David Anthony BPH-880505OT (WRPL) Wadesboro, NC
brian@apt.UUCP (Brian Litzinger) (08/18/89)
From article <12045@ulysses.homer.nj.att.com>, by ggs@ulysses.homer.nj.att.com (Griff Smith): > In article <428@ctycal.UUCP>, ingoldsb@ctycal.COM (Terry Ingoldsby) writes: >> In article <12027@ulysses.homer.nj.att.com>, ggs@ulysses.homer.nj.att.com (Griff Smith) writes: >> > I assume this is part of the attempt to get rid of interlace so the >> > computer graphics folks can avoid motion artifacts. >> > ... What is so evil about interlace? >> > >> Interlace can be a real pain if you want to draw thin horizontal lines, or >> diagonal lines that cause only a single pixel to be illuminated on a scan >> line. In these cases the refresh rate is only the frame rate (eg. 30 Hz), >> not the field rate, and flicker becomes quite annoying. You can >> occasionally see this on poorly designed text overlaid during TV sports >> programs. The flicker can be very visible. > > I don't think I've ever noticed this. If this is the kind of argument > that is being used, I think the industry is putting something over on > us. I've noticed interlaced video flicker quite abit. I've been involved in high resolution graphics (>1024x1024) since 1983. Flicker from interlaced video comes from two major causes. One, horizontal lines or diagonal lines with long segments that are horizontal tend to flicker because they are refreshed at have the effective rate of the overall screen. Blue usually flickers the most followed by red, then green. Two, the interfield registration of the two interlaced fields. Some people call this jitter. Your ability to sense flicker is also affected by several things. One, the persistence of the phosphers used in the display. Long persistence displays reduce flicker problems, but are generally dimmer and lower your effective resolution because of the slower response of the phospher. Two, some companies, such as IBM, substitute a phospher which produces a color called 'sky blue' rather than 'blue' in their normal persistence monitors. This reduces flicker problems in the most "flickery" color. Three, the age of your eyeballs. Older eyeballs often have a harder time detecting flicker on displays. Similar to how older ears can have difficulty detecting higher frequency sounds. Four, knowing what flicker looks like. Often times people never notice the flicker because no one ever pointed it out. Similar to people who have never noticed the panning that goes on in movies that have been transfered to television. > Interlace seems to be an excellent way to cut the bandwidth in half > while avoiding flicker and motion artifacts. There ain't no such thing as a free lunch. You definitely give up something with interlaced video. Have you seen the TOSHIBA double scanning television? <> Brian Litzinger @ APT Technology Inc., San Jose, CA <> UUCP: {apple,sun,pyramid}!daver!apt!brian brian@apt.UUCP <> VOICE: 408 370 9077 FAX: 408 370 9291
ggs@ulysses.homer.nj.att.com (Griff Smith) (08/18/89)
In article <1612@unccvax.UUCP>, dya@unccvax.UUCP (York David Anthony @ WKTD, Wilmington, NC) writes: > In article <12045@ulysses.homer.nj.att.com>, ggs@ulysses.homer.nj.att.com (Griff Smith) writes: > > > Interlace seems to be an excellent way to cut the bandwidth in half > > while avoiding flicker and motion artifacts. Other than pressure > > from the movie industry, what are the other arguments for eliminating > > interlace? [deleted some good arguments against interlace] > Yes, interlacing does save approximately half the bandwidth, > but it also cuts the information content in the diagonal and vertical > domain. > > York David Anthony > BPH-880505OT (WRPL) Wadesboro, NC Thanks, I think I learned something. This all seems familiar, though. Back in the 1970's, people in the computer industry were moaning about the evils of using NRZI encoding for magnetic data tapes: poor clock recovery, no skew correction, etc. The solution was (flourish of trumpets) Phase Encoding. Double the bit density, but reserve half the bits for flux change references. We all thought this was wonderful. But the next revolution went back to NRZI. In the interim, the advances in electronics had made it possible to conquer the NRZI dragon. They used GCR to make the flux change density high enough to ensure proper clock recovery. I would be delighted to see 60 hz progressive scan HDTV, but I assume we can't afford the bandwidth yet. Given a choice between using 30 Hz progressive scan with motion artifacts caused by frame doubling, and using 60Hz interlaced scan with alternate fields digitally synthesized to eliminate twitter, I'll take the latter. I think a lot of the problems you describe will go away as more intelligent receiving equipment becomes available. 24 Hz is a short-sighted standard, and I'm annoyed that it's even being considered. -- Griff Smith AT&T (Bell Laboratories), Murray Hill Phone: 1-201-582-7736 UUCP: {most AT&T sites}!ulysses!ggs Internet: ggs@ulysses.att.com
ingoldsb@ctycal.COM (Terry Ingoldsby) (08/19/89)
In article <12045@ulysses.homer.nj.att.com>, ggs@ulysses.homer.nj.att.com (Griff Smith) writes: > In article <428@ctycal.UUCP>, ingoldsb@ctycal.COM (Terry Ingoldsby) writes: > > In article <12027@ulysses.homer.nj.att.com>, ggs@ulysses.homer.nj.att.com (Griff Smith) writes: > > > ... What is so evil about interlace? > > > > > Interlace can be a real pain if you want to draw thin horizontal lines, or > > diagonal lines that cause only a single pixel to be illuminated on a scan > > line. In these cases the refresh rate is only the frame rate (eg. 30 Hz), > > not the field rate, and flicker becomes quite annoying. You can > > occasionally see this on poorly designed text overlaid during TV sports > > programs. The flicker can be very visible. ... > I don't think I've ever noticed this. If this is the kind of argument The TV people are very careful to avoid this situation, so you won't see it very often. Once in a while they screw up and it is visible. ... > Interlace seems to be an excellent way to cut the bandwidth in half > while avoiding flicker and motion artifacts. Other than pressure > from the movie industry, what are the other arguments for eliminating > interlace? Interlace IS a good way to cut the bandwidth in half and still give relatively good performance wrt flicker, etc.. It is not a *perfect* solution. Interlace is not bad for low resolution images (ie, almost everything spans two horizontal lines). High res images, or those created by a computer can be a pain with interlace. These problems can be overcome at the receiver by taking the interlaced info, storing it in a frame buffer, and clocking in out (non-interlaced) at 60 frames per second. Even though the odd/even info is only changing every 1/30 of a second, it is refreshed more often and so it doesn't flicker. Motion artifacts might still be visible. Note that this is roughly equivalent to a CRT with a 1/30 sec persistence on the phosphor. -- Terry Ingoldsby ctycal!ingoldsb@calgary.UUCP Land Information Systems or The City of Calgary ...{alberta,ubc-cs,utai}!calgary!ctycal!ingoldsb
paul@moncam.co.uk (Paul Hudson) (08/20/89)
Continuing the discussion ....
When we were discussing using interlace for a computer at my previous
work, we tried various lines& shapse in interlaced & non-interlaced
modes on a few people. One thing that came from this is that the
visibility of interlace flicker varies from person to person - I'm one
of those who find it very objectionable, ggs@ulysses.homer.nj.att.com
(Griff Smith) is presumably on the other end of the scale.
--
Paul Hudson These opinions void where prohibited by law.
Until 23 August (but (e)mail will be forwarded for a while)
MAIL: Monotype ADG, Science Park, Cambridge, CB4 4FQ, UK.
PHONE: +44 (223) 420018 EMAIL: paul@moncam.co.uk,
FAX: +44 (223) 420911 ...!ukc!acorn!moncam!paul
On vacation until September 6, then
MAIL: Ing. C. Olivetti & C. Spa, Via Cristoforo Columbo, 49,
20090 Trezzano Sul Naviglio, Milano, Italy.
EMAIL: ..!mcvax!i2unix!iconet!trzdor1!paul, paul@trzdor1.ico.olivetti.com
PHONE: 39 + 2 + 445701 FAX: 39 + 2 + 4454225yost@esquire.UUCP (David A. Yost) (08/22/89)
In article <12611@pur-ee.UUCP> 3ksnn64@pur-ee.UUCP (Joe Cychosz) writes: >Omnimax and Imax use 70mm film. And yes the image is allong the sprocket >holes instead of across. The size for the frame is approximately 70mm >by 50mm. I have seen a longer format in which a frame was about 2.5 >times longer than the standard Omnimax frame. I believe that this was used >in the GE pavilion at Epcot Center. The Epcot GE pavilion uses normal Omnimax format. Imax and Omnimax are 70mm film traveling sideways, 15 perfs per frame. (Traditional 70 mm travels vertically with 5 perfs per frame. Imax is 1.33 aspect ratio, and the weird-shaped Omnimax image fits within that frame. --dave yost <12611@pur-ee.UUCP> Logged in usenet category "I'm not sure, but..." as number 137,438 (for the month of August).