pinter@castor.bucknell.edu (01/15/91)
Two questions about DVI technology:
1) When authoring an application, what happens at the final step?
Let's say you have the application working on you 600 Meg
hard drive. Is it a simple matter to transfer it into a format
that is acceptable for pre-mastering, or does this take some
special program? If so, how much does such a program cost?
2) Does anyone know how the new DVI chips announced recently will
affect the price of boards, and when? After all, if the board set
that costs $4650 today will be available for $2000 in a month, I'd
just assume wait. Also, will the new boards be better in any way?
Marco Pinter
pinter@sol.bucknell.edutj@gpu.utcs.utoronto.ca (Terry Jones) (01/15/91)
This news group is new here so I don't know if I missed a lot but here goes. From recent announcements I would assume that DVI is looking like dated technology even though the chips are just announced. I sort of feel that JPEG and MPEG hardware is going to be more widely accepted. What are others feeling?
rkl@cbnewsh.att.com (kevin.laux) (01/15/91)
In article <573@hydra.bucknell.edu>, pinter@castor.bucknell.edu writes: > Two questions about DVI technology: > > 1) When authoring an application, what happens at the final step? > Let's say you have the application working on you 600 Meg > hard drive. Is it a simple matter to transfer it into a format > that is acceptable for pre-mastering, or does this take some > special program? If so, how much does such a program cost? > > 2) Does anyone know how the new DVI chips announced recently will > affect the price of boards, and when? After all, if the board set > that costs $4650 today will be available for $2000 in a month, I'd > just assume wait. Also, will the new boards be better in any way? > 1) If you have DVI, then you have their Production Tools. There is a program called VLayout that will add padding for CD-ROM. Although I have not actually mastered a CD-ROM yet, it would appear that after processing your files with VLayout, that it is merely a matter of running SD (Nortons Speed Disk) to insure contiguous disk space and then using SY-TOS to back up the files onto an Archive tape that then can be sent to a CD-ROM house for production. 2) The new DVI chips are VLSI whereas the chips used on the ActionMedia 750 boards are from a Silicon Compiler. Of course VLSI is better by definition (smaller geometry, faster, etc). The new chips are purported to be twice as fast as the current chips. I do know about pricing and availability, but I cannot say because of proprietary disclosure agreements I signed, so I suggest that you contact/talk to Intel directly. --rkl
rkl@cbnewsh.att.com (kevin.laux) (01/15/91)
In article <1991Jan15.040230.26507@gpu.utcs.utoronto.ca>, tj@gpu.utcs.utoronto.ca (Terry Jones) writes: > From recent announcements I would assume that DVI is looking like dated > technology even though the chips are just announced. I sort of feel > that JPEG and MPEG hardware is going to be more widely accepted. > I wouldn't make that assumption at all. JPEG and MPEG are proposed standards for Still and Motion Video respectively and haven't been approved yet (although we all know that there probably won't be any major alterations). Don't forget that those who have chips based on just *one* algorithm are just going to be first, not necessarily lasting. Those who can manufacture chipsets that can run *any* algorithm are going to be the ones with the most flexibility and will last in the long run. As far as DVI is concerned, it is a chip that runs Microcode (in a Very Long Instruction format, which is why it can do 1 instruction per clock cycle). If you want to implement different algorithms on the DVI chipset, you can. This is not necessarily true of, say C-Cubed chipset, which has JPEG built into the silicon itself. --rkl
young@brahms.udel.edu (Philip Young) (01/15/91)
In article <1991Jan15.040230.26507@gpu.utcs.utoronto.ca> tj@gpu.utcs.utoronto.ca (Terry Jones) writes: >From recent announcements I would assume that DVI is looking like dated >technology even though the chips are just announced. I sort of feel >that JPEG and MPEG hardware is going to be more widely accepted. DVI technology has been available for almost two years. The MPEG standard for digital video is still one or two years away from being formally adopted. Also, Intel has announced that they will provide MPEG compatability by the mid 90's. I wouldn't call DVI dated - it's just maturing faster then MPEG. Intel has received criticism for not waiting for the MPEG standard. Some people feel Intel is attempting to force a defacto standard on the industry by beating the standards to the market. However, even though the DCT based methods are theoretically superior; Intel has a viable product whose time has come. They can't really be expected to wait for the standards commitee?
jim@newmedia.UUCP (Jim Beveridge) (01/15/91)
In article <1991Jan15.040230.26507@gpu.utcs.utoronto.ca>, tj@gpu.utcs.utoronto.ca (Terry Jones) writes: > From recent announcements I would assume that DVI is looking like dated > technology even though the chips are just announced. I sort of feel > that JPEG and MPEG hardware is going to be more widely accepted. The first chip to do JPEG is from C-Cube, and they are currently only shipping the still frame version of the chip. The real time version is still not available. Even in compressed form, the bandwidth required for a full JPEG screen far exceeds the abilities of an IBM bus to transfer. (I don't believe it to be a problem for the Apple NuBus) JPEG still requires LOTS of data moving around. To keep track of it, you pretty much require the full resources of the system to move it off the hard disk and pump it into the chip fast enough. Of course, there are ways around this problem with a private bus and private hard drives, but that is $$$. The MPEG standard is still under discussion and won't be ready for at least a year. Don't expect commercially available MPEG boards for a couple of years. DVI is shipping now, but is VERY expensive, particularly for the production level video that requires that you send a tape to Intel. The "home-brew" comperssion that the DVI chips now do is very grainy and not suitable for production. The good news is that the production level does not require almost the entire power of the CPU to keep the picture running. Jim -- "If I wanted a .sig, I would have written one"
andrew@calvin.doc.ca (Andrew Patrick) (01/16/91)
In article <573@hydra.bucknell.edu> pinter@castor.bucknell.edu writes: > >Two questions about DVI technology: > > 1) When authoring an application, what happens at the final step? > Let's say you have the application working on you 600 Meg > hard drive. Is it a simple matter to transfer it into a format > that is acceptable for pre-mastering, or does this take some > special program? If so, how much does such a program cost? I am just a beginner with DVI, but I will try to tackle this one. The DVI boards you get for a PC allow you to do edit-quality capture and compression. Once you get your application working on a hard disk, you will still need to submit your source materials (1" video tape I believe) to be captured/compressed in the DVI lab. My guess is that you submit the video source material, a story board on how it fits together, and the controlling software, but I am guessing here. You see, I have not got all the DVI docs yet. -- Andrew Patrick, Ph.D. Department of Communications, Ottawa, CANADA andrew@calvin.doc.CA andrew@doccrc.BITNET Bill Watterson for President!
eric@mcrware.UUCP (Eric Miller) (02/05/91)
In article <809@anaxagoras.ils.nwu.edu> korcuska@plato (Michael Korcuska) writes: > >It seems that by >the time CD-I has full motion video we might see DVI supporting MPEG. DVI and CD-I may well end up both supporting MPEG (I hope so!). As I have said before, DVI and CD-I will probably co-exist peacefully for many years. The main difference is how you as a publisher perceive your audience. If you are publishing "consumer" titles you should publish in a format that will be supported by 10's of millions of consumer machines. If you are publishing "computer user" titles, then you should publish on a medium that is accessible to a MAC or an IBM/PC. With any luck, the standards for Full Motion Video will be universal enough so that publishers of "cross-over" titles don't have to worry about re-mastering all of their data. (Encyclopedias, reference books, some games...) >650 megs just doesn't provide the >space for a huge amount of video and the 150KB/sec data transfer rate for >CDs doesn't leave much room for improving video quality. You can fit 72 minutes of high-quality FMV on a CD-I or DV-I disc. That seems comparable to one side of a laser disc. What with multi-disc players becoming more popular, I can't see this as a realistic limitation. We are limited currently to 170 KB/sec of video data, but I will never underestimate the power of science to pack incredible amounts of data into CDs and to decode it in real time. Eric Miller Manager, New Media Systems Microware Systems Corp
rkl@cbnewsh.att.com (kevin.laux) (02/06/91)
In article <4926@mcrware.UUCP>, eric@mcrware.UUCP (Eric Miller) writes: > In article <809@anaxagoras.ils.nwu.edu> korcuska@plato (Michael Korcuska) writes: > >It seems that by > >the time CD-I has full motion video we might see DVI supporting MPEG. > > DVI and CD-I may well end up both supporting MPEG (I hope so!). As I have said > before, DVI and CD-I will probably co-exist peacefully for many years. The > main difference is how you as a publisher perceive your audience. If you are > publishing "consumer" titles you should publish in a format that will be > supported by 10's of millions of consumer machines. > > If you are publishing "computer user" titles, then you should publish on a > medium that is accessible to a MAC or an IBM/PC. With any luck, the standards > for Full Motion Video will be universal enough so that publishers of > "cross-over" titles don't have to worry about re-mastering all of their data. > (Encyclopedias, reference books, some games...) I think CD-I will just fade away. They have taken too long and not delivered enough. The i750 chipset runs microcode that is dynamically loaded and therefore will be able to run all sorts of compression/decompression algorithms. I can't see Intel not supporting JPEG or MPEG. If they don't write the microcode, I'm sure someone else will. As for computer/consumer titles, a standalone system with the i750 chipset could automatically detect the algorithm needed and load it. I'm not too worried about different publisher's formats. > >650 megs just doesn't provide the > >space for a huge amount of video and the 150KB/sec data transfer rate for > >CDs doesn't leave much room for improving video quality. > > You can fit 72 minutes of high-quality FMV on a CD-I or DV-I disc. That seems > comparable to one side of a laser disc. What with multi-disc players becoming > more popular, I can't see this as a realistic limitation. We are limited > currently to 170 KB/sec of video data, but I will never underestimate the > power of science to pack incredible amounts of data into CDs and to decode it > in real time. While the capacity/data transfer rate of the media is a factor, the decompression time is more critical for video quality. Consider that a frame of FMV must average 5KB in order to be played back at 30 fps. How compressed is that 5KB? The latest i750 chipset provides twice the decode time than the previous generation, allowing for more sophisticated algorithms to be run to decompress the video. This will certainly improve video quality (but doesn't mean that the quality will also be twice as good). 72 minutes of FMV will fit on a CD, but CD-I can only play back into a small window (I think I was told 100 x 64 at COMDEX last November). DVI provides for full screen playback (256 x 240). Lastly, I read in the Feb 91 Byte magazine's Microbytes column, that Iterated Systems has developed a hardware/software combination to deliver FMV on a standard AT computer with a VGA screen. Their system is based on Fractal Transforms. They claim 1.5 minutes of FMV will fit on a 1.44MB floppy, 40 minutes on a 40MB hard drive, and *10* hours on a CD. You need the hardware to compress it, but only the software to decompress it. -- ________________________________________________________________________________ R. Kevin Laux Email: rkl1@hound.att.com AT&T Bell Labs Voice: (908) 949-1160 Holmdel, NJ 07733 Fax: (908) 949-0959
eric@mcrware.UUCP (Eric Miller) (02/12/91)
In article <1991Feb6.140751.14909@cbnewsh.att.com> rkl@cbnewsh.att.com (kevin.laux) writes: > I think CD-I will just fade away. They have taken too long and not >delivered enough. You may be in for a surprise this summer. The number of publishers as well as manufacturers working behind the scenes is quite impressive. > 72 minutes of FMV will fit on a CD, but CD-I can only play back into >a small window (I think I was told 100 x 64 at COMDEX last November). DVI >provides for full screen playback (256 x 240). Wrong. CD-I will also deliver the FMV into a fullscreen window. As well, it can be encoded into many different resolutions including those larger (either vertically or horizontally) than the display, allowing the application to scroll around through the image. This data can be delivered from a CD, a hard disk, or a network. Eric Miller
gjh@ecs.soton.ac.uk (Gary Hill) (02/12/91)
In <4999@mcrware.UUCP> eric@mcrware.UUCP (Eric Miller) writes: >Wrong. CD-I will also deliver the FMV into a fullscreen window. As well, >it can be encoded into many different resolutions including those >larger (either vertically or horizontally) than the display, allowing the >application to scroll around through the image. >This data can be delivered from a CD, a hard disk, or a network. Yes, FMV is in the CD-I spec, but no-one has seen an implementation yet have they! (I've certainly not heard of >screen resolution video being available). The players I have seen certainly couldn't do FMV and the companies demonstrating them made no pretence that they could. Also Phillips have said they will be supporting MPEG and that isn't even defined yet so I think it will be some time before we see FMV from CD-I. If you have facts or references to prove me wrong so I can correct the report I'm writing at the moment please send them to me. Gary Hill, gjh@uk.ac.soton.ecs >Eric Miller
rkl@cbnewsh.att.com (kevin.laux) (02/12/91)
In article <4999@mcrware.UUCP>, eric@mcrware.UUCP (Eric Miller) writes: > In article <1991Feb6.140751.14909@cbnewsh.att.com> rkl@cbnewsh.att.com (kevin.laux) writes: > > > I think CD-I will just fade away. They have taken too long and not > >delivered enough. > > You may be in for a surprise this summer. The number of publishers as well > as manufacturers working behind the scenes is quite impressive. Behind the scenes, eh? ;-) > > 72 minutes of FMV will fit on a CD, but CD-I can only play back into > >a small window (I think I was told 100 x 64 at COMDEX last November). DVI > >provides for full screen playback (256 x 240). > > Wrong. CD-I will also deliver the FMV into a fullscreen window. As well, > it can be encoded into many different resolutions including those > larger (either vertically or horizontally) than the display, allowing the > application to scroll around through the image. Could you post references for this? I wonder why they didn't tell me this at COMDEX, since I specifically asked. I also asked about frame rate, which they said was up to 24 fps (from 15-18 a couple of years ago). 24 fps is acceptable for full motion video (that's the rate at which film is played). Can CD-I deliver at 30 fps, and at what resolution? -- ________________________________________________________________________________ R. Kevin Laux Email: rkl1@hound.att.com AT&T Bell Labs Voice: (908) 949-1160
nick@abblund.se (02/13/91)
In article <4999@mcrware.UUCP> eric@mcrware.UUCP (Eric Miller) writes: >In article <1991Feb6.140751.14909@cbnewsh.att.com> rkl@cbnewsh.att.com (kevin.laux) writes: > >> 72 minutes of FMV will fit on a CD, but CD-I can only play back into >>a small window (I think I was told 100 x 64 at COMDEX last November). DVI >>provides for full screen playback (256 x 240). > >Wrong. CD-I will also deliver the FMV into a fullscreen window. As well, >it can be encoded into many different resolutions including those >larger (either vertically or horizontally) than the display, allowing the >application to scroll around through the image. CD-I is doomed to failure because you can't digitize your own videos like you can with DVI. That's the only difference between CD-I and DVI. Otherwise, for all practical purposes, they're exactly the same, resolution-wise, full-screen-wise, frames-per-second-wise, storage-media-wise, and standards-wise. -- Nick Hoggard Phone: + 46 46 168524 Man-Machine Communication Lab Fax: + 46 46 145620 ABB Corporate Research, Dept KLL Email: nick@abblund.se Ideon Research Park, Ole Roemers vaeg 5, S-223 70 Lund, Sweden
walker@prlhp1.prl.philips.co.uk (David Walker) (02/18/91)
> >CD-I is doomed to failure because you can't digitize your own videos >like you can with DVI. That's the only difference between CD-I and DVI. >Otherwise, for all practical purposes, they're exactly the same, >resolution-wise, full-screen-wise, frames-per-second-wise, >storage-media-wise, and standards-wise. > What about cost-wise? CD-I is MUCH cheaper. It's marketed as a consumer player and will therefore be affordable in the home. david walker
brett@cayman.amd.com (Brett Stewart) (02/21/91)
I have heard it touted here that self-digitization is an advantage of DVI over CDI. Did I miss something, or is it no longer true for DVI that you either send your tapes to Princeton and pay $250 a minute for Intel to compress your FMV into DVI format, or you use the self-digitization which is low res, introduces visual artifacts, and consumes space faster, and therefore quite inferior to the quality that would be desired in a production multimedia product? (This information comes from Intel Publication 240694-001 which is a collection of article reprints) Also, I have looked carefully at the i750 specs. Again, am I missing something, or does the i750 set NOT have the ability to do MAC's, and therefore not have the ability to do at least fast JPEG compression/decompression, which has at its heart the Discrete Cosine Transform, which is best performed with a MAC capability? I believe the MPEG algorithms also use DCT-like MAC computations, with additional requirements also beyond the i750. The literature implies that the compression/decompression capabilities are one-sided, ie many seconds per frame to compress, but realtime playback. Therefore, the i750 would be inappropriate for certain kinds of groupware multimedia, that might feature realtime video captured, compressed and transmitted over a network. (Admittedly, this last is not a stated goal of DVI. However, the 'I' IS supposed to mean Interactive.) The i750 set looks like a very good product - I am not intending to disparage it but rather to understand its capabilities. If somebody from Intel is listening, can you say what sort of compression you do up there at Princeton? Is it something like a Fractal transform? Will it ever be published, so that the types of applications I just described might be made compatible with DVI by enterprising third parties? Any replies are welcome. If I receive mail that contains points of view which are not posted, I will summarize them for the benefit of the net unless requested otherwise. Brett Stewart Advanced Micro Devices, Inc. +1 512 462 5051 FAX 5900 E. Ben White Blvd MS561 +1 512 462 4336 Telephone Austin, Texas 78741 USA Brett.Stewart@AMD.com
gjh@ecs.soton.ac.uk (Gary Hill) (02/26/91)
In <1991Feb20.203009.27357@mozart.amd.com> brett@cayman.amd.com (Brett Stewart) writes: >I have heard it touted here that self-digitization is an advantage >of DVI over CDI. Did I miss something, or is it no longer true for >DVI that you either send your tapes to Princeton and pay $250 a >minute for Intel to compress your FMV into DVI format, or you use >the self-digitization which is low res, introduces visual artifacts, >and consumes space faster, and therefore quite inferior to the >quality that would be desired in a production multimedia product? >(This information comes from Intel Publication 240694-001 which is a >collection of article reprints) It is possible to perform compression on the i750 if you also have the capture board. This gives you edit level video. The info from Thorn-EMI who are marketing DVI in the UK say it will do full motion video at 10-30 fps. Thew compression is not as good as the off line compression which is called presentation level video, this is because elv does not use delta frame compression, but compresses each frame separately, as a result, the resolution is not so good for full screen video. Hope this helps, Gary Hill, gjh@uk.ac.soton.ecs
andrew@calvin.doc.ca (Andrew Patrick) (02/28/91)
In article <6938@ecs.soton.ac.uk> gjh@ecs.soton.ac.uk (Gary Hill) writes: >It is possible to perform compression on the i750 if you also have the >capture board. This gives you edit level video. The info from >Thorn-EMI who are marketing DVI in the UK say it will do full motion >video at 10-30 fps. Thew compression is not as good as the off line >compression which is called presentation level video, this is because >elv does not use delta frame compression, but compresses each frame >separately, as a result, the resolution is not so good for full screen >video. That sounds about right. We just got the new ActionMedia boards in (Capture & Playback) and are beginning to experiment with capture quality. It looks like the 30 fps capture will be usable if you use a smaller playback window (say 1/4 screen), but it is pretty ugly at full screen. However, our testing has been limited at this point. We have also found that with this new 2-board set that if your video source is anything other than a directly-connected camera, you are going to need a time-base corrector. Now a question: When you install both the capture and delivery boards, you actually plug your VGA monitor into the Capture board, and link the VGA board, Capture board, & Delivery boards together using ribbon cables. This works fine if you want the DVI playback to be on the VGA screen (you can even overlay your DOS text on the DVI images). But, how are you supposed to connect a standard NTSC screen and use it for playback of the DVI images? All you have left is the VGA connector on the Delivery Board and the VGA connector on your VGA card (which is terminated). -- Andrew Patrick, Ph.D. Department of Communications, Ottawa, CANADA andrew@calvin.doc.CA "The interface IS the program."
rkl@cbnewsh.att.com (kevin.laux) (02/28/91)
In article <1991Feb27.222824.6451@rick.doc.ca>, andrew@calvin.doc.ca (Andrew Patrick) writes: > > We just got the new ActionMedia boards in (Capture & Playback) and are > beginning to experiment with capture quality. It looks like the 30 fps > capture will be usable if you use a smaller playback window (say 1/4 > screen), but it is pretty ugly at full screen. However, our testing > has been limited at this point. > > We have also found that with this new 2-board set that if your video > source is anything other than a directly-connected camera, you are > going to need a time-base corrector. This is for the most part true, but not because of any fault on the part of the video digitizer. VTRs and VCRs have the notorious, ah, *feature*, of providing crummy sync signals, thus requiring the TBC. I have had no problem capturing CGA graphics though (from an old analog based version of my CPS workstation, preserved digitally for posterity :-)). > Now a question: > > When you install both the capture and delivery boards, you actually > plug your VGA monitor into the Capture board, and link the VGA board, > Capture board, & Delivery boards together using ribbon cables. This > works fine if you want the DVI playback to be on the VGA screen (you > can even overlay your DOS text on the DVI images). > > But, how are you supposed to connect a standard NTSC screen and use it > for playback of the DVI images? > > All you have left is the VGA connector on the Delivery Board and the > VGA connector on your VGA card (which is terminated). On page 3-2 of the ActionMedia 750 Board Installation Guide, there is a discussion about how to connect an analog 60 Hz RGB color monitor that can accept TTL sync in NTSC format. You will need to get a 9 to 15 pin adapter cable (not supplied). The pinout for the 9 pin side is given in Table 3-1. Regarding the VGA connector on the VGA card - in a development environment, you will need a monitor connected to it as well as the capture board, as most of the DVI production tools need dual screens, one for DVI and one for VGA (the monitor connected to the capture board has three basic modes: DVI only, VGA with DVI underneath, and VGA only). In a delivery environment, ie. no capture board, a single monitor is connected to the DVI delivery board. If the VGA connector on the VGA card has nothing connected to it, ie. it's unterminated, when the system boots, the VGA card will come up as monochrome only. That's why Intel provides the VGA terminator, a hardware solution. But it is *NOT* necessary, because there is a software solution. I have a little program that resets the VGA card back to full color operation. I may post it here if there is enough interest. -- ________________________________________________________________________________ R. Kevin Laux Email: rkl1@hound.att.com AT&T Bell Labs Voice: (908) 949-1160
andrew@calvin.doc.ca (Andrew Patrick) (03/01/91)
In article <1991Feb28.141127.17012@cbnewsh.att.com> rkl@cbnewsh.att.com (kevin.laux) writes: >In article <1991Feb27.222824.6451@rick.doc.ca>, andrew@calvin.doc.ca >(Andrew Patrick) writes: >> When you install both the capture and delivery boards, you actually >> plug your VGA monitor into the Capture board, and link the VGA board, >> Capture board, & Delivery boards together using ribbon cables. This >> works fine if you want the DVI playback to be on the VGA screen (you >> can even overlay your DOS text on the DVI images). >> >> But, how are you supposed to connect a standard NTSC screen and use it >> for playback of the DVI images? >> >> All you have left is the VGA connector on the Delivery Board and the >> VGA connector on your VGA card (which is terminated). > > On page 3-2 of the ActionMedia 750 Board Installation Guide, there >is a discussion about how to connect an analog 60 Hz RGB color monitor that >can accept TTL sync in NTSC format. You will need to get a 9 to 15 pin >adapter cable (not supplied). The pinout for the 9 pin side is given in >Table 3-1. OK, that part of the manual did not make much sense, but I think I understand it now (and I think I need a different NTSC monitor). Are these cables widely available, or am I going to have to go to the source suggested in the manual? (It is easier for me to buy it in Canada if I know what to ask for.) > Regarding the VGA connector on the VGA card - in a development >environment, you will need a monitor connected to it as well as the capture >board, as most of the DVI production tools need dual screens, one for DVI and >one for VGA (the monitor connected to the capture board has three basic modes: >DVI only, VGA with DVI underneath, and VGA only). In a delivery environment, >ie. no capture board, a single monitor is connected to the DVI delivery >board. OK, so what I should have is: - VGA monitor connected to VGA card - DB15 - DB9 cable connected to Capture Board and NTSC monitor correct? >If the VGA connector on the VGA card has nothing connected to it, >ie. it's unterminated, when the system boots, the VGA card will come up as >monochrome only. That's why Intel provides the VGA terminator, a hardware >solution. But it is *NOT* necessary, because there is a software solution. >I have a little program that resets the VGA card back to full color operation. >I may post it here if there is enough interest. If you don't post it, at least mail it to me. -- Andrew Patrick, Ph.D. Department of Communications, Ottawa, CANADA andrew@calvin.doc.CA "The interface IS the program."
tj@gpu.utcs.utoronto.ca (Terry Jones) (03/05/91)
>>I have a little program that resets the VGA card back to full color operation. >>I may post it here if there is enough interest. Please post, this has been a problem for me for a while! thanks.