It is currently 19 Mar 2024, 04:03

All times are UTC [ DST ]


Search found 20 matches

Author Message

 Jump to forum   Jump to topic

Posted: 12 Oct 2011, 20:02 

So, for some systems such as PAL+, technical information is readily available. Others, such as red-blue anaglyph 3D, are just obvious. I thought a thread with information on the less well known systems might be useful. I have some technical papers on MUSE Hi-Vision LD, & the two different types of HDVS discs, so I'll start with that.

 Jump to forum   Jump to topic

 Post subject: MUSE Hi-Vision LD, part 1
Posted: 12 Oct 2011, 20:42 

There were actually multiple iterations of MUSE optical videodiscs before the final consumer product was developed. The goal was to have 30 minutes per side playback time in CAV mode, or 60 minutes in CLV, the same as standard NTSC LD, but this was difficult to achieve. The MUSE signal has a bandwidth of 8.1 MHz, as opposed to 4~5 MHz for NTSC or PAL, meaning that the carrier frequency has to be higher, and the recorded bandpass wider.

The first prototype MUSE LD system was developed in 1984. (That's right, kids, before the first Digital Audio LDs were released!) At that time, the MUSE system was still under development, but it was recognized that packaged media would be important to the introduction of HDTV. The early MUSE system used a carrier-multiplex form of digital audio which was incompatible with LD recording, so a different type of multiplex was used on the disc, and processing was necessary in the player to change it back into the form accepted by the MUSE decoder. As a result, audio was limited to two channels of 32 kHz sampled 10-bit companded PCM.

The carrier frequency used was 14 MHz, with deviation of ±2 MHz, ie, black level at 12 MHz, white level at 16 MHz. The same emphasis used for broadcasting was applied. A second signal was also recorded, a "pilot tone" at a frequency of 2.28 MHz (67.5 times the line frequency) & an amplitude of -20 dB relative to the video carrier. This provided a reference for synchronization & to drive the moving-mirror tangential servo, so that jitter could be reduced to below 10 ns without the use of a digital time-base corrector.

The track pitch of the disc was 1.65 micron, and the minimum linear playing speed was 18 meters per second. The inner recorded radius in CAV mode was 95 millimeters, giving a play time of 17 minutes at the conventional 1800 revolutions per minute. In CLV mode, the inner recorded radius was 55 mm, the speed varied from 3100 to 1200 RPM, & the playing time was 30 minutes.

http://ura.caldc.com/stannum/disc/sanyo_1985_figure_9.jpg

Information & illustration from "Optical Videodisc for High-Definition Television by the MUSE", Tateo Toyama et al, SMPTE Journal , January 1986, pp 25 et seq. (Paper originally presented at 19th Annual SMPTE Television Conference, 16 February 1985.)

 Jump to forum   Jump to topic

Posted: 12 Oct 2011, 21:29 

Supposedly there was an effort to develop a MUSE version of VHD, but this seems to have gone nowhere.

Sanyo continued working on the MUSE LD problem, & the MUSE encoding format itself continued to evolve. A second Sanyo system was similar to the first, with the same pilot tone frequency (aplitude -28 dB) & play time, but used a deviation of only ±1.5 MHz, so that the black & white level frequencies were 12.5 & 15.5 MHz. This system again multiplexed 32 kHz, 14-10 bit companded 2-channel audio data into the MUSE signal, and also added a CD-type audio track, for a total of 4 channels. ( "Hi-Vision Optical Video Disc", Toshiaki Hioki et al, IEEE Transactions on Consumer Electronics , vol 34 no 1, February 1988, pp 72 et seq. )

By 1991, Sanyo had developed a disc with 60 minutes of play time per side. This used a pickup with a lens numerical aperture of 0.6 and a newly-developed laser diode with a wavelength of 670 nanometers (as opposed to 0.5 and 780 nm for NTSC players), providing a smaller scanning-spot size. This allowed reducing the linear velocity to 14 m/s, and the track pitch to 1.12 microns. The carrier frequency and deviation were changed to 13±1.5 MHz. The MUSE signal, which by that time had been finalized (daily broadcasts began in that same year), was used without alteration, and the CD-type audio signal was multiplexed at -27 dB. Thus, a total of 6 audio channels was available, four from the MUSE A-mode audio signal, & 2 from the Red Book track. ( "High Density Muse Videodisc", Hitoshi Terasaki et al, IEEE Transactions on Consumer Electronics , vol 37 no 3, August 1991. )

 Jump to forum   Jump to topic

Posted: 12 Oct 2011, 21:51 

At this point, other consumer electronics manufacturers became involved. A consortium of five firms, including Sanyo and Pioneer, was formed in September 1991 to develop & market the Hi-Vision LD format, & Pioneer introduced the HLD-V500 commercial player based on this format in May 1992.

The carrier frequency of the final format was 12.5 MHz, with deviation ±1.9 MHz, or a frequency range of 10.6—14.4 MHz. The pilot signal of 2.28 MHz was used, as before, and the Red Book CD-audio track was allowed as an option but not required. The inner playback radius was 55 mm in CLV, with a starting rotational speed of 2700 RPM, and 76 mm in (1800 RPM) CAV, with a track pitch of 1.1 micron, providing 60 or 30 minutes per side respectively (20 or 10 minutes on a 20 cm disc, although I have never heard of one). In order to deal with dropouts, the error correction code in the MUSE audio signal was changed from BCH(82,74) to BCH(82,67). (The extra bits were taken from those set aside for "independent data" in the MUSE specification.) These codes are closely related, & it is believed that all consumer MUSE decoders can accept both types of audio signal. A 76-bit control and address signal was inserted, using biphase modulation at 2.7 megabaud, in line 564 of each frame, which is left blank and reserved in the MUSE specification for exactly such uses.

http://ura.caldc.com/stannum/disc/pioneer_1993_figure_1.jpg

Information & illustration from "MUSE Videodisc System", Takashi Sakakibara, 1993 NAB HDTV World Conference Proceedings , pp 147 et seq.

 Jump to forum   Jump to topic

 Post subject: Re: Squeeze LD
Posted: 14 Nov 2011, 22:35 

Let's be clear on the basic math here.

If you letterbox a 16:9 image into a 4:3 video frame, which is the conventional way to do it, you are only using 360 of the 480 visible scan lines of NTSC. If you zoom in on that letterboxed image, to fill the screen of a display with a 16:9 native ratio, you still only have 360 "real" lines of resolution, even if the display has 480, 720, 1080, or some other number of scan lines. If, on the other hand, you "squeeze" the 16:9 image, that is, use all 480 lines, then you obviously have more vertical resolution available.

As far as horizontal resolution is concerned, the nominal bandwidth of broadcast NTSC is 4.25 MHz, which corresponds to about 425 resolvable elements across the width of the screen. In 4:3 aspect ratio, this means about 320 elements per picture height ; in 16:9 aspect ratio, it gives only about 240 resolvable elements per picture height. (Reducing the horizontal resolution compared to the vertical resolution is justified, in line-scanned & especially interlaced displays, by something called the "Kell Factor" which relates to the fact that the image is continuous in one direction & discontinuous in the other.) Now, some LDs have a wider video bandwidth, but typically most of the other equipment in the production chain was designed to work within the 4.25 MHz envelope, so the degree of improvement there is limited, & the 2/R analog audio carrier had a nasty tendency to interfere with the video. I believe the HLD-X0 demo disc is an example of a wide-band LD with only digital audio.

The fact to observe, however, is that whether you are zooming a letterboxed image or displaying a squeeze image at its proper ratio, either way you have the same ~240 elements per picture height to play with. Hence the squeeze image has, as we would have expected, a higher total resolution. This is the exact same math which applies to DVDs, although the DVD has an allowable video bandwidth of about 5 MHz, making for about 500 elements across the width of the screen, or 280 per screen height in widescreen mode.

Now, given a fixed resolution, does a larger image or a smaller one look sharper? The smaller one, of course. But, on the other hand, due to the phenomenon of finite visual acuity, that perceived sharpness is bought at the expense of being able to resolve visible detail.

 Jump to forum   Jump to topic

 Post subject: Decoder accessories
Posted: 25 Nov 2011, 10:03 

So, I'm preparing to start serious practical research into the MUSE format. It won't be for another couple of months, but I want to prepare the way. I've started thinking about possible pieces of hardware I might build in the course of verifying my understanding of the format, & I wanted to get some feedback from people as to whether they sound useful. If I could sell a couple of these gizmos, to recoup some of the costs of my experiments, of course that would be pleasing as well.

I'm not ready to start dealing with video yet, & when I do it'll mostly be on the encoding side, but the way the existing decoders handle the 4-channel A-Mode audio is kind of inconvenient. Unless you have something which can handle two simultaneous SPDIF PCM inputs, which is virtually unheard-of, you have to deal with an analog multichannel connexion, which can get awkward. And not all decoders (HV-VMD1, I'm looking at you) even have the four-channel analog outs! On the other hand, every decoder has the "MUSE bitstream" output, which dumps the 1350 kbps DANCE datastream. This was meant for operation with outboard equipment for some kind of data services which I don't think were ever implemented, but it could just as easily be used with an outboard audio decoder.

So I thought of two gadgets. One, which seems useful for diagnostic purposes but probably wouldn't appeal to anyone but me, is a sort of headphone amp, with LEDs to show which channels (1-4) are present, & buttons to assign one to each side of a stereo headphone jack. The other is probably more useful, putting the audio onto some kind of connector which can carry multichannel PCM. I guess this would be HDMI for most users, although I think there are some components which can accept it over some kind of USB connexion.

Anyway, tell me what you think.

 Jump to forum   Jump to topic

Posted: 10 Jun 2013, 04:19 

This is not a digital mastering problem of any kind, or even a video problem per se. What you are seeing is a sort of image lag . Basically, when the film-to-video transfers were done, the image sensor used (probably a plumbicon camera element) had a certain tendency to retain an image. So, in addition to the current film frame, the image captured on videotape incorporated the three or four previous film frames at a substantially reduced intensity. You can really see it, as I recall, on the first of the L.A. Hero Giant Robo discs, at places where a scene cuts onto solid black. The phenomenon, in fact, is exactly the same as the afterimage which lingers in your eye after looking at something.

Using a better transfer system, such as a flying-spot scanner instead of a film chain (projector-camera combination), or even a different choice of film chain pickup, would have prevented this. But another part of the problem is simply that, with the particular system I suspect was used, the overall amount of incident light has an effect on how rapidly the afterimage is discharged. It would require some skill, in a program with a lot of dark backgrounds, to get the black level right while simultaneously minimizing the lag. If setup for the transfer was done carelessly, or in haste, or using settings worked out for some other type of program material, I would expect the results you describe.

Then I have another problem that is, IMO, even more annoying, and its all over every disc of Dirty Pair: Perfect TV (1985) [VPLY-70266] .

While the previous sort of artifacting was probably the result of too much light being shined into the telecine machine and %100 analog in nature, the problem with the Dirty Pair discs is clearly digital in nature. Basically, the discs look like they were made from VCDs or something. Again this is visible in high contrast areas. A great way to set it off is to watch a white spaceship traveling through black space. The ship will leave a trail. Its really looks like MPEG artifacting or maybe even DIVX AVI. If I were mastering this myself I would say "not enough key frames!". When everything on the screen is bright its fine, but its really friggn' annoying when you see situations like the space ship I mentioned, which sadly happens constantly since this is a scifi show!

This box consists of the entire TV series and a bunch of OVAs. The OVAs were also sold separately years before this set came out but I've never viewed them myself so I don't know if the problem was introduced in this set or if it was there all along. The OVAs originally came out after the TV series aired and were from a different era of production so it seems weird that the same problem would be there all along in the original tapes. My assumption is that this sort of artifacting was introduced by some terrible early digital mastering system that thankfully didn't get used very often. I REALLY HATE this problem and it really brings down an otherwise totally amazing box set. For the record this show was probably shot on 16mm and then transferred to...I don't know, betacam or something for TV broadcast and/or transfer to home video. My point being it wasn't shot to digital tape, or broadcast digitally, or available in any digital format until DVD came about. I think this LD set is from 1994.

 Jump to forum   Jump to topic

Posted: 30 Jul 2015, 09:34 

Just recently I have acquired a "Blackmagic Intensity Pro" video capture card, capable of accepting an analog-component high-definition input. I'm still struggling to make it work, but I have had at least a little success so far, with grabbing a still frame. So this is the test pattern from "The Test Disk", captured via Sony HIL-C2EX and MSC-3000.
Click for full-size version (3.3 MB .PNG)
http://ura.caldc.com/stannum/decoders/msc3000_still_small.png

 Jump to forum   Jump to topic

Posted: 31 Mar 2016, 19:57 

For a real trip, you could try "Gestalt". It was computer-animated (in a manner which imitates cel work) at 30 fps. Not quite what you're looking for, I'm sure!

 Jump to forum   Jump to topic

Posted: 28 Oct 2017, 04:03 

What I would like to know, & may end up paying money (alas!) to find out, is whether a CD-Graphics decoder with an external video input would also work. The first-generation CD-G decoders seem to have, not a SPDIF digital audio input, but an input for the EFM signal off the disc, which some of the Pioneer players have an output for.

 Jump to forum   Jump to topic

Posted: 30 Oct 2017, 16:40 

Sorry to me a service manual is usually something that goes way over my head when I look at it. Just endless board diagrams that make no sense to me.
If you want to work on LD players, the service manual is not just your best friend, it's an indispensable guide. These machines are complex enough that even a very experienced electronics service tech can do little without extensive documentation.
It's not shameful if this is not for you.

 Jump to forum   Jump to topic

Posted: 04 Nov 2017, 17:21 

I can't figure out what you're trying to say here.

Are these "discs you can never find a copy of at less than a ridiculous price"? Or is it "discs you will never buy, no matter how cheap, or how many copies you encounter"?
Because, seriously, "The Venus Wars" is neither rare nor expensive.

 Jump to forum   Jump to topic

Posted: 07 Dec 2017, 00:22 

Shortly put, the whole thing about 3:2 pulldown that makes it work is that, while video runs at 30 frames per second, it does so in two steps. First half the lines are scanned, in 1/60 second, then the remaining lines in the remaining 1/60 second. The two sets of lines are "interlaced" as you do with your fingers, first a line in the first field, then a line in the second field, then back to the first field, & so on.

So, as a result, the easiest way to deal with film running at 24 frames per second and video at 30 is to recognize that both are multiples of 120. Each video frame occupies 4/120 second, & each video field occupies 2/120, while each film frame occupies 5/120. So, your first frame is held in the film gate for the period of two fields or one frame, that is, 4/120, and then "pulled down" to make way for the next one, which is held for three fields or 6/120. Since 4+6=10, the two successive film frames occupy 10/120 second, & two-and-a-half successive video frames occupy 10/120, so everything works out fine.

The main thing to realize is that each & every film frame is represented either by two or by three successive video fields. In other words, either a complete picture in two halves, or a complete picture plus a half of a picture repeating the first half of the previous picture. It is, therefore, possible to recover the original film frames. The only exception occurs when video editing has been done after the film-to-video transfer step, in which case you may be left with occasional partial frames. This is, however, rare. Much more common are "cadence shifts", in which the regular 3:2:3:2:3:2 sequence of fields stutters, giving 3:2:2:3 or 2:3:3:2 sequences. These cause no problems to the viewer, but schemes which rely on detecting the sequence of fields often react badly.

 Jump to forum   Jump to topic

Posted: 13 Dec 2017, 08:41 

Looking at a schematic diagram (which, unhelpfully, is mostly in Chinese), it appears that you want to tap at the junction of Q203 & R208. From that point, the signal flows to both input pins of IC201, the audio FM demodulator. Much upstream of that, you'd be getting some version of the full off-disc RF signal, which is likely not to be all that helpful.

I guess the Panasonic design works rather differently from the Pioneer one, in that the two audio FM signals are separated at the demodulation stage, by different time constants in the two parallel detectors. At least, without a proper datasheet for the PA0034A, I think that's what's going on. The only difference between the networks on the two sides of the chip is C216, 270 pF capacitor attached to pin 7, vs C215, 330 pF on pin 36.

 Jump to forum   Jump to topic

Posted: 14 Dec 2017, 07:56 

I suggest taking the output through the otherwise useless mono audio jack (the one tagged for an RF modulator).

 Jump to forum   Jump to topic

Posted: 06 Jan 2018, 02:34 

You see in this image a block marked FL301. This is the band-pass filter that splits the audio FM into left & right channels for demodulation. Just before this, at the output of Q303, looks to be where you should be picking off the AFM signal for your AC-3 output. I suspect the big circled dot on the trace from Q303 to FL301 marks a test point where you can very easily attach.

 Jump to forum   Jump to topic

Posted: 08 Jan 2018, 10:02 

Title is "Chinese Light" or "The Light of China", & the subtitle is something like "New First-generation Original Hit/Gold Songs from China". I'm guessing it's probably a karaoke disc, but the song titles are very hard to read.

 Jump to forum   Jump to topic

Posted: 08 Jan 2018, 18:52 

The suggestion I had was to use a very thin, even layer of Plasti-Dip or something like it, a liquid rubber compound. If necessary it could be "tooled" for extra grip.
But, happily, I've not needed to test that so far.

 Jump to forum   Jump to topic

Posted: 12 Feb 2018, 08:25 

Without telling what player you're using, & what other equipment it's passing through on the way to the display, & what the display is, I don't think anyone can help.

 Jump to forum   Jump to topic

Posted: 22 Jun 2019, 20:48 

Hi folks, I was recently given a CLD-1080, manufacture date 1990. I opened it up & it was clean, aside from a faint fishy odour. That can sometimes mean bad caps, but I didn't see anything obvious. Anyway, after exercising the pickup slide by hand a little bit, it works fine & passes all functional tests. I'd say it's at 95% operation.

The 5% is a crackle in the digital audio output. This doesn't happen all the time. It's disc-dependent, & it only occurs during some sections of the disc. So, for instance, it occurs during the first & last 20 minutes of side A of my copy of Heavy Metal (that side is pretty speckly). It only occurs during the first 5 minutes of side B. It's all through my copy of LaserVision Demonstration (CAV), doesn't happen at all during Good-Bye Mama (12cm VSD), or on CDs so far as I have yet found.

It's not a huge problem, because I rarely need to play the digital soundtrack in opposition to the analog one. But if there's an easy fix, such as turning up the RF level pot, I'd like to apply it. This particular player is most likely going to be my choice for the A-Kon Classic Anime Festival next weekend.
Page 1 of 1 [ Search found 20 matches ]


All times are UTC [ DST ]


Jump to: