How did RCA SelectaVision work? (1 Viewer)

[Kevin Farley]

I know it used a stylus and it was a "Capacitance" disc, but after that I'm baffled. What is a Capacitance disc? (I've mostly finished my BSEE so I'm not afraid of a deep explanation...)
And I guess I'll post this separately if need be, but how does laserdisc work? the disc is read by a laser, but if it's not digital, (I've heard it's analog), how is it read? Is it pits that deflect the laser to different photoelectronics? or does it somehow reflect a continuous stream to reproduce a signal?

 

[Robert Silge]

Can't help you on SelectaVision, but (not being an EE) the easiest explanation I have for LD is that it's like a phonograph turned 90 deg. Phonographs have various depths to the groove, thus producing a range of frequency responses. The video on LD is similar, but the length of the pit determines the frrequency. The audio is digital just like a CD. Someone can correct me if that's a poor explanation.
------------------
-Rob
My meager collection:
www.dvdtracker.com/~lazarus.asp

 

[Leo Kerr]

Okay, I can help a little bit on both LD and Selectavision.
First, I have a literary question. The standard convention is that a magnetic platter like in a hard drive is called a "Disk", while an optical platter, like a CD or LD is called a "Disc." What is an analog platter like a record called? A Disk or a Disc?
Selectavision (or whatever.)
There is, as I understand it, two concentric "grooves" on the platter. One is for the sapphire stylus to ride in to guide the read-head, and contains no actual information. The other "groove" isn't really a groove, but is more a "space" between the guide-grooves. On this "groovy-space," there are shaped "pits" in the vinal that are read by the capacitance-sensitive head. I believe the coding is not unlike a laserdisc's coding, although it was always all analog (never supported digital audio.)
Laserdisc
I'm trying to remember this from Bob Nyland's excellent series of articles on LD technology. Does anyone still have them archived? Anyway...
Laserdisc is analog. All of the information is "packed" into a single analog waveform; some things are shifted up or down frequency, are boosted or compressed, placed on different carrier frequencies, and all sorts of other things. This waveform is then "clipped," resulting in what might otherwise look like a digital bitstream, except it is meant to be read in an analog fashion.
That is, a transition from a high to a low state represents a zero-crossing of the original master-waveform, while the center of the plateau or valley represents... well, it does represent the peak of the waveform, but I can't recall how one found the actual final height. Unless it was also "mushed" in such a way to reduce all of the variations into a uniform wave. I'll really have to look in my archives and see if I still have those articles...
Hope this helps.
Leo Kerr
[email protected]

 

[Leo Kerr]

Okay, I was wrong. But here are the words of the master. Enter...(drumroll,) the Laserdisc!
LD#96: How is LaserDisc "analog"? Edition: 21 Jan 92
re: > In what way is the data on an LD not digital. I agree that what eventually comes out the RCA cable is analog, but what is read off the disc is still numbers isn't it?
from a technical summary of LD to be written someday...
Despite what the promoters of LD might allow you to believe, LD is not digital video. Is not always digital audio, either. This is not surprising considering that "LaserVision" is basically a 1972 technology.
There is a single track consisting of a continuous spiral of pits and "lands" (non-pits). Encoded on that track are the composite video (luminance, chrominance, sync, info), two analog audio channels and a RedBook-style stereo digital audio data stream complete with subscodes. Side, track, time, frame and other control signals are encoded, during pre-mastering, in both the RedBook subcodes (if present) and in non-displayed scan lines in the vertical interval of the composite video signal.
* The pre-mastered baseband composite video signal (about 4.2 MHz wide for NTSC) is used to modulate an FM carrier signal with a nominal frequency of 8.5 MHz. The amplitude of the NTSC signal deviates the FM carrier over the range 7.6 MHz (video sync tip) to 9.3 MHz (video white level).
* The two analog audio channels each modulate a pre-emphasized FM audio subcarrier (2.3Mhz 1/left, 2.8MHz 2/right).
* The RedBook-style digital audio bits are pre-emphasized and used directly. They represent a signal envelope with a bandwidth under 1.4 MHz.
* The three FM carriers (video, analog 1/L, 2/R) and the digital audio are all summed, then clipped. The audio stuff is all attenuated 26dB before the summing.
The resulting clipped signal looks like square waves, and the low level, low frequency components (analog and digital audio) cause a variation in +/- duty cycle. This signal is then more-or-less directly recorded as pits and lands.
In over-simple terms, the video signal is represented by the center-to- center pit spacing and the audio is represented in the difference between pit/land length.

 

[Leo Kerr]

Okay, and now for to answer the original intent of the thread.
Bob Nyland's artical on CED:
LD#16: Capacitance Electronic Disc system Revised: 05 Mar 92
In 1990, I stumbled across some used CED discs, bought one and got curious about this now-antique technology.
CED was marketed between 1978 and 1984, and was an incompatible videodisc format that competed with LaserDisc (LD, the Philips/MCA system then called LaserVision or LV). A few test PAL CEDs were also made (bizarrely, AFTER CED was cancelled in the US), but the only major CED market was North America. Although Japan uses NTSC, they adopted a grooveless capacitance system called VHD (reportedly declining but still in use). CED was tested there, but never established a foothold.
CED titles sold for $5 to $10 less than LDs, and both disc formats have always been priced substantially less than tape. There were 4 or 5 player brands, including Radio Shack. (Curiously, the Radio Shack laserdisc player introduced in early 1991 has a catalog number which is the old CED player number plus one.)
The Technology...
Each CED platter is delivered in a 14 x 12.25 x .25 inch opaque plastic caddy that looks somewhat like an oversized version of today's Philips/LMSI CD-ROM caddy. If a video title required two or more platters, they were provided as two or more caddies (with some risk of being separated - when I bought my sample, I wanted to get a title that was a duplicate, and I initially grabbed one that turned out to be "disc 1 of 2" - I returned it.)
The entire caddy was (is) inserted in the player, and the outer sleeve withdrawn. The exposed platter was not supposed to be handled. You can remove the carrier and the disc platter by releasing two recessed latches on the top edge; however, this will void your media warranty There is a small soft textile strip on the lips of the sleeve for collecting debris (or spreading it, depending on your point of view).
The disc itself is just under 12 inches in diameter. Apart from the large 1.3-inch center hole, it looks somewhat like a typical LP made of black vinyl. My references tell me that this is actually a conductive carbon- filled PVC. Early discs were reportedly unfilled vinyl substrate, with an inconel-on-copper data layer and an oiled styrene surface. Problems with delamination led to the use of solid conductive vinyl.
Open question: Did any metallized discs reach market?
Looking at the vinyl disc, colorful light diffractions suggest that the grooves are significantly smaller than on an LP. Also, the vertical interval of the video signal produces visible spokes in the diffractions, which confirms two things:
- There are 4 frames (8 fields) per rotation. This resulted in pronounced motion during "still frames". I hesitate to call it "field motion", as there would be a burst of 4 complete video frames (8 fields), which could be 3 or 3-1/2 film frames, depending on field dominance. LDs are either 1 video frame per rotation (CAV) or between 1 and 3 (CLV).
Deliberately mastered CED still-frame material repeated for 8 fields.
- CED discs are CAV. I have seen no evidence of CLV in the CED format. Since CED had an inherent capacity of 1 hour per side (compared to 30 minutes for LD/CAV), it's not clear that CLV would have been advantageous. LD requires CLV to reach 1 hour per side. RCA is reported to have demonstrated 3 and 4 hours per side on experimental CEDs.
There are numerous blemishes on the surface of my sample, some of which look like fractures and others like spalling or flaking. There are none of the fingerprints and scuffs that are common on rental/used LDs, so I presume that these spots are either manufacturing defects or the result of normal use.
The active data area on the CED spans the radii 2.9 inches to 5.8 (compared to 2.2 to 5.7 for LD). CEDs play from the outside in, like LPs. LDs play from the inside out, like CDs. Also like LPs, the grooves for each CED side are on the same side as the jacket label for that side. On LD, the data surface is on the opposing side from the side label. Discs were flipped manually. Some players had auto-insert and eject.
CEDs spin at 450 rpm (vs 1800 rpm for LDs, with CLV slowing to 600 rpm at the outer edge.) This provides a linear velocity of between 136 and 273 inches/second for CED (vs 390 ips for LD/CLV and up to 1000 ips or so near the outer edge of an LD/CAV). This means that CED has only 1/3 as much linear space per frame as LD. I would expect this to result in lower bandwidth and reduced signal/noise.
The CED signal is a single spiral of pits (as is LD). On CED the pickup head (which is electrical, not mechanical) is guided by a groove. The arm is linear tracking, as are some LP turntables. Arm deflection triggers a drive servo to move the arm. Unlike an LP, the groove is smooth and shallow. This is also where CED and VHD depart. VHD has a an ungrooved flat surface. The VHD pickup is servo controlled and is guided by pilot signals between the data pit tracks. On LD, the surface is smooth, and the pit track is followed optically by servo.
To put it more crudely, unlike an LP, the CED groove does not "wiggle". The bottom of the groove follows the pits, but the pickup stylus contact area is longer than several pits, and rides on the "lands" between the pits. It does not track the pits or otherwise vibrate. This should minimize wear, but I would still expect the stylus to be polished down over time. Early stylus assemblies were user- replaceable, and reportedly had a 400 hour life.
I would also expect groove ridge damage from fast-fwd/rev, still-frame and other operations that require skipping tracks, since the pickup is not raised during these operations. I would further expect self-skipping problems in the face of debris in the groove and general surface damage. One owner reports that his manual warned against over-use of still-frame, lest it result in excess skipping. Nonetheless, another reports having owned a couple of "viewer controlled" (interactive) entertainment titles that had self-still action points.
The stylus senses the pits electronically. The stylus tip is shaped something like a flat-bottom rowboat. It is solid diamond or saphire, with a thin sheet of metal on the "stern". This electrode is stimulated with a 915 MHz signal. The signal strength is modulated by the change in capacitance caused by the pits flying by (remember, the CED disc is conductive). The various video and audio signals are extracted by demodulating and detecting the envelope of the 915 MHz carrier.
Warps and rotational eccentricities can result in data rate errors large enough to cause incorrect color decoding. Time-base-correction (TBC) on CED is accomplished by changing the length of the pickup arm. On LD, time-base is handled by swinging mirrors, optical delay lines or more recently, with digital TBCs.
The pits are recorded in pulse-FM, as on LD. The CED video signal is recorded "color-under", just like standard VHS (and unlike LD, which is composite). CED luminance is modulated at 4.3-6.3 MHz and chrominance at 3.6-2.6 MHz (a heterodyned signal with 670KHz bandwidth). LD composite is 7.5-9.2 Mhz. RF (ch 3/4) output was standard. No one has reported seeing a even composite video output on CED, much less Y/C out (S-video).
CED is reported to get only 250 lines of horizontal resolution, compared to the 425 claimed for LD. Although the VHS-like color-under scheme has well less than full NTSC chroma resolution, CED transfers were generally of higher quality than VHS from that period. RCA was one of the earliest users of flying spot cinetele scanners, and used films negatives where possible. CED owners confirm that quality (on "good" discs) was somewhere between early-80s VHS and Beta.
Although intended for pressing on ordinary LP-style equipment, CEDs required clean-room conditions. Reportedly, only RCA and CBS ever made them.
CED has stereo analog sound with CX noise reduction. CED did not last long enough to benefit from digital sound (added to LD around 1985). Audio on CED, as on LD analog, is two FM subcarriers. CED deviates them by 75 KHz (vs 100 for LD). CED defines them FM-multiplex broadcast-style, as stereo L+R and L-R. On LD they are discrete L and R. Not all CED players had stereo outputs (all LD players do).
Re-creating separate L and R channels from the CED sum and difference channels is not difficult, but does increase the risk of cross-talk. Cross-talk is not a serious problem for a stereo program, but might make dual-analog impractical. I have no information on whether any dual-program CEDs were attempted.
Different players offered operations similar to LD, including: * medium and high speed search/scan, fore & aft;
* vertical interval time codes with one second resolution; * time seek;
* "band" seek (similar to chapter on LD);
* "page" (4-frame) seek +/- 4 fields (similar to frame on LD); * auto-still-frame (stop codes).
There was no normal-speed reverse play. And obviously, CED expired before economical digital field store was available. However, since all the titles were CAV, this is an arguable disadvantage.
So why did CED fail?
* It didn't have enough market share to keep the attention of the bottom-line obsessed revolving-door management at RCA.
* Couldn't record, but cost as much as a VCR, and VCR prices were dropping rapidly in the '80s.
* Failed to enter Japanese market at all (LD has significant market share there, which helped keep LD alive here).
* Sensitivity to contamination. Skipping was perceived to be a serious problem. Also a perception of wear, even though RCA claimed 2000 plays per platter. Several respondents reported a claim of only 100 plays.
* Did not have a dedicated vendor like Pioneer to keep format alive. RCA was just beginning its final descent into self destruction and eventual buyout. There was no stable, committed management present to run a videodisc crusade.
* No niche: CED was not sufficiently superior to tape (esp. Beta, which was its early competitor), and was clearly inferior to LD.
* CED did not appeal to "early adopter" videophiles, whose snap opinion was that laser was cool, but stylus-in-groove was technologically primitive. They used the perjorative name "needle vision" when discussing CED.
* The public was confused by two "disc" formats (when I demonstrate LD in my home, many people ask "Can you still get discs?") They only remember that "video disc died in 1983".
* Open systems: RCA tried to get a lock on the CED market. Early players were expensive and only available at RCA dealers.
Of course, had CED not been introduced, LD might have been a bigger success (or a complete disaster, given the media quality problems up through 1987).
For more information, see "RCA and the Videodisc: The Business Of Research", by Margaret Graham (which I have not read).

 

[Leo Kerr]

Last post here. I promise.
FYI, the entire collection of Bob Nyland's articles are still archived on the internet, available via FTP (which you can do through your web browser.)
nic.funet.fi (128.214.6.100, Finland, dir: /pub/culture/tv+film/laserdisc)
Or, for those of you unfamiliar with the inner workings of FTP, enter this address into your web browser:
Link Removed
Leo Kerr
[email protected]

 

[Kevin Farley]

Wow, wow, wow, wow, wow! Thank you so much!!! I was hoping to get at most a paragraph. This was so helpful and informative. It's fascinating to learn how they managed to use a groove to record video. I've always wondered about the use of the same concepts/tools used to "sweeten" audio, how reverbs, delays, dbx type stuff(dbx, bbe, audax, etc.) would affect video signals; it's interesting to think about video signal manipulation; changing pictures. Thank you again, and I'm going to check out that FTP site you recommended. I've puzzled about how a laserdisc could store an analog waveform (versus on/off bits) ever since I learned that they were analog.
Thank you again,
Kevin Farley