LD#97: CAV vs CLV formats Revised: 24 Jan 94 Minor edits: 2002-07-15 > Does a CAV disk produce a better picture than a CLV disk? Short answer: For free-run playback, not really. The quality difference between CAV and CLV is small, and much less dramatic than say, S-VHS vs VHS. (Of course, for seek/still/step/slow/fast, CAV is at least twice the image quality of CLV, and on most players, offers functions that CLV discs can't duplicate.) The bottom line is that CAV and CLV editions of the same title generally have more significant differences than just the rotation mode. This is particularly true when the CLV is from a generic studio and the CAV is, say, from Voyager Company (Criterion Collection). The CAV edition is also apt to be widescreen, have supplements (esp. still-frame), dual-audio commentaries, extended/restored footage, more extensive liner notes and documentation, and perhaps be a higher quality transfer. Long answer: 0. See also LaserDisc standard articles LD#01 (overview) and LD#12 (CAV still-frame/pulldown details). 1. CAV has a slightly better video signal-to-noise than CLV. The ratio starts out the same on CAV and CLV, but on CAV the s/n constantly (although slightly) improves as play progresses to the end of the side, where 2.6 times as much linear space is available for each pit/land. This amounts to an ultimate 3 dB improvement, and is generally only noticeable on defective discs, where the noise visibly decreases during CAV play. So, on the average, CAV is less than 2 dB better. 2. Cross-talk is generally invisible on CAV. Cross-talk usually appears as herringbone patterns on the screen, and can result from an LD mastering defect (A), player misadjustment (B), or even master tape problems. In cases A & B, CAV can make the problem invisible, because On CAV, the adjacent scan lines are: - aligned in time - often identical to the current line - at least contextually related to the current line, so the visible effects of crosstalk are minimal or invisible. On CLV, the adjacent scan lines are: - misaligned in time even if identical in content at start-of-side - from distant parts of the frame image (contextually unrelated) as the side progresses - usually contain horizontal retrace interval junk (including large video pulses) during the active line period of the desired scan line, so any crosstalk is likely to be quite visible. 3. CAV requires simpler time-base correction than CLV. CAV theoretically requires less time-base correction (TBC), since it is easier to servo for constant angular speed (fixed rpm) than for continuously changing angular (constant linear) speed. Time-base errors can affect image stability (esp vertical lines/edges), and color decoding, which is determined by small phase angle differences between the initial colorburst and the main signal in each scan line. However, all LD players have some form of TBC and I haven't noticed any difference between CLV and CAV image stability, even on low-end players. There may be a slight difference in color decoding accuracy on CAV. (On the other hand, few consumer VHS or S-VHS machines have TBC, and I expect that some part of their gross vertical edge instability is due to this.) 4. Point defects on CAV tend to be stationary; they scroll on CLV. They affect the same pixel, scan line or group of lines in a sequence of frames. CLV defects tend to scroll (faster toward side end). Which of these is more obnoxious is a matter of opinion. 5. CAV is generally mechanically noiser. This is because the player runs continously at 1800 rpm, whereas CLV slows from 1800 to 600 as it plays. On the other hand, because CLV is constantly changing rpm, it can trigger harmonic vibration (thrumming) in the player as it passes through the resonant frequencies of the various mechanical subsystems (and the supporting furniture). 6. CAV is logical frame addressable, and has true FRAME still/step. CLV, on players with field store, is addressable to the nearest second, and the "still" is a field (1/2 the vertical resolution of a frame). CLV stepping on consumer players is typically track-stepped, which can be up to 3 frames near the outer edge of the disc. CAV is not usually time addressable. 7. And of course there is the "environmental noise" of having to change CAV sides twice as often as CLV. Drifting along... The CAV advantage would not have been enough to justify buying a CAV title if we had had available: * A player with CLV-freeze capability that was frame-addressable, displayed full-frame (2 field) information, had a high sampling resolution (D1, D2, or better), honored white flags (if present) & stop codes, and was also selectable for field dominance (in case the flags were incorrect or missing). Even the industrial LD-V8000 couldn't do all of this on all CLV discs. If such a consumer machine had existed, it would have been expensive. Even at the twilight of LD, you usually had to buy 2-sided play just to get medium-res freeze-field with sloppy time addressing. * Full-featured CLV titles. Even if the hypothetical CLV-frame consumer machine existed, big CAV houses (like Voyager) could not (and would not) have switched to single all-CLV editions. That would have lockd-out frame/stop functions for cheaper machines as well as the entire installed base of non-field-store ("CLV effects") machines. Since they would have continued to make both CAV and CLV editions, they would likely have continued the practice of minimizing the features and supplements on the CLV editions, both to keep the pre-mastering and mastering charges down on the CLV, and promote the CAV. We never got advanced CLV players that could eclipse the need for CAV. We got DVD instead (always CLV, but with full frame store usually). Make your CAV vs CLV buying decisions on the merits of the transfer and pressing of each title. Regards, PO Box 248 Bob Niland Enterprise mailto:name@isp-name.domain Kansas which, due to spam, is: 67441-0248 USA rjn AT access DASH one DOT com Unless otherwise specifically stated expressing personal opinions and NOT speaking for any employer, client or Internet Service Provider.