1080p24 ("24p") on Blu-ray

One of the best things about Blu-ray is that, for the first time on home video, a motion picture is rendered on disc just the way it was recorded on film. Both actual movie film and Blu-ray Discs that contain movies or TV shows shot of film record images at 24 frames per second. On a Blu-ray Disc, accordingly, film is transferred to video at a one-to-one ratio of film frames to video frames. When a Blu-ray Disc containing a movie or a TV show shot on film is output to an HDTV at, specifically, 1080p24 (a.k.a "24p") resolution, playback is notably smooth and free of distracting video artifacts.

On a DVD, film is transferred to video in a way that uses a different ratio of film frames to video frames. There are 30 video frames per second on a DVD, not the 24-fps rate of film. This frame-rate mismatch compromises a movie's ability to be displayed in a truly film-like fashion on a progressive-scan HDTV. DVD players do not support the 1080p24 output resolution because DVDs themselves do not.

The frame-rate mismatch makes no major difference when a DVD is played into a now-obsolete TV that uses a picture tube. But today's HDTVs don't use picture tubes, so the frame-rate mismatch causes the displayed picture to be slightly impaired. Blu-ray players and discs avoid the mismatch problem entirely.


Practical considerations:

In order to take advantage of the 24-fps frame rate of movies recorded on Blu-ray Discs, you need:

• A Blu-ray player that supports 1080p24 output on HDMI
• An HDTV that supports 1080p24 input on HDMI
• An HDMI cable to connect the player to the HDTV
• A Blu-ray Disc containing a movie or TV show shot on film
• The right setup option(s) in the player and TV
  

(Ideally, the TV will automatically accept 1080p24 on HDMI and use it without anything special being done to its setup options. The same is typically true of a Blu-ray player when it is using its default "Auto" — for "automatic" — output resolution selection. Only in special cases would the player's automatic output setting need to be manually overridden.)


Background information:

The 30-fps method used to transfer films to DVD is used due to the fact that, in the United States, TV was originally standardized to provide 30 video frames each second, even though film uses 24 frames per second.

To avoid image flicker, in the NTSC television system that was used for decades until the recent digital television revolution, and is also used on DVDs sold in the U.S., each video frame is divided into two separate "fields" transmitted 1/60 second apart. Though they represent slightly different moments in time, the two fields of each video frame can be thought of as "interlaced" together, much as the fingers of two hands folded together are interlaced together into a single unit.

Today, the NTSC standard is being replaced in the "digital conversion." This conversion has actually been going on for several years, though today we are hearing a lot about the problems it will cause for consumers who are not ready for it when, on June 12, 2009, the plug is finally pulled on analog NTSC TV broadcasts. But all DVDs sold in the U.S. will continue to be recorded in a digital version of NTSC even after the digital conversion is done.

NTSC (which is analog video) and DVDs (on which NTSC-like video is recorded digitally) use two interlaced fields per video frame, whether the interlaced analog or digital video comes right from an NTSC video camera or is transferred to video from motion picture film. Each 1/30-sec. frame of interlaced video is made up of two fields, with each field taking 1/60 second.

Frames of NTSC or DVD video (and the two fields they are each subdivided into) are composed of numerous horizontal scan lines. These lines run across the screen from left to right, but since they are stacked on top of each other in the vertical direction to make a two-dimensional picture, they serve as a measure of the "vertical resolution" of the TV picture. In each NTSC frame there are 525 scan lines, though only 480 of them are visible on the TV screen (the others are in the so-called "vertical blanking interval" and don't show up on the screen). The vertical resolution of NTSC video (and of DVDs in the U.S.) is 480 lines.

The first field of each video frame contains that frame's even-numbered scan lines, starting at line 0 at the top of the screen: 0, 2, 4, etc. The second field contains the video frame's odd-numbered scan lines: 1, 3, 5, etc.

That is the way the original NTSC TV system's video signal, which is analog, not digital, is handled. For digital video such as is recorded on a DVD, the scan lines become pixel rows, and the blanking intervals are no longer needed, but the idea remains the same.

Thus, on a DVD sold in the U.S., there are video frames lasting 1/30 second each, with each frame containing two interlaced fields. Because the picture is recorded digitally, the scan lines are now rows of pixels: 480 pixel rows per video frame. The even-numbered pixel rows (0, 2, 4, etc.) are recorded as the first field of the video frame, while the odd-numbered pixel rows (1, 3, 5, etc.) are recorded as the other field of the same frame.

The scan lines or pixel rows of a single field of interlaced video are like an arrangement of slices of a scene viewed through a window, when the slices of the scene are separated by the partially opened slats of a venetian blind. For the second of the two fields in each video frame, the view and the slats simply exchange roles: the slats now contain picture information, while what was picture information becomes slats. This picture-slats alternation happens over and over again, with one alternation every 1/30 second.

Each field of normal, non-film-originated interlaced video records information from the video scene that occurs 1/60 second after the previous field, whether that field is in the same video frame or in the prior video frame. This delay between the two fields of a video frame (or between the first field of a new frame and the last field of the immediately prior frame) worked well with old, picture-tube based TVs, since those TVs would draw the two successive fields on the screen one after the other, at the exact same 60-fields-per-second rate at which the video camera captured them in the first place. Though the two fields were drawn on the TV screen one after another, 1/60 second apart, the fact that the glow of the phosphors on the picture tube's face decayed slowly, abetted by the "persistence of vision" of the human eye, made the two fields seem simultaneous.

When film at 24 frames per second is transferred to interlaced video at 60 fps, as for an analog NTSC broadcast or a standard digital DVD in the U.S., "2:3 pulldown" is used to match the disparate frame rates of film and video. But on a Blu-ray Disc, the same film is digitally recorded at its original 24 fps. A typical Blu-ray player today (but not the early Blu-ray player models from a few years ago when Blu-ray was first introduced) can play that recording into an up-to-date 1080p HDTV at 24 fps, using a digital HDMI connection between the Blu-ray player and the HDTV. The result is much better than even the best DVD can offer, even when the DVD is played on the same Blu-ray player and viewed on the very same TV.

With 2:3 pulldown for motion-picture film that is being transferred to video — a technique which is also called "3:2 pulldown" or "telecine" — film frames are scanned to make a new video field (not a video frame, a video field) every 1/60 second. (Find out more about telecine here; the word "telecine," by the way, is pronounced either with four syllables, or with three.)

After 2:3 pulldown, every group of five sequential video frames contains exactly three of those video frames whose fields both come from just one film frame. But two of the five video frames in a five-frame group contain fields that derive from two different film frames. These two (of every five) video frames can be called "dirty frames," since they do not come from a single, "clean" film frame.



In the illustration above, the first two of the five resulting video frames are "clean," coming entirely from film frames A and B, respectively. The fifth video frame is also "clean," since both of its fields come entirely from film frame D. But the third video frame has one field scanned from film frame B and one from film frame C, and so that video frame is "dirty." So is the fourth video frame, since it has one field from film frame C and one field from film frame D.

Film frames B and C (or C and D) may come from a single film camera shot and be only slightly different from one another — or even identical, if there is no motion in the scene. However, in a far worse case, they can cross an editing splice between two different film camera shots. If they do not cross a splice and are only slightly different, objects that are in motion from one film frame to the next film frame can develop edges that look serrated rather than smooth, when video frame 3 or video frame 4 in the illustration above is viewed on a progressive TV. If they come from different film shots and are accordingly grossly different, then when viewed on a progressive TV they can look a complete mess. They are not just "dirty," they are downright "filthy." Video frames that are "dirty," even "filthy," occur typically with 2:3 pulldown and can cause problems on progressive TV displays.

What does that name, "2:3 pulldown," actually mean? It means that the machine (called a "telecine") which scans the physical film and converts its frames into interlaced video for television transmission or DVD "pulls down" the film, using its sprocket holes, in a way that holds one frame of the film in place for 1/30 second, which is the time it takes to scan two 1/60-sec. video fields. Then it pulls the next film frame into position and holds it steady for the time it takes to scan not two but three 1/60-sec. video fields: 3/60, or 1/20, second. Then the next film frame generates two video fields, and then the next generates three fields ... and so on, and so on. That simple 2-3 alternation in the pulldown of the film in the telecine machine winds up creating a mixture or "clean" and "dirty" video frames like those in the illustration above.

Today's digital TVs are "progressive," not interlaced. They are designed to present each video frame intact, not divided into two alternating sets of interlaced scan lines. Progressive TVs can be thought of as using the scan lines in the picture all at once, 0, 1, 2, 3, etc., rather than using the even-numbered scan lines and then the odd-numbered scan lines.

The letter "p" is used to designate progressive scan. When we read that an HDTV is "1080p" (or "720p") the "p" stands for the fact that the TV uses a progressive, non-interlaced display. When we read that a certain TV or cable channel is "1080i," that means that its signal has the same 1,920-pixel by 1,080-pixel resolution as 1080p, but each video frame is transmitted as two interlaced fields, 1/60 second apart, not a progressive frame with all its scan lines used at once. (1080i is like NTSC video in that regard, in fact, except that NTSC, when digitized and recorded on DVD, is "480i" rather than 1080i. It has far fewer pixel rows/scan lines and much lower resolution.)

When video that has been recorded in interlaced fashion (first the even scan lines, then the odd) is shown on a progressive TV, it must first be "deinterlaced." For a 1080p HDTV, deinterlacing a received 1080i signal is a snap: just combine each pair of successive fields, transmitted 1/60 second apart, into a single frame, and show one frame every 1/30 second.

This results in a "1080i60" video input being displayed on the TV screen as "1080p30." The "30" designation after the "1080p" means that there are 30 progressive frames shown on screen each second, rather than 60 interlaced fields per second. Likewise, the designation "1080i60" for the input video signal means video with 60 interlaced fields per second, ostensibly intended for an interlaced display. (Sometimes these and similar designations contain an embedded slash: "1080p/30" or "1080i/60". And sometimes we see "1080p@30 Hz" or "1080i@60 Hz", indicating the screen "refresh rate" in cycles per second or Hertz.)

Typically in a 1080i60 video input stream, in each span of time that contains four film frames (4/24, or 1/6, second) there are five video frames that, after being deinterlaced, need to be flashed on a progressive HDTV's screen. The extra video frame magically appears because 2:3 pulldown is done, producing "dirty" video frames by creating two interlaced-video fields from some of the film frames and three interlaced-video fields from others. This is the same as what happens with 480i video, except it is done at an HDTV resolution.

The 2:3 pulldown cadence not only creates "dirty" frames when deinterlaced and displayed progressively, it also results in a somewhat herky-jerky rendition of the motion in the picture, called "telecine judder."


Films on Blu-ray discs avoid both "dirty" frames and telecine judder because a Blu-ray Disc is encoded at 1080p24, right on the disc. This means there are 24 full progressive-video frames recorded per second recorded on the disc, just as there are 24 film frames per second recorded on film. No 2:3 pulldown is done. (Often, by the way, the official designation "1080p24" is shortened to "24p".)

Provided that a Blu-ray player is capable of sending 1080p24 video direct to an HDTV, and provided that the TV is capable of accepting it and can put the received 1080p24 video right on screen without any sort of change being made to the picture content, there is no need for the player or TV to do 2:3 pulldown compensation (sometimes called "inverse telecine") at all. Nothing must be done to hide the serrated edges and other picture defects that are commonly associated with 2:3 pulldown's "dirty" frames, when they are used by a progressive display. Telecine judder disappears. Hence, Blu-ray Discs of a Hollywood feature film can be viewed with stunning, unprecedented fidelity to the original film.

Blu-ray players actually transmit 1080p24 video to the HDTV at 24 Hz, but the TV may use each field (say) three times in refreshing the screen at a refresh rate of 72 Hz. This can reduce visible flicker, which can fatigue the eyes. However, the actual content of the video changes at a rate of one field every 1/24 second. Since 72 Hz is a whole-number multiple of 24 Hz, there is no change to the picture content.

Some HDTVs using LCD panels now refresh the screen using a 120-Hz refresh rate, such that each frame of 1080p24 is used five times. (Likewise, each field of 1080i60 video input is deinterlaced, producing 1080p60, and then each frame of that is used twice to produce a 120-Hz refresh rate.)

Whatever the actual refresh rate, whether 72 Hz or 120 Hz, the video coming out of the Blu-ray player and being displayed on the screen is actually 1080p24. Its frames match those of the original film one-for-one.


That same film on DVD is recorded at 480i60, which means that each interlaced frame consists of two fields, each with 480 ÷ 2 = 240 rows of pixels. Each field takes 1/60 second, so the two fields of any given video frame take 1/30 second.

When 480i60 is shown on a progressive TV display, it must be deinterlaced. The simplest way is to use a "line doubler." A line doubler simply combines each pair of 480i60 fields into one whole frame and shows each whole frame twice, achieving in effect 480p60 video on the screen.

Line doublers work fine for 480i that derives from a video camera, but for 480i that derives from a film that has been subjected to 2:3 pulldown, line doublers make two of every five displayed video frames "dirty" ones. Even when these frames are not "filthy" — spanning splices between film camera shots — objects in motion tend to develop serrated edges. Fast motion results in edges with pronounced comblike "teeth." The result can be the picture artifact or defect known as visible "combing":



Combing artifacts are sometimes not-so-affectionately referred to as "jaggies." The problem shows up when line-doubler deinterlacing is done for progressive TV displays. For an interlaced display such as a CRT ("picture tube") there is no need for deinterlacing. Each field is drawn sequentially onto the phosphors of the picture tube, scan line by scan line, with each field appearing 1/60 sec. after the previous one. There is no visible combing.

With a progressive TV display, combing that results from deinterlacing can be reduced by having the TV do more sophisticated deinterlacing than a line doubler does. But it can often be eliminated entirely by a DVD player, before the TV even "sees" the video signal.

Digital 480i video on DVD can optionally be recorded using a "soft telecine" type of 2:3 pulldown. Recall that in 2:3 pulldown certain video fields are used twice, once in each of two successive video frames. "Soft telecine" is a way to use "flags" recorded on the DVD to say, in effect, "The video field you would normally expect to occur here is missing. Repeat the corresponding video field from the previous video frame instead."

A "progressive-scan DVD player" can change 480i on the disc to 480p for output to the TV (assuming the TV can accept the doubling of the rate at which the TV signal is input to it). Such a player can typically use the "soft telecine" flags, if they are present on a DVD, to eliminate "dirty" progressive frames in the deinterlaced output. The basic logic is simple: just ignore repeated fields entirely. However, the frame rate generally has to be padded out from 60 fields/30 frames per second to 60 progressive frames per second, so certain of these whole frames are repeated as necessary, instead of repeating fields as needed. This eliminates combing/jaggies, but it can result in a noticeable judder on screen.

There is one further problem with 2:3 pulldown on DVD. Some DVDs use a "hard telecine" instead of flags to accomplish 2:3 pulldown, such that repeated fields are actually encoded twice, and few DVD players are able to remove it. Likewise, many "soft telecine" DVDs are poorly encoded: the flags that tell when to repeat fields are missing or inserted in wrong places. This often happens near a chapter break, at which point even a sophisticated progressive-scan player can produce on-screen jaggies.

Blu-ray players, when they are used to play DVDs, have exactly the same problems with jaggies and judder as DVD player do. But when they are used to play Blu-Ray Discs made from movies or TV shows shot on film, playback is stunningly smooth and artifact free.