Saturday, December 12, 2009

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

Not too long ago I posted 1080p24 ("24p") on Blu-ray to talk about how, for the first time on home video, a motion picture is rendered on Blu-ray disc just the way it was recorded on film: at 24 full video frames per second. Using "progressive" 1080p video, films on Blu-ray can be stunningly realistic when viewed on a 1080p HDTV connected to the Blu-ray player via HDMI. In contrast, I said, DVDs generally don't store progressive video, at the 24 frames per second used by motion picture film, or at any other frame rate.

Recently, an anonymous person added this comment to my post:

Dear Bonehead,

I stopped reading your article partway through because you obviously know NOTHING about the DVD standard. NTSC DVD's support both "30fps" (interlaced video) and 24fps (progressive scan video); in fact, most Hollywood DVD's use 24fps when the original source is film to save on disc space (by eliminating 6 fps of interpolated data), as the DVD player will do 2:3 pulldown as necessary when outputting to an interlaced display. Blu-ray just adds extra resolution to the image, not a "breakthru" in frame rates or progressive scanning. Bother to do your research before blathering on the internet like you are some sort of expert. I don't know everything, but I do know something about DVD's, as I work with encoding and authoring them everyday for my job.

Now, I can take being called names like "Bonehead" when my name-caller has any kind of a good point to make. In this case though, not only was the anonymous commenter rude, but he was just plain wrong, to boot. So here is my rejoinder:

Sorry, you who so impolitely called me "Bonehead," but I beg to differ:

On DVDs sold in the United States, film-derived video is recorded in such a way as to result in 480i output from a standard DVD player — that is, interlaced video with two fields per video frame, separated by a time interval of approx. 1/60 second between fields. The frame rate is thus approx. 30 frames per second.

The typical "progressive scan" DVD player (one that does no "upconversion") can derive 30 complete, non-interlaced frames per second from film-based video on an NTSC DVD and output those frames as 480p video on a component-video connection, into a TV whose component-video input can handle the bandwidth of 480p. It cannot, however, output progressive video at 24 fps. Only a Blu-ray player, using a Blu-ray disc, can do that.

To derive progressive output at 30 frames per second from film-based material, the progressive scan DVD player can simply take each video field and "double" the scan lines in it, to make up a full frame with 480 scan lines. This is the "line doubler" approach.

A smarter approach is to do "3:2 pulldown compensation," a.k.a. "2:3 pulldown compensation" or "inverse telecine." Ideally, this process faithfully recreates the 24 frames per second of the original film —— but then 6 of those 24 frames are repeated(!) to bring the video-output frame rate up to 30 frames per second. This is done because few if any of the TVs that were available when progressive scan DVD players were introduced were able to accept video at 24 frames per second.

On the DVD itself, the encoding is 480i. However, film-based material is usually — but not always — encoded in such a way that the video fields that need to be repeated (in "2:3 pulldown") to change the 24 frames per second of film into the 30 frames (or 60 fields) per second of NTSC video are flagged, with the DVD player being expected to use the flags to re-output the fields as necessary.

This use of flags to tell the player how to do 2:3 pulldown is called "soft telecine." In "hard telecine," the fields are actually repeated on the disc. Few progressive scan DVD players know how to compensate for "hard telecine." They typically do know how to compensate for "soft telecine" — but, unfortunately, many DVDs using that type of encoding have portions where the flags are missing or improperly used, resulting in imperfections in the output cadence until the flags get back in sync.

"Soft telecine" 480i DVDs record exactly 24 frames per second, or 48 fields per second. They accordingly bear a superficial resemblance to the "24p" recording of a film on Blu-ray, but the latter is truly recorded as progressive video at 24 frames per second, while the former records discrete fields of interlaced video at 48 fields per second.

In saying the above, I realize that I am in disagreement with the Wikipedia article on 24p.

The section "24p on DVD" states, "DVDs, however, are capable of storing the native 24p frames." This is possibly true. It also states, "Every Hollywood movie is laid to disc as a 24p ... stream." This is definitely not true.

True 24p is at best an optional way to encode DVDs that is rarely if ever used. It is not actually used for "every Hollywood movie." See this page at the website dedicated to the Handbrake video transcoder software, if you don't believe me about most or all film-based NTSC DVDs being either "soft" or "hard" telecined. Also see this page about the MPlayer and MEncoder software. The format discussed at "11.2.2.2. Telecined" is the one used on virtually every film-based NTSC DVD.

Still don't believe me? Let me refer you to perhaps the ultimate authority on DVDs, Jim Taylor, who wrote the book DVD Demystified and maintains the Official DVD FAQ. He says, in "What's a progressive DVD player?":


A progressive-scan DVD player converts the interlaced (480i or 576i) video from DVD into progressive (480p or 576p) format for connection to a progressive-scan display (31.5 kHz or higher) ... There's enormous confusion about whether DVD video is progressive or interlaced. Here's the one true answer: Progressive-source video (such as from film) is usually encoded on DVD as interlaced field pairs that can be reinterleaved by a progressive player to recreate the original progressive video.

OK, that's about the size of it, then. Virtually all NTSC DVDs in the United States that were sourced from film have telecined video encoded on them, which means interlaced, not progressive, video. Any more questions, Anonymous?

19 comments:

michal pfeil said...

Love the rebuttal! And thanks for the handy links.

Bobby said...

I looked at the frame rate of some of my DVDs (movies that were originally 24fps) in MPEG Streamclip, and it lists them as 23.976. Does this mean that the file on the DVD is 23.976, or is the movie actually split into 47.952 fields per second (twice the rate of 23.976)? Is the program just telling me what the original frame rate is, but the actual file on the DVD is interlaced? If so, that would mean each 24fps movie is stored on a DVD using 48 fields per second (not exactly that, but close enough.

Also, when I click through these DVDs on my computer or TV, it takes 24 frames for one second. Why isn't it 30fps? Is that because my computer and TV are progressive displays and its undoing the pull down and displaying them at 24fps? Would it take 30 frames for one second to pass on an analog TV? Thanks in advance for any info!

eric said...

Bobby, I haven't really thought about this subject for years, so my recollection is fuzzy. Also, MPEG Streamclip no longer seems to work on my Mac, which is now running Mac OS El Capitan 10.11.6. As best I can recall, all Region 1 NTSC DVDs, whether or not they are sourced from 24-fps film, all record interlaced video at slightly less than 60 fields (or slightly less than 30 frames) per second — never 48 fields (24 frames) per second. This was done to make them compatible with "ordinary" NTSC TVs which can handle only (slightly less than) 60 fields per second.

24-fps film is transferred to DVD using a process called, alternately, "telecine" or "3:2 pulldown" or "2:3 pulldown." This process generates duplicates of certain of the video fields.

Ideally, the presence of duplicated fields from the telecine process can be avoided by the use of 1-bit "flags" on the DVD. The DVD player knows how to honor those flags, which say that the field the flag accompanies is to be rendered a second time when needed.

But progressive-scan DVD players, when used with compatible TVs, and computer apps like MPEG Streamclip and VLC need to render entire video frames, not video fields, so they have to put pairs of video fields together into single frames. Depending on the app and the DVD itself, if you go frame-by-frame through a movie-derived disc on your computer, you can sometimes see frames that are stitched together from two separate film frames, offset slightly in time, and such frames don't look quite right when they are presented one frame at a time.

The frame rate reported in the Media Information window by a computer app such as VLC (I can't be sure about MPEG Streamclip) is probably going to be that at which a film-derived DVD is nominally recorded — not that at which it is currently being displayed. The actual frame rate you see displayed at "normal" playback speed on your computer screen may be different, depending on whether the app is told to "deinterlace" the video and what mode of deinterlacing is chosen.

I'm guessing that this leaves you with further questions, so if you want to pose them, feel free to leave another comment.

Bobby said...

Thanks for the reply Eric! I’ve been reading a lot about this issue and talked with some editors online about it (I’m an editor too) and one thing that I’ve heard is that the most common method to put a 24fps movie on a DVD is to first slow down the movie to 23.976 fps, which is basically a progressive file, although technically it’s interlaced because each frame is split up into two fields, with both fields in each frame representing the same moment in time. These are called “progressive segmented frames,” but are still considered interlaced since they are separate fields with one containing the odd lines and the other the even lines. This is why I thought that a 24 fps movie on a DVD is actually 48 fields (progressive segmented frames) per second. Although technically it would be 47.952 fields per second (23.976 x 2).

Then this “progressive” file is encoded on the DVD with flags that tell an interlaced DVD player to repeat certain fields so that the movie will play at 30 fps, or 60 fields per second (actually 59.94). I’ve read that it’s rare for a filmed movie to be encoded as an interlaced movie with 2-3 pulldown because then more space is being used up on the DVD. In this case, the movie really is a 30 fps movie (or 29.97 fps, aka 59.94 fields per second). But in the majority of cases, instead of encoding the repeated fields, the flags tell the player to repeat them, which would mean the file on the DVD is 24 fps (I’ll use whole numbers from now on, since it’s simpler), which is 48 progressive segmented frames per second.

So when played using an interlaced DVD player on an interlaced display, the DVD player will do the 2-3 pulldown and output 59.94 fields per second. Something I’m not sure of is how many frames represent one second when watching a movie on an interlaced display? Will it take 30 frames to go one second, or will it take 24? By “how many frames will it take?” I mean how many clicks of the remote. When I click through a DVD on my LCD TV, it takes 24 clicks for one second to go by. But I wonder if it would take 30 clicks with an interlaced player and interlaced display.

Using a progressive player with a progressive display, the DVD player ignores the flags encoded on the DVD and just shows whole frames. If the movie has the 2-3 pulldown encoded on the disc, then the film will look worse on a progressive display than on an interlaced display because on a progressive display the fields aren’t separated, so each frame will be two distinct moments in time combined, which might produce the “jagged edges” effect.

Now, I’m a bit confused because in your reply you said ideally the presence of duplicated fields can be avoided using flags, and you also said all 24 fps movies on a DVD are 30 fps. If the duplicated fields are encoded then it would be 30 fps, but if they aren’t, wouldn’t the file itself be 24 fps?

Bobby said...
This comment has been removed by the author.
eric said...

Bobby, I'm getting confused because you originally extended your comment above with a lot more material about your topic, which is an interesting one. I confess it has been years since I thought about such questions. I'm going to post what will probably turn into a series of comments, owing to the fuzziness of my current thinking on the topic at hand and the fact that there is a limit to how many characters I can put in any one comment.

You asked whether “progressive segmented frames” (PSF) are used on DVDs. I’d never heard of “progressive segmented frames” (PSF) until now. I’ve just tried to find some authoritative source to tell me whether PSF is ever used in encoding DVDs. (Although the Wikipedia article on PSF says the DVD of “Terminator 2: Judgment Day” contains that type of encoding, the article that is referenced as a source for this supposed ”fact" is actually about the BluRay disc.) So I am dubious about the existence of actual PSF encoding on DVDs.

Let's keep in mind a crucial difference between PSF and interlaced video. In PSF, the two fields of a video frame are from the same moment in time, because they derive from a single frame of film. In interlaced video, the two fields of a video frame are separated by a small amount of time, because that's how an interlaced video camera (NTSC, for instance) generates its output.

The video output from a standard (as opposed to progressive) DVD player, if it is being used with a standard-definition TV, has to be interlaced NTSC video, at least here in this country where NTSC was once the only broadcast TV standard.

What's encoded as MPEG-2 video on a Region 1 DVD itself, according to Jim Taylor's 1998 book "DVD Demystified," pp. 116-117, can be stored in either progressive or interlaced format (no mention of PSF). Signals originating from video cameras — presumably NTSC video cameras, that is — "are usually stored as a combination of interlaced and progressive blocks, depending on the motion between blocks."

But, Taylor says, "inherently progressive" film "is usually stored on DVD as frames along with the information to tell the player how to split and repeat the frames as fields." This information is basically in the form of flags, apparently. The flags cue the player as to how to do 3:2 pulldown on the fly. Taylor says, ”Unfortunately, this causes motion artifacts, since two of every five television frames contain fields derived from two different film frames." The motion artifacts being referred to are generally ”jagged edges."

By "usually stored on DVD," I think Taylor means that some DVDs actually store all of the NTSC video fields that 3:2 pulldown would create, and they fail to include flags that enable the player to do 3:2 pulldown on the fly. Such discs use "hard telecine," that is, while the discs with flags use "soft telecine."

I remember hearing many complaints from technically savvy folks in the early days of DVDs that lots of soft-telecined movie titles were improperly flagged and caused various playback imperfections.

When progressive-scan DVD players came along, TVs needed to provide component video inputs in order for progressive DVD output to work. Then the player could simply skip doing internal 3:2 pulldown for movie-based material ... but the (presumably soft-telecined) disc needed to contain properly organized flag bits to get optimal results.

None of these scenarios, however, employ progressive segmented frames.

(End of Part 1.)

eric said...

(Part 2 of my answer to Bobby.)

Still, I'm trying to think of a good reason why PSF video couldn’t be stored as such on a Region 1 DVD. The encoding could fake "soft telecine" such that there are two fields stored per film frame, along with the appropriate flags that would allow the player to construct 30 fps interlaced video output when necessary. Progressive-scan output would seemingly work fine as well.

Here is a thought experiment relating to the possibility of “progressive segmented frames” (PSF) on Region 1 DVDs that need to work with the old DVD players of yore and with the old analog NTSC TVs. The video connection between the player and the TV is assumed to be a composite video cable with the usual yellow plugs:

(Stage 1.) Suppose I point a stationary film camera at a painting in an art gallery. I let it capture a two-hour movie of the painting. Because the camera is not moving, every film frame is exactly the same as every other film frame. Now I take my film and go through the process of transferring it to video and putting the video on a standard Region 1 DVD. To do the latter, I use the soft-telecine method. The DVD I make is, I assume, properly flagged so that when the DVD player generates interlaced NTSC video output at 30 fps, the internal ability of the player to output the necessary “extra” video fields — i.e., fields that would be stored redundantly on the DVD itself in “hard telecine” — works fine. And the picture on an old-fashioned interlaced analog NTSC TV looks fine.

(Stage 2.) I take the same exact film from Stage 1 and put it on a second DVD using the output of some specifically PSF-oriented process instead of the output of a “soft telecine” process. Clearly, this DVD will also need to be properly flagged, for the same reason that the first DVD did: so that when the DVD player generates interlaced NTSC video output at 30 fps from the DVD, the internal ability of the player to output the necessary “extra” video fields works fine.

I could be wrong about this conclusion, but I believe the two DVDs might be bit-for-bit identical! In part, this is because there is no visible motion, and accordingly no MPEG-2 motion vectors have to be encoded on the DVDs. If the two DVDs are bit-for-bit identical, then it may be the case that PSF-oriented video on DVD and “soft telecine” video on DVD are, or at least may be, synonymous. (I’m assuming that the “soft telecine” video on DVD is properly flagged, of course.)

eric said...

(Part 3 of my answer to Bobby.)

At one point you said in your original second comment, now deleted, “This tells me all video content on a DVD is interlaced, but not necessarily 30 fps (60 fields per second).” I think you may be mistaken about all video content on a Region 1 NTSC DVD being interlaced, because Jim Taylor says in a footnote on p. 116 which is associated with the discussion I quoted above: “In MPEG-2 encoding, the decision between progressive and interlaced format can be made all the way down at the macroblock level. The DVD-Video specification limits MPEG-2 video to nonprogressive sequences, which can contain both progressive and interlaced frames. Interlaced frames can further include both progressive and interlaced macroblocks. Since progressive macroblocks are more efficient (using one motion vector instead of two), even an interlaced source is typically encoded by using more than 50 percent progressive macroblocks.”

Also, you asked about the number of frames per second ("clicks") using frame-by-frame advance with a standard player on an interlaced TV. Would there be 30 clicks per second of film content? My guess is yes, but I can't test it because I no longer have a standard DVD player or an interlaced TV.

I think I should accordingly add here that the questions you are asking make sense from the point of view of today's digital HDTVs, given that they have the ability to handle so many different video formats, resolutions, frame rates, refresh rates, etc. But at the time DVDs and the first players first arrived, it was crucial, when designing the DVD-Video specification using MPEG-2, to insist that Region 1 DVDs and their players offer NTSC compatibility over composite video connections. That could have been implemented via "hard telecine" and simply not specifying any other method, but clearly someone was thinking about the need to not waste space on the discs. So they came up with the idea of "soft telecine" performed by the player itself, based on flags on the discs. It seems to be the case that someone invented the idea of “progressive segmented frames” slightly later than the time frame in which the DVD-Video spec was laid down. PSF seems to be more straightforward to me, with a lot more applications than just how movies get rendered on DVD. But "telecine" (including "hard" and, later on, "soft" varieties) was an idea with a long history when it came to NTSC video.

Anyway, if your interest persists in this topic or if I've failed to answer all your questions, I'd be happy to hear back from you.

Bobby said...
This comment has been removed by the author.
Bobby said...

Eric, your three posts in reply to me are very interesting! I must’ve been confused about this topic. I thought that since DVDs are 480i, this must mean all content on a DVD is interlaced. And the idea of progressive segmented frames made sense to me because that would be how to turn progressive material into interlaced material. But if truly progressive content can be stored on a DVD, why are all DVDs 480i? Is it impossible for a DVD to be 480p?

In your “Revisiting 1080p24 on Blu-ray” blog post you said…
"Soft telecine" 480i DVDs record exactly 24 frames per second, or 48 fields per second. They accordingly bear a superficial resemblance to the "24p" recording of a film on Blu-ray, but the latter is truly recorded as progressive video at 24 frames per second, while the former records discrete fields of interlaced video at 48 fields per second.”
The 48 fields that you mentioned in this quote sound like progressive segmented frames to me. So I took this to mean that for a 24p film DVD, the file can be stored as 48 progressive segmented frames per second, with flags to repeat fields as necessary for interlaced display. But for progressive players and displays, the flags could be ignored and the 48 fields could easily be combined to show true 24 progressive frames per second.

I’m confused because in your first reply to me you said “all Region 1 NTSC DVDs, whether or not they are sourced from 24-fps film, all record interlaced video at slightly less than 60 fields (or slightly less than 30 frames) per second — never 48 fields (24 frames) per second.” These two quotes sound contradictory to me, unless I’m misunderstanding something…?

Since I was under the impression that all content on a DVD had to be interlaced, I thought that the only way to have a progressive film on a DVD was to split up each film into two fields, which would technically be interlaced, although still represent the same moment in time. But then you said something I thought was very enlightening. You quoted Jm Taylor saying that progressive content can be stored on a DVD along with information to tell the player how to split and repeat the frames as fields. I thought that the flags encoded on a DVD only tell a player how to repeat certain fields; I didn’t know that there were also flags that tell the player how to split progressive frames into fields. If this is true, then shouldn’t it be pretty easy to for a progressive player to ignore the flags and output the 24 fps on a progressive TV? If the TV had a refresh rate of 72 Hz or 120 Hz, or any number divisible by 24, then the TV player itself wouldn’t have to do the 2:3 pulldown (like it would on a 60 Hz display), and you’d get true 24 fps content.

So, when people say that all DVDs are 480i, doesn’t this actually mean they were intended for interlaced display but could also displayed as 480p?

On a side note, I was wondering if Blu-rays are true 24 fps content or if they’re actually 23.976 fps. Since filmed movies on DVD are 23.976 (so that with all the interlacing and pulldown they could be displayed on interlaced TVs with a refresh rate of 59.94Hz), does this mean most movies on Blu-ray are 23.976 too? Is a TV with a refresh rate of, say, 120 Hz really 120, or is it really 119.88 ( which is double the amount of 59.94)? It seems to me that the refresh rate of TVs is what makes a company decide what frame rate to choose.

Bobby said...
This comment has been removed by the author.
Bobby said...

Eric, I am also confused! When I tried to publish my post it told me it was too long, so I split it up into two posts. Then I decided that I wanted to add something to my second post so I deleted it, and tried to post the updated one. But after I posted it, it got deleted and I have no idea why! I tried several more times to publish it and kept getting deleted. Anyway, I thought I’d try to post it again for the sake of completeness. Luckily I had it saved in a text file. I hope it works this time!

“I had to split up my post because I was told it was too long! I also wanted to add something so I deleted the last post and updated it here:

In your original “1080p24 on Blu-ray” article you said “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.”

I’ve heard this is the least common method for putting a film on DVD because then there’s no progressive file on the DVD that would look good on a progressive TV using a progressive player. Isn’t this only one method? The other method is to leave the film as it is with the player reading encoded flags and doing the 2-3 pulldown itself.

Also, you mentioned that a progressive DVD player used with a compatible display might show frames that are stitched together from two separate film frames. I agree with this, but this would only happen on a 60 Hz TV, right? If the TV is 72 Hz or 120 Hz (which is common now) then every single frame will be seen without any frames being mixed, because a 120 Hz display will show every frame intact 5 times each for 1/120 of a second, which would be the same as 24 fps. You described this process for Blu-rays and I read that the same process will happen with a DVD if the file on the DVD is a 24 fps file with flags telling the player to do 2-3 pulldown, instead of the 2-3 pulldown being encoded (which results in repeated fields being encoded).

You told me that "all Region 1 NTSC DVDs, whether or not they are sourced from 24-fps film, all record interlaced video at slightly less than 60 fields (or slightly less than 30 frames) per second — never 48 fields (24 frames) per second." But in your blog post you said "Soft telecine" 480i DVDs record exactly 24 frames per second, or 48 fields per second. They accordingly bear a superficial resemblance to the "24p" recording of a film on Blu-ray, but the latter is truly recorded as progressive video at 24 frames per second, while the former records discrete fields of interlaced video at 48 fields per second." So then DVDs can store 48 fields per second?

You quoted Jim Taylor’s DVD Demystified website with: “There's enormous confusion about whether DVD video is progressive or interlaced. Here's the one true answer: Progressive-source video (such as from film) is usually encoded on DVD as interlaced field pairs that can be reinterleaved by a progressive player to recreate the original progressive video.”

This tells me all video content on a DVD is interlaced, but not necessarily 30 fps (60 fields per second). How can a 24 fps movie be 30 fps if the repeated fields aren’t actually encoded on the DVD? Wouldn’t the movie be 24 fps and only seen as 30 fps on an analog TV, but shown at its native 24 fps on a progressive TV with a refresh rate that’s 72, 120, or any other rate in which 24 can be divided evenly into?

Thanks!”

eric said...

Bobby, All DVDs are 480i, not 480p, because of the need to make the DVD-Video specification work commercially. 480i is the digital equivalent of analog NTSC, the sole broadcast TV standard in the U.S. back then. No one had digital TVs or TVs that could reproduce 480p, when DVD-Video was first standardized. Also, the honchos wanted to make sure the new disc format could fit an entire movie on a single-layer DVD, since it took a while for dual-layer DVD technology to become commercially available. I think that's why they came up with the idea of putting flags rather than redundant video frames on discs, and requiring that DVD players be able to generate 3:2 pulldown on the fly from those so-called "soft telecine" discs. Also, progressive DVD players and TVs that could take advantage of them were envisioned, but not yet a reality. So that's why both "soft telecine" and "hard telecine" (but apparently not PSF) were incorporated in the DVD-Video specification.

I slightly misspoke when I said all Region 1 NTSC DVDs contain interlaced video. Taylor's "DVD Demystified" book, in a footnote on page 116, says that all MPEG-2 "sequences" of frames on DVD-Video have to be nonprogressive — albeit that the broader MPEG-2 specification does allow progressive sequences — but the frames in each sequence can be both progressive and interlaced, as can the "macroblocks" that each frame is composed of. Progressive macroblocks are more efficient because there is just one "motion vector," not two. So "even an interlaced source is typically encoded by using more that 50 percent progressive macroblocks." That said, the video output of the decoder in the basic, non-progressive DVD player is always rendered as interlaced so as to be NTSC-compatible. Get it?

You asked: "I didn’t know that there were also flags that tell the player how to split progressive frames into fields. If this is true, then shouldn’t it be pretty easy to for a progressive player to ignore the flags and output the 24 fps on a progressive TV?" Go to http://www.dvxuser.com/V6/archive/index.php/t-186811.html and look at David Jimerson's post towards the bottom for a discussion of the flags. It's in the section labeled "3-2 Pulldown." And yes, ignoring the "repeat_first_field" flag is how progressive players do their thing. But I assume the progressive player still needs to honor the "top_field_first" flag so as to know which field of an interlaced frame has the odd-numbered scan lines and which has the even-numbered. Jimerson is assuming here that all DVD-Video is interlaced, despite that Taylor says otherwise.

"So, when people say that all DVDs are 480i, doesn’t this actually mean they were intended for interlaced display but could also displayed as 480p?" Well, yes, but you may be tending to mix up how DVDs are encoded internally and how they can be displayed externally.

As for true 24 fps vs. 23.976 fps on Blu-ray, Blu-ray is capable of either one. See http://forum.blu-ray.com/showthread.php?t=102488 for more. Apparently, most movie-based Blu-rays use 23.976 fps rather than 24 fps.

Bobby said...

Thanks, Eric. In that post by David Jimerson in the link you gave, he said all 24 fps movies must be converted to 60 fields per second for DVD. It’s hard for me to understand this when flags are involved to tell the DVD player to repeat fields. I’m trying to visualize this concept. It’s easier for me to visualize it when the 2-3 pulldown is already encoded in the video file so that no flags are used. In this case, I’m imagining one frame being split into two fields (which I think of as progressive segmented frames) and then having certain fields duplicated. PsF are technically considered interlaced because even though they represent the same moment in time, one field has all the odd lines and the other has all the even lines. So four frames are turned into eight fields, and then there are two repeated fields, making a total of ten fields (five video frames for every four film frames), with duplicate fields being stored on the DVD. I understand how this is 60 fields per second.

But instead of encoding duplicate fields (which uses up more space), the other method is to split up each frame into two fields, and have flags that tell the player to repeat certain fields. I’m trying to visualize this as being 60 fields per second on the DVD and I don’t understand it. The only way I can think of it is that for every second, the 24 frames are turned into 48 fields with instructions (aka flags) that tell the player to repeat certain fields. But the file on the disc itself is 48 fields per second. So when people say all DVDs are 60i, does this mean they are displayed as 60i but can be encoded as 48i?

There are two things Jim Taylor said about this that sound contradictory to me.
The first one is the quote “Progressive-source video (such as from film) is usually encoded on DVD as interlaced field pairs that can be reinterleaved by a progressive player to recreate the original progressive video.”

And in one of your replies to me you said that Taylor said "inherently progressive" film "is usually stored on DVD as frames along with the information to tell the player how to split and repeat the frames as fields."

In the first case he’s saying each frame is split into fields and encoded that way, which would be progressive segmented frames. In the second case, it sounds like he’s saying entire frames are stored, and it’s the player that splits up the frames into fields with the use of flags. So I’m just trying to understand which of these is true, or if both methods are used.

eric said...

Bobby, try reading http://www.bigdish.info/other/pgscan.html for quite a detailed rundown (title: "DVD Benchmark - Part 5 - Progressive Scan DVD - October, 2000") on the various ways 24 fps film-mode video can be stored on a Region 1, NTSC-compatible DVD. You might want to search the page for the name of a flag such as "top_field_first" so that you can home in on the specific material that interests you. Or, search for the section title "How progressive players work".

The flags that are key are "picture_structure", "progressive_frame", "repeat_first_field", and "top_field_first". Look at the series of 3 tables showing 3 different ways that 4 film frames can be encoded on a DVD. And note that the discussion says "And there are dozens of other legal variations".

Notice that the term "MPEG picture" is being used rather than "video field" or "NTSC field". So maybe it's best to think of what's encoded on the DVD as a series of "MPEG pictures" rather than a series of either frames or fields — even if the term "field" appears in the flag names "repeat_first_field" and "top_field_first", and the term "frame" shows up in "progressive_frame". Once the mental focus shifts to "MPEG pictures," the question of whether "frames" are "progressive segmented" or not recedes in importance. Remember: "progressive segmented frames" were not yet on the radar screen when DVD became a reality.

Also keep in mind that some DVDs fail to set their various flags — e.g., "progressive_frame" — correctly. You should read the section on "Film-mode de-interlacing ("Re-Interleaving")" for more about the problems that real-world progressive-scan DVD players can run into when trying to deal with film-mode video on DVD, and the various methods they use to set aside those concerns.

It appears, then, that there are no simple answers to your questions!


Bobby said...

That is a great link Eric! I actually read that article a couple weeks ago when I was reading about 2-3 pulldown and I forgot how informative it was. After reading the parts of it that you suggested I feel like I understand it better, even though it’s more complicated than I thought! Previously I was trying to picture what progressive and interlaced sequences “looked like” since I’m used to video editing on a timeline, but it seems like a sequence on a DVD is more of a concept than a visual representation.

I have enjoyed talking with you about this subject and thanks for all the useful info! Take care!

Bobby

Bobby said...

One last thing. According to http://www.mplayerhq.hu/DOCS/HTML/en/menc-feat-telecine.html, which you gave a link to in your blog post, all video intended for NTSC television that’s put on DVD is stored as frames, not fields. It says video that was originally 60 fields per second is turned into 30 frames per second, and video that came from 24 fps film usually stays at 24 fps. The only time it doesn’t stay as 24 fps is when it was telecined before being stored on DVD, in which case the film would be 30 fps on DVD. So everything on DVD is either 24 fps or 30 fps (actually 23.976 and 29.97), but when displayed on an interlaced display becomes 60 fields per second.

If this is true, I wonder why NTSC video is stored on DVD as frames instead of fields. I understand keeping video that came from 24 fps film as frames rather than fields, but why change video that was originally 60 fields per second to 30 fps on a DVD if the TV is going to display it as 60 fields per second anyway? Is there some advantage to storing the content as frames instead of fields on the DVD?

eric said...

Bobby, Keep in mind that in NTSC video, every frame is composed of two fields. So NTSC video is stored on DVD as both frames and fields. The two are not mutually contradictory!

Bobby said...

Very good point Eric!