Why 480p and not 1080? (1 Viewer)

[John_Berger]

Now i am unsure if the video on DVD is stored at 23.976(film) or 29.97/25 fps NTSC/PAL.

I've found that some DVDs are 29.97, some are 24, some are 23.976. I haven't really found any serious pattern, although DVDs from similar studios or groups (like the Monty Python episode DVDs) usually have the same frame rate.

 

[Vince Maskeeper]

Just out of curiosity- where and how are you determining frame rate? Anything passing to a standard TV from a set top DVD player should be a ntsc compatible 30fps.

-Vince

 

[Justin Ward]

I was just thinking. If the 60hz AC could interfere with the 30fps video, why is it that PAL AC(50hz) does not interfere with its video(which is 25fps exactly)?

 

[Robert Cook]

Drew Eckhardt wrote:

Quote:

Let's not make too many associations that are not necessarily true. In the NTSC world, movies and TV shows that are 24 fps or 23.976 progressive (there are such TV shows, which were shot on film) are typically stored as 23.976 progressive on DVD. This does not have to be the case, however, and I'm sure there are exceptions, which would waste space. Interlaced video running at 59.94 fields per second is stored as interlaced video at 59.94 fields per second, which is also directly supported by MPEG-2 and DVD.
In the PAL world, movies and TV shows shot at 24 fps or 25 fps are (or can be) stored at 25 fps. Interlaced video running at 50 fields per second are stored as interlaced video at 50 fields per second.
And in case anyone has any questions in their mind, from the point of view of the DVD format, 16:9 (so-called anamorphic) and 4:3 pictures are treated exactly the same on disc--it's just 720x480 (or 720x576 for PAL) pixels. There's an embedded flag that tells the player how to handle it, and that's it. Any squishing or stretching of the picture that is necessary for proper 16:9 or 4:3 representation on disc must be done prior to encoding.

 

[Vince Maskeeper]

So does the DVD standard have a specification concerning how the frames are encoded (interlaced vs. Progressive)? It seems that many people say the frames are "stored" progressively- while I have been repeatedly told that the frames are interlaced with flagging which would help in reassembling a progressive stream.

This also begs the question concerning the 24 frame encoding- is this raw progressive frames, and if so the player actually builds a 30fps interlaced NTSC output from 24 frame progressive? If so, how come my HTPC has such a problem playing it back at 24 progressive frames (72hz)- if that is the native way it is stored.

Also, if this is true- what fps rates can DVD players handle- just 24 and 30? What made them decide to make the player able to alter the frame, but not process the resolution to a different output?

This is very interesting- and I wonder if you don't mind pointing to the source for this info? I still find it amazing that they would strike a 24 frmae telecine for DVD (I'm unsure- does HD support s straight 24 frame encoding?)- which might only be useful for the DVD format. Also, if NTSC is the intended output of DVD- why put the strain on the player by encoding at 24 in the first place, when you could simply telecine at 30 and strike a master compatible with all NTSC formats?

Thanks
Vince

 

[John_Berger]

Just out of curiosity- where and how are you determining frame rate? Anything passing to a standard TV from a set top DVD player should be a ntsc compatible 30fps.

Since my laptop doesn't have a DVD player :frowning: I used to convert some of my more-watched DVDs to AVI to compensate for the terrible movie selection in hotels when I travelled. Although it's been about a year since I've done so, I remember that some of the resulting files were out of sync, so I needed to adjust them, usually to 24 fps or something just slower. Now that I think about it, 23.976 was very common. I remember not knowing about that frame rate and wondering why I had to tweak it to such an odd number. Now I know. It wasn't as odd as I thought.
My Monty Python episode DVDs, however, were very easy. I made every episode into its own AVI, and every single file had to be changed to 29.97 fps. No exceptions. I didn't even have to think about it.

 

[Robert Cook]

Vince Maskeeper wrote:

Also, if NTSC is the intended output of DVD- why put the strain on the player by encoding at 24 in the first place, when you could simply telecine at 30 and strike a master compatible with all NTSC formats?

It potentially provides 25% more bandwidth or running time (depending on how you use it), which is a major benefit, in my opinion.

 

[GregK]

Hi Robert & company,
I am going to take the theoretical devil's advocate here (It is theoretical, as I don't personally author DVDs so don't take my statements as gospel !) ..and make the following bold statement: ALL current DVDs (meaning 24fps film shot material) are encoded as interlaced. I am a television engineer and have been reading up on the MPEG-2 standards for the last ten years. So does that make me an expert? HA! ..I wish!! Are there still parts I don't understand? You bet!! So with that humble intro, let's dive in!!
My understanding of the DVD format (meaning for the most part: the MPEG-2 bitstream) is that when video is encoded it can be flagged if it's NTSC/PAL interlaced material @ 60 or 50 fields per seconds respectively, or at 30fps for film or computer graphics video. ..Or if it's originally 24fps film. I don't believe this mode (24fps "true" progressive) is currently in use, as compressionist would rather interlace an image themselves and filter for artifacts that may arise when needed. But if it is 24fps encoded as 60 field NTSC, obviously 3-2 pulldown has been applied to fit 24 frames per second into 60 fields. Now MPEG-2 is always trying to find ways to compress video efficiently, and if it's flagged properly, the encoder can recognize 3-2 pulldown during compression and waste fewer bits on extra fields. It also (in theory) doesn't add additional compression artifacts due to the 3-2 pulldown if it knows how it was originally encoded, as the case of 24fps to 60 fields per second ala 3-2 pulldown. Some of the early pro-scan DVD players relied solely on these "flags" to convert interlaced video to progressive. The problem is, sometimes when some DVDs were encoded, these helper "flags" were set wrong. While this was only a minor detriment in encoding efficiency, it becomes a major snag when a pro-scan DVD player relies on these flags to make a progressive image. This is why most pro-scan players now have an "auto" mode, so it looks at the video material itself vs relying on the MPEG-2 identification flags.
Now some will say that even *if* 24fps material w/3-2 pulldown was authored as interlaced, because an external or internal DVD line doubler converts it to 480p properly, it's the same as a natively stored 480p format. That is simply not the case. When progressive images / video is converted from progressive (be it 1080p or 480p) to interlace, some scenes can show artifacts from "derezing" and need to be filtered properly.. by a human on a scene to scene basis, not by a chip or algorithm, although I'm sure that too is often done. Now let's face it... the vast majority of DVD playback it still interlaced so it makes perfect sense to continue to filter, monitor, and author DVDs in the interlace 480i format.
I could be wrong, but it wouldn't surprise me if this is why D-VHS is avoiding the 1080p format, because downrezing to interlace may introduce occasional artifacts to 1080i. But if the D-VHS title is converted to 1080i by the studio, these x-factors are eliminated. And die-hard home theater buffs can still do inverse telecine and achieve a true 1080p image, but without all of the vertical resolution of the original 1080p master. But if this *is* the case, perhaps they will introduce a cheap in-player filter code for 1080i playback, but then vertical resolution is chopped more often. This is ditto for 480p DVD transfers. (Hmm.. maybe a dynamic filter with an instruction code/channel could be used???)
OK boys! ..It's a slow day, so have at it!!