1080i vs 1080p - Page 6

108024PsF is not an encoding standard. It is purely a transport standard that says that each field is trasmitted separately over the wire by being clocked out of a progressive frame buffer with the odd lines comprising the first field and the even lines comprising the second field.

This is no different from sending a 1080I signal down the wire, except that each pair of fields are guaranteed to be from the same frame, originally captured with no temporal (that means time, it nothing to do with tearing) displacement. (IOW, the entire frame buffer was originally captured from the same film frame).


I don't agree with this. 24PsF was created to allow progressive frames to be processed by the same cabling and switching infrastructure (HD-SDI, for example)already in place in HD production facilities. Without it we probably wouldn't have 24P encoded on HD discs to begin with, because every telecine bay and production house would have had to rip out every existing piece of gear and cabling and replace it with new 60I / 24P capable gear at a very great expense.

Vern

Thanks Ron. I was being really sloppy (and incorrect) using the term 1080p24sf in a compression context. I was trying to distinguish between interlaced and progressive sequences while avoiding the confusion that terminology usually produces.

But now that I got your attention ... Is it true that Blu-ray MPEG-2 encoding is still using interlaced sequences of frame pictures for film-sources? Is that still the only way to include top_field_first and repeat_first_field flags in MPEG-2, or has that changed? Is VC-1 similar in that respect?

Edit: Also thanks Vern and Stacey, I didn't see your replies until after I posted this.

If all interlacing was against the law for the wole universe this would be a better world.

Actually, it's the other way around. HD-DVD only allows 1080i@29.97 fps sequences while Blu-ray allows 1080p@23.976, 1080p@24 and 1080i@29.97 fps sequences (independent of codecs).

The current scheme for VC-1 encoding on HD-DVD is to code the source at 1080p@23.976 and then run the bitstream through a tool that changes the syntax of the bitstream from 1080p@23.976 to 1080i@29.97 with perfect telecine flags (repeat_first_field, top_field_first and progressive_sequence in MPEG-2 lingo but probably some other names in VC-1 and H.264). Stacey can verify that this is the case.

If you are interested in direct 1080p23.976 or 1080p24 output from the player, then Blu-ray has the advantage. The Blu-ray player knows that the stream is 24p at the beginning of the movie and will not change for the entire movie. It can output 24p without concern. There's still some question as to how the player handles other content on the disc and whether annoying output frame rate changes might occur. We'll see when an actual Blu-ray player ships with 24p output.

HD-DVD players cannot know that the stream is 24p. The stream is always signalled 1080i@29.97. It may have perfect telecine flags (and the VC-1 titles do because of the tool mentioned above), but the player cannot know beforehand that the entire stream has perfect flags. It could be 1080i@29.97 video, 1080i@29.97 with telecine flags dropping to video at times or 1080i@29.97 with perfect flags.

So for HD-DVD, it's a user interface problem. You could provide a 24p output mode, but it's not guaranteed to work properly with all content. Most designers are reluctant to provide an option that possibly doesn't work. It just leads to confusion and lot's of support calls from the field. But never say never.

Bottom line, the 24p output story is yet to be told by either format. I expect Blu-ray to deliver first (and possibly HD-DVD never), but we'll see.

Ron

The statement is correct. Your understanding of it is not complete.

Ron

OK, I got it now. Your key statement is that the player detecting and outputting 24FPS from an HD DVD is not bullet proof.
I agree with that.

Since every HD DVD I have played through my CII scaler and Sony Qualia has exibited perfectly smooth pans and scrolls with no sign of 3:2 induced studder, I'm thinking it's not a big deal if we don't get 24PsF or 24P out of the player anyhow.

Deleted previous post as it clearly doesn't apply here.

Vern

Here a web site guys that compared 1080i vs. 1080p. Very interesting.
http://blog.hometheatermag.com/geoff...061080iv1080p/

By the way if the new blue laser content (HD-DVD and Blu-Ray) disc is store at 1080p/24 then what is DVD? 480p/60?

I've got a CII and Sony Ruby, and at 1080p60 I'm seeing plenty of stuttering pans. Going in via DVI at 1080p48 fixed that, luckily.

Well, I am going into my Qualia with 60I since it doesn't support 60P without an upgrade. The Qualia also has film pulldown detection logic.

Vern

UGH....Always a step a head you guys!!!

that's not very kind and supportive!!!

i guess time will tell huh?

So let me ask this--say Blue-Ray conmes out with a player that outputs 1080p/24 and say you send that to a set that accepts 1080p/24 and has a REFRESH of 72.

In THAT case and IF the Blue Ray Player did a GREAT job of ouputting 1080p/24 and SINCE the set would REFRESH at 72 instead of 60 wouldn't you have the best picture possible as far as a film would go?

And since video is shot Not at 24 like film but at 30 or 60 why don't they just make PROGRESSIVE sets that ACCEPT 1080p/24 and REFRESH EVERYTHING Video and Film at 120 so there never is freakin Judder!

What's so hard about having Blue-Ray output 1080p/24--having sets that accept it and having everything refreshed to 120 so everything in the video universe--film and video would look great?!

Artwood -

I'm pretty it would look better because of the 3:3 pulldown. The real diiference is if we ever get 1080p/60 content on disc which probably won't happen anytime soon. 1080p is marketing guys.

Hieu

Why is it that the Japanese can broadcast 1080p but we can't? Are we just retarded?

What if they had only 1 channel of 1080p and charged anybody who wanted to put prgraming on it? Who would PAY to use it?

If someone recorded it would the universe end?

A 1080p feed was setup a couple of years ago at CES. A group went around every booth who offered an ASTC tuner and asked them to tune to the channel. Less than 10% of the ATSC tuners were able to decode it. It was decided at this time that 1080p was not an option for ATSC in the US due to lack of support from the box makers.

Did the box makers take their marching orders from the MPAA?

I don't mean to be a spelling/grammar nazi, but the term you're looking for is "all intents and purposes", not "intensive purposes."

But you just became one by posting

I hear people all the time say irregardless, and supposebly, but I stay mum even though these "words" do not exist. So for all intensive purposes, I am considerate.

-T

You are considerate, but can you tell us the heighth of the new blu ray palyers

This cannot hold for games correct? A 60fps 1080P game cannot be transmitted in 1080i and displayed in 1080P without a loss in IQ or a drop to 30fps or am I missing something?

Hi guys,

I'm hoping Ron (or someone) can help me understand this better. Let's use an example of an HD-DVD player outputing 1080i HDMI to a 1080p display. I understand that a movie will be encoded at 1080p/24. Are you saying that inside the player, this must be decoded at 1080i/29.97? If so, how is this accomplished? I undersand 2:3 pulldown, but can't make the connection here.

And, if the player has a 1080i/29.97 signal, does it simply double each field to ouput the final 1080i/60?

Finally, what is the frame pattern of 1080i/60 that is reconstructed to 1080p/60?

Who and what exactly is being broadcast in 1080p?

Are all of us suckers who just wanted to get into HDTV watching obsolete sets just 1-year into the purchase?

Much of the series / sitcom programming on CBS and NBC is a 1080p24 sourced and delivered at 1080i60. If your TV does a decent job at deinterlace, it can reconstruct the original 1080p image.

The bitstream describes a 1080i@29.97 output from 1080p@23.976 frames. Here's how it's done:
TFF alternates after every time RFF is 1. That is, if you repeat the first field (top or bottom), the next frame must start on the opposite parity (bottom or top).

I use the term 1080i@29.97 to mean 29.97 interlaced frames per second. That's how it's signalled in the bitstream. Everything is normalized to frames. However, lots of folks call this 1080i59.94 (or 1080i60 for short), or 59.94 interlaced fields per second.

If the output rate after a deinterlacer is 1080p@59.94, then the repeated fields are tossed, and the original 1080p@23.976 frames are repeated either 2 or 3 (in a 2:3 cadence) times to get to 1080p@59.94.

Ron

It's somehow contradictory.
Blu-ray has to deal with 1080i@29.97 fps sequences as well.

Thank you for the explaination, Ron.

MOst LCD/Plasma TVs today take the 1080 60i signal, and use only half of that information because it is unable to deinterlace properly. Every other frame is skipped, and your resolution essentially becomes 1920x540 @ 30p. That means that about 50% of the transmitted information is used.

The TV uses that converted 540p signal and scales it to the native display resolution.

If the TV has a 1366x768 resolution, even more pixels are lost because it cannot display the 1920 horizontal pixels that are available from the converted 540p signal. The TV has to scale the picture down to 1366 horizaontal pixels, which is about 71% of the 1920x540 pixels.

I may be worng with these calculations, but it seems to me that most HDTV screens sold today use about 35% of the pixel information sent by the broadcasters.

It is still better than standard TV though.

Pleae correct me if I made some incorrect assumptions.

Okay now. We don't want anyone to loose there heads over this apparently sensitive issue.

rambo, I believe this information was based on the previous generation of TV sets. In the newest generation, I believe that there are far fewer sets that do this. More sets today properly handle deinterlacing 1080I. The sets that still do not are typically not 1080 resolution sets, so some loss of resolution is guaranteed, regardless of how the deinterlacing is performed.

Vern