1. Join Now

    AVForums.com uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Progressive scan clarifications

Discussion in 'Projectors, Screens & Video Processors' started by SimonMW, Jul 17, 2005.

  1. SimonMW

    SimonMW
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Hi,

    I've been reading a few comments about progressive scan and deinterlacing I'd like to try and clarify, as well as trying to find out some more technical information so that I can see if there are any differences between DVD players and televisions. Some people appear to be confusing deinterlacing with true progressive scan and are using the terms interchangably.

    As I understand it, from my own work in video (I'm a professional cameraman and editor as my day job) a DVD can be encoded as either interlaced or as progressive scan. Now lets take some of my own footage as an example.

    I usually shoot in true 25 fps progressive scan for many projects I produce. This means that the camera is recording the full 576 lines. When I edit, because the signal is recorded to an interlaced format (PAL), but with both fields carrying information from the same frame, I need to tell my software I am using progressive footage. Then, when I encode the final DVD I need to specify to the MPEG2 encoder that I am using progressive scan footage. This way I keep the absolute full resolution of the progressive scan picture. Many commercial film DVDs are also encoded in true progressive scan.

    This means that when you use these DVD's on a progressive scan DVD player they simply output the full progressive signal through component for example. There should be NO deinterlacing taking place whatsoever, contrary to what has been said in other threads.

    Why have I mentioned this? De-interlacing usually involves ditching half the resolution of an interlaced picture by taking the dominant field and doubling it. Some very clever deinterlacing systems will look at the motion of the picture and only deinterlace those aspects which are moving. This still removes resolution but is slightly better in some cases.

    Regardless of the method deinterlacing removes resolution. On some of the other threads people have said that whether you buy a progressive DVD player or not depends on whether your TV does a better deinterlacing job than the player. I find this highly odd, although I am open for correction if anybody knows any better. In order for your TV to display the true progressive scan picture from a progressive scan DVD it would need to read the odd and even fields and combine them into one frame (because 25 frame progressive displayed in a 50i signal the odd and even fields will be the same).

    However I doubt most televisions deinterlacing circuits work this way because many TV's these days convert broadcast signals to progressive scan as well. The combining of fields into one frame will not work for a true interlaced source. Line doubling or interpolating will however, but this loses resolution as I mentioned.

    The long and short of it is that IMO the only way you will get the true full progressive scan picture off a progressive scan DVD is to use a progressive scan DVD player through component into a TV that can accept and display the incoming images progressively with no field manipulation. This is because there is NO deinterlacing involved as the DVD is encoded in progressive form to begin with.

    Some people have mentioned jagged edges on progressive scan pictures. This could be down to the fact that you are seeing the picture resolution in one go rather than with the interlace flicker to cover it up. 576 lines is not an awful lot. The other problem could be down to edge enhancement added either during the DVD encoding process, or by the DVD player, or by the TV, or a combination of all three. This would have the effect of increasing aliasing on high frequency edges.

    So to sum up, the reason I posted this is because true progressive scan and deinterlacing are two completely different things that seem to becoming muddled up on some of the other threads in this forum.
     
  2. Gordon @ Convergent AV

    Gordon @ Convergent AV
    Distinguished Member AVForums Sponsor

    Joined:
    Jul 13, 2000
    Messages:
    14,011
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Location:
    Living in Surrey, covering UK!
    Ratings:
    +2,808
    Just want to add some further information you may find useful . DVD players with progressive scan ouptuts have to de-interlace the interlaced output of the mpeg decoder, even when the decoder is outputting progressive as well. They can do this in digital domain by varying methods. Often it's by looking at flags on disc to see whether the source was originally recorded progressively (like yours) or not. If it was they will reconstruct the progressive frame and output that accordingly....however the flag information is often wrong or missing. The best progressive scan DVD players actually use dedicated chipsets to look at the picture information to ascertain how best to de-interlace to progressive format. These are fed an interlaced digital signal off the MPEG decoder in the player.

    You will find that most progressive scan dvd players, tv's displays and video processors will have 2:3 inverse telecine detection to allow them to accurately reconstruct film frames from 60Hz video sources. They may or may not have 2:2 inverse telecine for 50Hz progressive sources. As you say the better de-interlacers will do motion adaptive work on material recorded with a video camera(interlaced). It should also be noted that the above is for SD material. For HD material it's virtually all just BOB de-interlacing at present with the resultant reduction in resolution from progressive capture sources. There are new generations of products coming that will do all these clever tricks on HD too.

    Hope this is of some use to everyone.

    Gordon
     
  3. SimonMW

    SimonMW
    Guest

    Products Owned:
    0
    Products Wanted:
    0
    Ratings:
    +0
    Thanks Gordon. I wasn't aware of some of the problems of missing or incorrect flags and the potential problems this could cause. I can see this will drive me insane finding DVDs that were not encoded properly! I just found this article which people may or may not have seen regarding progressive scan DVD player problems.

    http://www.hometheaterhifi.com/volume_7_4/dvd-benchmark-part-5-progressive-10-2000.html

    Although at least we know that when a DVD player does recombine the fields properly into progressive scan frames it does it pretty well. The problem with TV's is that we have no reference as to what kind of method it is capable of to create the progressive picture.
     
  4. Gordon @ Convergent AV

    Gordon @ Convergent AV
    Distinguished Member AVForums Sponsor

    Joined:
    Jul 13, 2000
    Messages:
    14,011
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Location:
    Living in Surrey, covering UK!
    Ratings:
    +2,808
    Simon: That article is often referenced on this forum......

    There are other issues as well with regard to DVD. Bad edits can cause different cadences and different programme material can have non 2:2 or 2:3 cadence. The chipsets used to do film detection (inverse telecine) all have their pro's and cons in how well they lock on and what happens when they loose lock. The proof of the pudding is in the eating....as you say the specs don't give you enough info.

    Gordon
     
  5. Welwynnick

    Welwynnick
    Well-known Member

    Joined:
    Mar 16, 2005
    Messages:
    7,087
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    133
    Location:
    Welwyn, Herts
    Ratings:
    +825
    Well, this is big news for me.

    I don't consider myself ignorant, but I thought DVDs were interlaced only, and all "progressive" players used de-interlacers. It never, ever, occured to me that some players could simply pick whole frames off the disc.

    What happens to a progressive disc being played on a non-progressive player? Does it INTERLACE the video to output it? That sounds unlikely.

    It's difficult enough to find DTS discs, but how do we know what discs are progressive. I often read the small print on the back, but I have never seen it say progressive. That would have caught my attention.

    Nick
     
  6. Gordon @ Convergent AV

    Gordon @ Convergent AV
    Distinguished Member AVForums Sponsor

    Joined:
    Jul 13, 2000
    Messages:
    14,011
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    166
    Location:
    Living in Surrey, covering UK!
    Ratings:
    +2,808
    Nick.

    All progressive players do use de-interlacers. That was the point of my post. The original poster was wrong in his assumptions.

    Gordon
     
  7. Carl Ed

    Carl Ed
    Standard Member

    Joined:
    Mar 9, 2005
    Messages:
    54
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    6
    Ratings:
    +0
    All material is stored on DVDs as fields, not frames or segmented frames. This is due to a combination of backwards compatibility and the age of the DVD spec.

    Hopefully BD and HD-DVD will use 1080/24p (or 1080/25p) and not have the issues that DVD has.
     
  8. kurtz

    kurtz
    Active Member

    Joined:
    Feb 15, 2004
    Messages:
    171
    Products Owned:
    0
    Products Wanted:
    0
    Trophy Points:
    21
    Location:
    last place you'd look
    Ratings:
    +21
    I thought segmented frames are stored as fields, where one frame = 2 fields. It sounds like Simon is describing segmented frames:
    I thought that this was 576sf25 segmented frame format where a frame was split into two fields - but with the important proviso that both fields contained lines from the same progressive source frame. Therefore the fields can be easily recombined, so only trivial deinterlacing is required (2-2 pulldown).

    The DVD studios' telecine machines output pairs of fields based on the same frame too. So we can effectively have 576sf25 dvds today.

    I have stayed away from the 480 line 3-2 interlaced case because that clearly isn't segmented frame - but in PAL-land we must be very close to it.

    This is stretching my understanding of segmented frames and 2-2 cadence to the limit :rolleyes: - is there a killer exception case out there I don't know about? Is this just a case of standards and nomenclature preventing the disc content being officially called 576sf25 because it can't guarantee that the flags will be correct?
     

Share This Page

Loading...