Chroma bug only in progressive scan?

Discussion in 'Playback Devices' started by Sami Kallio, Apr 13, 2004.

  1. Sami Kallio

    Sami Kallio Screenwriter

    Joined:
    Jan 6, 2004
    Messages:
    1,035
    Likes Received:
    0
    Trophy Points:
    0
    After reading all these articles about chroma bug, I got the understanding that this only affects progressive scan, not interlaced. Am I wrong?

    http://www.hometheaterhifi.com/volum...ug-4-2001.html

    "One should be used for frames marked "interlaced", and a different one should be used for frames marked "progressive." Most decoders use only one algorithm for all frames, and that's where they go wrong."

    I have an interlaced picture (large, over 100") where some of the red colours flow over and a friend told me that looks exactly like a chroma bug. I was hoping that would be the case as I have heard Sony players tend to have this bug but from all I read it seems to be progressive only problem.
     
  2. Allan Jayne

    Allan Jayne Cinematographer

    Joined:
    Nov 1, 1998
    Messages:
    2,404
    Likes Received:
    0
    Trophy Points:
    0
    The "chroma bug" may be present during interlaced scan viewing also. It is less obtrusive during interlaced display because the odd scan lines start to fade while the even scan lines are being drawn and vice versa.

    You will almost always see closely and uniformly spaced thin horizontal dark streaks throughout the picture from time to time when viewing a large interlaced picture, this is the interlace fading I described, not the chroma bug.

    Broad flow over of color is not the bug, rather the bug is thin one scan line thick discolored horizontal streaks. You will get minute flow over, called blooming, if your contrast is set too high. Broad flowover is also seen if the source material was just NTSC broadcast quality at some time in its earlier life, or if you use composite or (not as much with) S-video connections.

    1. If the DVD player shows the chroma bug during progressive scan, the player will also show the bug when you set it to interlaced mode.

    2. Some progressive scan players correct the chroma bug (actually use a scan line blending procedure that softens the color boundaries slightly) in the de-interlacing stage in which case the chroma bug will still be there in interlaced mode.

    3. It is possible for the MPEG decoder (the first stage in the player) to yield video without the chroma bug where interlaced output is also free of the bug.

    Video hints:
    http://members.aol.com/ajaynejr/vidbug2.htm
     
  3. Sami Kallio

    Sami Kallio Screenwriter

    Joined:
    Jan 6, 2004
    Messages:
    1,035
    Likes Received:
    0
    Trophy Points:
    0
    Thanks Allan! I guess I will have to find myself a bug free DVD player and test it out. Too bad the Denon's aren't easily made region free, otherwise I would be buying one of the refurb 910's form ecost.
     
  4. Chris PC

    Chris PC Producer

    Joined:
    May 12, 2001
    Messages:
    3,975
    Likes Received:
    1
    Trophy Points:
    0
    Ok, so the MPEG decoder can cause the Chroma Bug and so can the de-interlacer? If the MPEG decoder is free of the bug then the interlaced and progressive output is free of the bug? I think the Interlaced Chroma Problem (ICP) is talking about interlaced video material (as opposed to film flagged DVD's) and not interlaced output on a player? I'm a little curious if the Panasonic DVD F85 and F87 have much of this bug.
     
  5. greg_t

    greg_t Screenwriter

    Joined:
    Jan 18, 2001
    Messages:
    1,650
    Likes Received:
    0
    Trophy Points:
    110
    The chroma bug is caused by a fault in MPEG decoders, not deinterlacers. A MPEG decoder that has the chroma bug will display it in both interlaced and progressive. Some deinterlacers, however, can mask the chroma bug, basically hiding it. The Faroudja chip has the ability to mask the chroma bug. Here is a link to give you more info than you ever wanted to know about the chroma bug.

    http://128.121.62.219/volume_8_2/dvd...ug-4-2001.html
     
  6. Jim_F

    Jim_F Screenwriter

    Joined:
    May 15, 2000
    Messages:
    1,077
    Likes Received:
    0
    Trophy Points:
    0
    I see it much more often on my cable than I ever see it on my Sony 9000ES (where I see it rarely and usually only during the credits on a DVD)

    It doesn't bother me too much either way.
     
  7. Dale Adams

    Dale Adams Extra

    Joined:
    Jun 16, 1999
    Messages:
    16
    Likes Received:
    0
    Trophy Points:
    0
    There's at least a bit of misinformation being posted here. CUE and ICP, while related in terms of the net visual effect and the fundamental cause (interlacing), are different beasts.

    CUE is introduced in the MPEG decoder when video material encoded as (effectively) progressive is decoded as interlaced. ICP is a inherent limitation of all interlaced material subsampled using a 4:2:0 scheme (i.e., pretty much all DVDs). It is possible to have an MPEG decoder and DVD player with CUE. They all have ICP, however, and this bears no relation to whether or not there's a CUE problem.

    CUE and ICP artifacts can be largely filtered out, with a small loss of vertical chroma resolution. It's always better to have no CUE, however, as there is some image quality degradation when CUE artifacts are generated and then filtered. ICP is always there, so the best you can do is try to filter out the artifacts.

    - Dale Adams
     
  8. Chris PC

    Chris PC Producer

    Joined:
    May 12, 2001
    Messages:
    3,975
    Likes Received:
    1
    Trophy Points:
    0


    OK I think I get it. Material on DVD's is encoded as interlaced, with many movies using progressive film flags when the material originates from a 24 fps progressive movie. So you're saying when the MPEG decoder goes to work making the interlaced output, regardless of it being flagged as film, the MPEG decoder makes fields with the CUE in them? Some MPEG decoders, for instance, the older Panasonic ones, DO NOT create the CUE? I wonder if the deinterlacing is critical. For instance, Faroudja and SiI50x chips are ok with poor flagging, but do bad film flags or generally bad de-interlacing affect any CUE that may be present?

    So if the MPEG decoder creates CUE, some de-interlacers like the Faroudja (which masks it because of its method of sampling chroma from only certain fields) or whatever it is that the SiI504 chip does to filter out the bug.

    Either way, its confusing.

    So how much difference is there between:

    1) Using 480i output which contains the CUE when you use a Faroudja or SiI504 to remove the bug?
    2) Using 480i output without any CUE to begin with.

    My questions relate to my thread concering the Panasonic DVD F85 and F87.

    Thanx for the helpful info.

    [​IMG]
     
  9. Dale Adams

    Dale Adams Extra

    Joined:
    Jun 16, 1999
    Messages:
    16
    Likes Received:
    0
    Trophy Points:
    0

    The 504 does not remove CUE artifacts. The Faroudja deinterlacers do, but only incidentally as they only use a single field of chroma even for film sources. While this has the (unintended) benefit of largely removing CUE artifacts, it also introduces other problems, such as flicker with sharp vertical chroma transitions.

    It's always better not to have CUE artifacts to start with. Filtering out the artifacts does not yield the same quality signal as an unfiltered signal with no CUE artifacts to start with. The difference may be subtle, but it's there.

    - Dale Adams
     

Share This Page