Personal View site logo
Make sure to join PV on Telegram or Facebook! Perfect to keep up with community on your smartphone.
Please, support PV!
It allows to keep PV going, with more focus towards AI, but keeping be one of the few truly independent places.
Official Low GOP topic
  • 1003 Replies sorted by
  • @robbin You need to register/login to get them... Al
  • @APXmusic
    thanks for Tipp - Media Player Classic runs - VLC does not
  • @kae
    i noticed that the last 4 frames are always (even the ones with normal footage) not ingested by 5dtorgb (it causes the freeze of transcoding) and most players have many difficulties to play the clips (for me it's impossible to get to the end). No issues or problems after transcoding.
  • Hi all. New here and wanted to try some of these firmware settings. Can someone explain to me how to download the linked zip files from some of your comments. My computer sees them as PNG picture files.
  • @svart

    Just tested the above settings with audio counting up to 39 on REGULAR subject matter, not the death chart and WMP played the entire clip back fine. Also the entire clip ingested into the avid. Maybe the end glitch is more pronounced with death chart shots because it's stressing the codec so much? Did you experience end glitch problems with normal footage, death chart footage or both?
  • @bwhitz, what film mode setting were you on? Smooth, nostalgic, cinema, etc? -2 on everything? Great footage.
  • Ok. Just tested my first settings that bwhitz posted on the death chart and media player does not play the last second or so as svart revealed. But it plays the clip fine, does not stutter. So until this is solved I will stick with my latest settings (attached here with the overall bitrates deleted) as the data rate on the death chart is 59mill and the visual quality is better. In WMP at least, the clip does not stutter on playback at all but the last second or so will not play.

    When I ingest these clips into the Avid (I don't have FCP) the clip transcodes fine and it seems the last second ingests. Can someone do a test shooting visual timecode or a clock to make sure? For now, I will continue shooting tests with these settings as the footage is incredible and the workaround for the head and tails glitch is simply to roll sooner and cut later. Clumsy, I know, but better than crappy 1080p footage.

    Stumped as to what to try next to fix this, Vitaliy????
    setbRev072411.zip
    439B
  • +1 with @svart. Really appreciate @kae & @butt. Yes low GOP seems nice. We gotta discover more patterns and hints for Vitaliy to squeeze more juice out of GH2. i.e. Those settings are trasitory. Better PTool release is the true fruition from testings.
  • Thanks for posting them so fast! I see what I did. These had the 2.2x high/low/top/bottom multiplier instead of the 2x and I capped 24H at 50 instead of 60 mill on my most recent settings. I'm going to test these now and see if I can eliminate the end glitch. Did you have any experience with the end glitch on your footage or any problems ingesting them into FCP?
  • @Kae Thanks! ...and here's the settings I used... great work on the patch!
    seti.zip
    445B
  • @bwhitz

    WOW. Stunning! And I know how much better the 3 GOP looks without the vimeo compression. You can really light! Add some slight grain and it's indistinguishable from film. Will you attach the settings you used, I lost track of what I posted. I'm trying to see if I can eliminate the end glitch svart just posted about.
  • Someone suggested I also post this in this thread as well as it applies to the 3GOP setting... this was done with the first settings Kae posted.



  • @vart
    What I'm saying is that it's not completely stable.

    Ahhh, I see what you mean, although my players don't choke at all. I knew about the blocky couple of frames up front but my players seem to stop playing it a little before the end. Easy workaround - just roll a little before you need to and cut a little after you need to - just like shooting film. Hopefully, we can figure out what's doing this and fix it. But I am the opposite of you. I'll take predictable quality and predictable instability (and that's what this is) any day over rock solid stability but less than optimum quality.

    This was my problem with all the GH1 testers. Everyone was shying away from 24P n because once and awhile it crashed. I was shooting 48 Mb/s 1080 24P N with 3 GOP for 7 months in real world scenarios and it only crashed a couple of times with some stuttering frames. The footage was stunning - better than any of Lpowel "stable" patches without 24N. No right or wrong here, just different priorities.

    Agreed though, would be great to get rid of the blockies and inability to play the last couple of seconds of the clip. I don't know what is 'choking' your players though. Processing power?

    I am going to test the settings again now without:
    Video Bitrate 24H ADD=70000000
    Overall Bitrate=70000000
    Overall Bitrate 2=54000000

    which Vitaliy proved redundant to see if that makes the clips more stable. Maybe this will make a difference.
  • What I'm saying is that it's not completely stable. Something about it chokes up all of my players and there is always 2 seconds of nothing at the end of the files.

    In any case, I'm testing for stability, not for best quality. Quality will come when we figure out how this works.

    I have to remind folks that we aren't striving for perfection yet. We need folks to just try a bunch of stuff, preferably on VK's testing lists so that he can understand how the changes make the codec work differently, not to try to hurry and get to the good stuff.
  • @svart Media Player Classic seems to play all clips from a modified GH2 just fine.
  • @bwhitz
    i think no - unfortunately!

    maybe create Vitaly a new PTool . . . MJPeg 30p<>24p
  • @Butt Is it possible to change the Mjpeg framerates yet? Obviously to 24p... ;)
  • my favorite is this 1080/24pN GOP3 patch (in camera playback ok)
    including 720p GOP2 Patch (in camera playback ok)
    including 105MBit MJPeg 1080/30p (no in camera playback possible)
    sete.zip
    586B
  • @Angry_C
    for the lowGOP Patch use Panasonic Class 10 60Mbp/s
  • @svart

    try the VLC player version 1.1.11for GOP3 Files
  • Perhaps now that more detail is encoded moiré is more obvious...
  • Running Kae's settings:

    Recorded 20 seconds

    First 3 seconds strong macroblocking
    VLC player chokes 6 seconds into file
    WMP chokes at end of file
    Splash player plays file but video stops 2 seconds from end

    The file is 20 seconds long but 2 seconds at the end is missing.

  • @BenjaminDUPLAIX

    The moire has nothing to do with the hack or the settings. It's a limitation of the chip and the way it processes detail. It's been there on the GH1 and is on GH2.
  • @BenjaminDUPLAIX

    why should the higher number of high-quality I frames lead to more moire?
  • 44mbps and gop 3 looks good. But I can see moire I think
This topic is closed.
← All Discussions