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 GH2 "Stalin" hack development topic
  • 939 Replies sorted by
  • Fantastic!
  • Again, nice work!
  • Congratulations, that looks amazing. May I ask how is the bit rate for 1080/24 in low light setting?
  • This charts are for low light medium detailed scene.
  • @Vitaliy

    So what is your opinion so far? Does the GH2 benefit more from the hack than the GH1 does?
  • >Does the GH2 benefit more from the hack than the GH1 does?

    GH2 hack will be far ahead from GH1, of course.
  • Yoooohooo I am so relieved, I have been a bit wary with the failure of the first batch of test with the 24p output that Panasonic had had left some bad surprise in case the gh2 was hacked. That will be really incredible, great job Vitaliy.
  • @Vitaliy

    I wouldn't expect the GH2 to be that much better in terms of overall image since 40 mb/sec and above on the GH1 was pretty close to perfect anyway. Are you really seeing that much improvement on the GH2 for the 1080p @ 24 FPS mode?
  • @mpgxsvcd Remember GH1 had no b-frames, 40Mbit+ was necessary just to get rid of the compression mud. GH2 has better scaler, better compression efficiency, better sensor... 40Mbit will be great in AVCHD on the GH2
  • "I wouldn't expect the GH2 to be that much better in terms of overall image since 40 mb/sec and above on the GH1 was pretty close to perfect anyway."

    Yes, but I thought the GH2 produces a bit better picture, so with a higher bit rate I could only imagine the GH2 should be even better!
  • >Are you really seeing that much improvement on the GH2 for the 1080p @ 24 FPS mode?

    It is be better in high detailed scenes and if you use high ISO, os it preserves noise accurately.
    GH2 42Mp is much better, not because of B frames (so multiply by 1.3x GH1 bitrate), but also because of better motion estimation and better codec bitrate correction (so, it won't drop that much as in GH1).
    Plus you must remember much better SD controller, so people get 160-180Mb/s MJPEGs with fast cards.
  • >Plus you must remember much better SD controller, so people get 160-180Mb/s MJPEGs with fast cards.

    That is fantastic news. Is the MJPEG still hardcoded to 30P? Or any possibility to change that to 24P? That high bit-rate would be great for green-screen.
  • @Vitaliy

    Thanks for the explanation. Now I really want the next Ptool release. Any idea how long we will have to wait. I have some time tonight and tomorrow and I would love to do some tripod testing with each individual setting.
  • @Isaac_B
    Yes, 30p. We need big time to understand how sensor, scaler and compresser are working together.

    Btw, 422 patch do not work, so not so much use for big bitrates. Again, we need to look at whole MJPEG encoder to find change performed by Panasonic.
  • >Any idea how long we will have to wait.

    As I said, it depends not on me only.
    New PTool will be released only after we'll move forum and domain to VPS.
    Otherwise it'll be extremely sluggish thing :-)
  • VPS? Is it a hoster? Why don`t you release 3.61 independent of server transfer on here on this site?
  • Probably still an overload of visitors. That said, amazon S3 now supports domains. Upload a simple html website for it and the updated PTool there and you won't have to worry about too many visitors ever again.
  • I know it's not a priority for some, but any word on 1080/25p? Anxious pal users have fingers crossed..
  • @Vitaliy_Kiselev

    How is gradation, is there the same banding over smooth tones like in a blue sky we have with all 8bit AVC H.264 codec DSLRs or is that fixed by the higher bitrates?
  • @Vitaliy_Kiselev "Again, we need to look at whole MJPEG encoder to find change performed by Panasonic."

    MJPEG mode is disabled when HDMI monitor connected, that might give a clue... seems HDMI and AVCHD encoder chip closely linked, and MJPEG is separate feed. Or maybe they just did this for standards compliance reasons.
  • >Btw, 422 patch do not work, so not so much use for big bitrates. Again, we need to look at whole MJPEG encoder to find change performed by Panasonic.

    That'd be quite interesting, seeing the hardware-possibilities we have there...

    Something else: HDMI-related questions are not only important because of external recording - for me, universal EX-tele availability is more relevant. So you may not need HDMI recording with ± 50 Mbit AVCHD; but I'd love to use EX-Tele mode, working with an external monitor, exactly this combination is really interesting for many purposes (like recording concerts in low light, where you want really fast telelenses, and a big screen, because it's hard to see in semi-darc scene anyway)
  • @EOSHD Just to re-roast that old chestnut, isn't fixing the banding on smooth tones (because of 8-bit depth) a function of dithering, not bitrate? I've done some experiments where the banding was much less on high-iso shots (because of sensor noise acting as dither, I believe).

    However, I'm all for high bitrate if it gives a good quality increase in an efficient way...
  • Thats what I thought @Mark_the_Harp (due to 8bit) Just wondered if issue can be mitigated slightly with better compression. We shall see!
  • Exciting times ahead!