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.
Truthfulness of short GOP superiority claims
  • dkitsovdkitsov
    I am sure that this topic will start a few flame wars, but I am after an extensive testing not so sure about the advantages of the short GOP schemes.
    I fail to see any difference between short and long GOP recordings. I have made a test recording using three different settings: @kae 3GOP setting (as of July 22), same but with 6GOP, same but with 12GOP. All of them were of the same scene with a repeatable motion, same camera settings etc. I fail to see any difference as far as temporal resolution and accuracy are concerned. I have encoded three of the recordings with an intra-frame codec (cineform i frame only) to remove original h264 GOP, edited them down to a 5 seconds clips, run them through a denoiser (to remove a noise signature that might aid in the ID-ing the GOP), and finally encoded them at 50mbps and uploaded to Vimeo.
    Please look at these clips (A, B, C) and comment below about which one out of A,B,C do you think is 3,6, 12 GOP. Originals (not re-encoded by Vimeo) are available for download on Vimeo
    Thanks.
    A.

    B

    C

    http://vimeo.com/thedima/shortgopa
    http://vimeo.com/thedima/shortgopb
    http://vimeo.com/thedima/shortgopc
  • 100 Replies sorted by
  • svartsvart
    That's what I mentioned in the other thread, I've been peeping these files pretty hard and I pretty much can't tell the difference, even in the files that people swear they see a difference.

    I tried the 3gop stuff today and then immediately put the stock firmware back on the camera and took a vid of the same trees/grass in the wind and split screened them. I couldn't tell a difference. Even the MUD was very similar.
  • bwhitzbwhitz
    It doesn't show up on all shots... I notice it on people and hand-held shooting more than anything else. But hey, if it doesn't make a difference to you, then it doesn't really matter...
  • brianlucebrianluce
    Does the shorter GOP grade more easily?
  • ButtButt
    @bwhitz
    yes I see the same way
  • dkitsovdkitsov
    So far four responses and yet not a single attempt to ID the GOP. I wonder if it is because there is no difference...
    @brianluce GOP is not important for grading, bitrate is. At too low of a bitrate, when you are trying, let's say, to grade something up from the shadows, instead of detail you get macroblocking. With enough bitrate you will get information, albeit noisy, but the information non the less.
  • qwerty123qwerty123
    I'm on a bad computer to look at the settings, but I think one IMPORTANT thing about testing the differences is this --> if you want to test whether there is any difference, you need to test under circumstances in which you would MOST EXPECT to see a difference if there is any.

    I'm unsure whether your short clips satisfy that condition.

    It's not a criticism--it's just something that needs to be thought about in getting a good test.
  • DanDOFDanDOF
    Well, I'll give it a try: I guess B is GOP 3 ( http://vimeo.com/thedima/shortgopb )
    Am I right?
  • dkitsovdkitsov
    @qwerty123 These clips cover the entire range of the object motion with an overlap. This motion range covers multiple Groups Of Pictures: 10 in case of 12GOP, 20 in case of 6GOP, 40 in case of 3GOP. Should be more than enough to compare. Any temporal quality difference in GOP of different length would be obvios in a clip as short as the length of 2 or even 1 GOP that is longest, ie: 0.5 a second or a second long.
  • dkitsovdkitsov
    @DanDOF
    I will not answer this question just yet. I want to get more responses to have a significant sample for a study.
  • BuellerBueller
    There is no difference. I think they are all the same clip. lol
  • bwhitzbwhitz
    @dkitsov Again, locked off shots really don't explain what we're talking about. It's the more subtle movements that long-GOP footage seems to effect... but it's all subjective. This whole thing only started because a few of us noticed that the 3-GOP setting looks more like Red, Alexa, Film motion. We're not making it up for no reason...
  • dkitsovdkitsov
    @bwhits Sure, the only reason you say that locked off shots do not exhibit the phenomena is because when I will run the hand-held test of the same scene you will say that one cannot compare the clips because of the variable of motion of the camera operator.
    Still, as of now I accept your surrender, I guess you are too proud to admit that you cannot see a difference.
  • ButtButt
    the perception is subjective
    who prefers the GOP3, why not?
    I prefer GOP3
    the spaghetti - I'm not saying what is GOP3, why?
    do a Lowlight comparison with people in motion . . .

  • bwhitzbwhitz
    @dkitsov No, I'm just saying it because lock off shots don't show a difference. Especially without people and organic motion. I can't tell a difference here really... it's fast moving spaghetti against and a still background. Low-GOP motion is all about how the footage moves in more subtle conditions.

    Also, I don't know why this is being made into a battle. Why does anyone have to win or surrender? Just because some of us perceive and prefer short-GOP motion doesn't mean that everything else it bad... just use what you like and what is best suited for the project. Everyone is making this into more than it is. This isn't a contest. Hell, I'll probably use the longer-GOP settings for some stuff as well eventually. It's just not that big of a deal.
  • driftwooddriftwood
    C = 3GOP, A=6 GOP, B= 12 GOP
  • dkitsovdkitsov
    @driftwood "3GOP wins - more detail"
    It's an opinion. Facts please.
  • dkitsovdkitsov
    By the way I am neither pro or against short GOP. I want to know the truth. I do not care about the opinions. I just wonder if short GOP is a Monster Cable of AVCHD on GH2.
  • Ian_TIan_T
    @dkitsov - nothing wrong with having an opinion. If one perceives something to be different they shouldn't have to prove it? Perception is reality. What you think is real might not be to the next person. If you say you don't see a difference....I believe you. But if I say I see a difference...I would hope you believe me. It's just one of those things that you can't quite put your fingers on.

    Honestly though...I can't tell the difference between the three examples you've posted. Then again...I have not downloaded yet.
  • bleach551bleach551
    I say A=3GOP, B=12GOP and C=6GOP.
  • VMTVMT
    Is it a superiority claim? It's more of a preference claim. Its the same as preferring the "film" look over the "video" look. It's subjective and one is not superior to the other.
  • libertaslibertas
    a=6, b=3, c=12
  • fehafeha
    @dkitsov
    I just wonder if there is enough motion and details to trigger (challenge) codec to use GOP ?!
  • dkitsovdkitsov
    @feha fair point, will construct a more strenuous test.
  • Angry_CAngry_C
    a=6
    b=3
    c=12

    Gop 3 is more smooth for my eyes.

  • DazzaDazza
    I pick the same as libertas and Angry_C. I haven't really tested what GOP I like better, but I do like them in that order B, C then A.
Add a Comment