Personal View site logo
Make sure to join PV on Telegram or Facebook! Perfect to keep up with community on your smartphone.
GH2 Hack Myths and Realities
  • I have seen a lot of really good testing going on here. Some of the interpretations of results, however, seem a little off to me - or at least a little simplistic. Although testing things like high bitrates and low GOP are clearly valuable, interpreting results based on maximizing just one thing doesn’t effectively address the issue of how each of these affect a usable final outcome. No one thing will produce the results we want, and some of the things that need to be considered seem to never even come up. Some of the issues I see:

    High bitrate means good results. Not necessarily. Once the bitrate has been set high enough to outperform other factors in the camera (e.g. the sensor), there is no need to go farther.

    High AQ settings produce noise. Not true – high AQ simply reproduces noise more faithfully.

    Setting noise reduction in the camera is bad. Not necessarily true. If you are shooting at high ISO’s more noise will be produced. The codec has to work hard reproducing this noise, and you’ll end up just crushing the blacks or doing noise reduction in post anyway. Applying some noise reduction in camera under these circumstances can be very helpful because the NR is applied before video is encoded and you are not consuming codec bandwidth reproducing noise.

    The camera should always be set with Contrast at -2. Not true, the camera’s contrast should be set so the histogram shows the entire range being used. If you always set Contrast to -2 low contrast scenes will have poor gradations because you have set the dynamic range artificially low thus limiting the number of bits being used for the image.

    In camera saturation should always be set to -2. Not true, for the same reason given above.

    In camera sharpness should always be set to -2. Absolutely correct! Sharpening produces fake detail which the codec will have to work hard to encode. Sharpening is always done better in post.

    Bigger P and B frames are always good. Not true – with static scenes B and P frames serve no purpose insofar as image quality is concerned because they primarily encode changes to the image. If things are correct, static scenes will produce small P and B frames.

    Bigger P and B frames are always bad. Also not true – with highly dynamic scenes P and B frames get bigger because lots is changing.

    Smaller I frames coupled with big P frames is bad. No, it’s perfectly normal behavior because with highly dynamic scenes because there is usually some motion blurring – which reduces I frame size – combined with big changes from frame to frame – which increases the size of P and B frames.

    Individual frame quality can be improved without raising AQ. Mostly not true. The exception to this is when the codec runs out of bandwidth before completing a frame, resulting in image degradation in lower parts of the frame (macroblocking). The quality of motion encoding, on the other hand, benefits greatly from bitrate increases irrespective of increases in AQ.

    Bigger I frames means better individual frame image quality. True.

    Bigger P and B frames means better individual frame image quality. Only if there is motion, otherwise big P and B frames do little to contribute to individual frame IQ.

    Bigger P and B frames means better motion encoding. When they appear in high motion scenes, absolutely true.

    Improvements to image quality are readily visible. Absolutely not true. Some of the most important improvements only become visible in post, when color grading, etc…

    To put all this together; consider a test where high bitrates coupled with AQ 4 produces an unacceptable amount of noise at ISO 3200. Moreover, the stream is exhibiting artifacts, such as odd looking pulsing in the frame sizes. What is happening is that the codec is being overly stresses trying to encode an inordinate amount of noise. Raising the bitrate won’t help; that just stresses the codec even more. Lowering AQ will seem to help – but it isn’t really doing anything except lowering the quality, which would be better achieved by just lowering the bitrate. The solution here is to apply a little in camera noise reduction. If you just lower AQ, all gradients and image subtleties will be degraded, whereas NR is only applied to the darker parts of the image near the noise floor.

    Of course, all this assumes that you will be doing some post processing. I assume most people interested in advantages brought about by the hack do that.

  • 45 Replies sorted by
  • Only one word. Brilliant.
  • I was just about to write the exact same thing*. Oh well...

    * Yeah, yeah, Vitaliy's response... thanks Chris.

  • Spot on...
  • Excellent write up Chris.

    Lol... @astraban...yeah....right.
  • +1000

    Great write up Chris.
  • Super helpful, thanks Chris!
  • I like. Thanks Chris. It is truly the thought process shining.

    One day... I hope AQ value can be changed from a camera menu on the fly. Also GOP value.
  • @cbrandin

    The only question I have is how does the contrast setting affect dynamic range? I thought that only the sensor could affect dynamic range? Doesn't the contrast setting just move the black point or white point up and down the scale? Therefore, adjusting the contrast is only affective if you prevent it from clipping the highlights or clipping the shadows. Otherwise the setting should be irrelevant in terms of post processing.

    I may be way off base with that. That was just my understanding of it.
  • Check out tips from @stefanos. He covered the contrast setting well. I think I read his comment a dozen times.

    Prolly I would need to read Chris's post a dozen times as well.
  • Thanks a lot Chris, very helpful, I hope everyone reads this.
  • Always sharpness -2... might not be true. As @stefanos pointed out, it might depend. Then he said always -2 NR. That might not be true. I'm not saying who's right or wrong. It's really open to discussion. Food for thought.

    Contrast, sharpness, saturation, noise reduction.... all depends... I guess. There seems no fixed value for all cases. The Art of Film Mode Settings!!! Each time I think I got it... then I find out I don't know crap about it. Haha.
  • @cbrandin -- Yeah I am gonna print this and keep it in my sir are as Vitality said "brilliant"...
  • I read the post by @stefanos (thanks, @stonebat). I found his take on things pretty consistent with mine. Issues like contrast, saturation etc... Ideally you would always want to set them to use the full dynamic range - but knowing when that happens is more of an art than a science because real time vectorscopes aren't available (although the histogram can tell you a little about contrast). I was just trying to get to the point that low contrast isn't always good - just sometimes, which is what I think he is saying.

    I have to disagree with him about sharpening, though. I think the advantage he sees is not because of the camera sharpening, rather it's because of the advantages of multi-pass sharpening. He just happens to be doing the first pass in the camera. I contend that it would still be better to do two passes in post.

    When you consider these issues in light of increased bitrates, lower QP settings, etc... I think the reasoning changes somewhat.

  • By the way, multi-pass noise reduction also works well. Doing a little NR in camera can help the codec to apply bandwidth to more useful aspects of the image. I would never suggest doing all the NR in camera - just some of it;)

    To me the issue is that sharpening never, under any circumstances, adds real detail; whereas noise reduction does remove real noise.

  • Thanks Chris! BTW would be interested in ppl's favorite sharpening apps for post if sharpening is turned down in cam.
  • I just use the sharpening plugin that comes with the editor. Are there better 3rd party plug-ins? I know that there are great 3rd party NR plugins, so I guess it would follow that somebody has a better sharpening plugin too. Good question...

  • All good points so how about three threads for best overall settings for movement, best overall settings for image, and best overall settings for movement & image. Whereby bitrates, QA, GOP, in camera settings, etc... are nailed to the most recommended settings for the three requirements... Cant wait to get stuck into testing from tomoz when I get back from Greece!
  • @cbrandin In Vegas I use convolution. Very quick - processes in real-time in HD. Most video editors will have a convolution kernel.
  • Unfortunately there are no best overall settings. You have to consider lighting, subject, nature of movement, etc... That's why it's important to understand some of the underlying issues. I can see where there could be best overall settings for your particular style of shooting, though. You have to put that together for yourself.

    I can see where eventually most people will end up selecting from a relatively small collection of settings - but I don't think we're there yet. Short GOP, for example, still needs work. Until the issues with short GOP are solved, we won't see the progress in that area we want.

    As for motion vs. detail - that is what the AQ settings are for.

  • Hmmm... multi-pass sharpening and mult-pass noise reduction. photo can be easily processed by Adobe Camera Raw and Photoshop CS5. For video I had to google.

    4. Apply Neat Video noise reduction process (enable ‘very low freq’)

    * Very low freq is for very dark noisy shots like the one shown in the demo. If you are cleaning up well exposed footage you don't really have to check that unless there's large gradients (like a blue sky).

    5. Apply ‘blur & sharpen>unsharp mask’ (default settings)
    6. Apply ‘blur & sharpen>sharpen’ (12)

    *These settings are to be adjusted to your need I really don't recommend "12" of sharpening on every shots. Unsharp mask default of 50 is a good starting point but again adjust to your needs.

    Sharpening could also be done when doing final grading in your NLE. It would be better to do it at that point so you can adjust sharpness for each shot and go back and forth for matching the look of your sequences. In Premiere CS5 you have access to the same sharpening tools as in After Effects.

    In extreme cases of noise reduction where 1 part of the frame is well exposed and the rest not so much I recommend cleaning up the shot and then dragging another instance of the same footage but with no processing and masking the well exposed parts (add some feather around your mask to soften it up). That way you clean up the bad compression blocking but you retain the original sharpness of your well exposed elements. That might require roto if your subject is moving but it doesn't need to be super detailed roto...

    7. Apply ‘noise and grain>noise’ (set amount to 0.4)

    In-camera NR + the step 4 for multi-pass noise reduction. The step 5 & 6 for two pass sharpening?

    Both Premiere Pro CS5 and After Effects CS5 support unsharp mask and sharpen effects. Neat Video plug-in supports sharpening feature, too. Neither sharpening nor noise reduction takes advantage of GPU acceleration.
  • "Improvements to image quality are readily visible. Absolutely not true. Some of the most important improvements only become visible in post, when color grading, etc…"

    I hadn't put 2 and 2 together on this... very enlightening. For a novice trying to learn the right way, it's a privilege to have access to this community. Thanks for posting!
  • @stonebat

    I'm a big fan of NeatVideo!