GH4 100 mb/s better than 200 mb/s??

[Giulio Cosmo Calisse]

I just did some 1080p tests in both 200mbs ALL-I and 100mbs, and the 100mbs shots actually grade a lot better!

Not sure why it's operating this way - is 200/ALL-I just a complete gimmick based on bitrate theory, but actually degrades image quality?

I decided to do a test where I bumped shadows/highlights, saturation and sharpness to strain the codec.
Here are a couple of stills!

GH4 200mb/s

 

GH4 100mb/s

 

[Cassius McGowan]

Links not working

[MattH]

4k reduced to 1080p is so much better in terms of detail and lack of artifacts on the GH4 that there is no point shooting 1080p anyway.  All I gives more accurate temporal samples, but any one frame might have more compression because there are more frames to encode.

[Julian]

He's talking about 1080p 100 mbps.

 

Wasn't the conclusion with the GH3 that AVCHD 50 mbps was better than MOV 72 mbps? And with the 5D3 (h264) also the ALL-I had some troubles if I recall correctly?

 

Don't ask me why... and I think it's weird indeed. Maybe it's not so much the grading that makes a difference but macro blocking or movement?

[Ben Corwin]

Very interesting. Thanks for the post! IPB compression for the win!

[FilmBrute]

Nice, post. Where I see compression is macroblocking in moving images, particularly shadows. A test like that would be harder to recreate. Also, I would image the higher you go, the more diminishing the returns are. I don't know the science of compression, but it is pretty remarkable how you can compress a file 10:1 and it won't look 10x worse.

[Michael Ma]

It looks like IPB is a clear winner for this test.  I'm definitely an IPB convert on my GH3.  I was shooting ALL-I for a few months, but aside from the better quality image on IPB, the processing benefit of ALL-I on a descent computer is negligible.  

In Premiere Pro it converts to 32bit 4:4:4 on the fly in the NLE* or while encoding, before doing any post processing.  This on the fly conversion is really not what's causing the slowdown in the encoding.  It's the post processing you apply on it.  On a modern computer i7 with at least 8GB of ram, you would only save a few seconds encoding a 5 minute video shot with ALL-I instead of IPB.

 

*Source: http://tv.adobe.com/watch/premiere-pro-cs5-feature-tour/staying-native-or-going-intermediate-transcoding-and-premiere-pro-cs5-/

Jump to 5:50 mark.

 

In short, very little reason to use ALL-I on the GH3.   It looks like GH4 is going to be the same from these preliminary reports.

[jcs]

If no motion artifacts, IPB is more efficient and tends to work better. The same is true for the 5D3.

[Andrew Reid]

Did you grade first frame or a middle one? First frame can be blocky with some codecs. I saw this on D5200 and GX7, various others.

 

Better check that.

 

I won't be shooting in any of the 1080p modes in-camera on the GH4 because you can easily do 1080p in post with the 4K and get a much better image that way.

[Giulio Cosmo Calisse]

I did notice the first frames were blocky so I picked a middle one on purpouse.

I did not expect IPB to have THIS much more color information on still shots!

So ALL-I should be better with motion because it saves individual (albeit more compressed) frames instead of interpolating?

[pietz]

im not surprised by this result at all. it would have been my first guess that 100mbps ipb is better than 200mbps all i on the GH4 and any other camera to begin with. especially in a comparison like the one you made. let me explain:

 

most of you know what i, p and b frames are and how they work, so i wont get into it. for the 200mbps option to be better in image quality than the 100mbps one, all B and P frames from the lower setting need to be larger (in file size) than half the size of an i frame. lets talk numbers: an i frame in the 200mbps is pretty much exactly 1MB in file size, that means the b and p frames from the 100mbps setting need to be between 500kB and 1MB, so the 200mbps option is superior in picture quality. but thats just rarely the case. b and p frames are usually smaller than half the size of an i frame.

 

in an example like yours with a locked camera the b and p frames will probably around 10 to 20% the size of an i frame, because you have very little change from one frame to the other. if the camera uses 100% of its available bandwidth, which my theory is based upon, the 100mbps will bring you much better quality, because the i frames in the 100mbps setting are "allowed" to be larger than in the 200mbps option.

 

the only advantage of shooting only i frames is that your computer has to work and calculate less. meaning it works better and its response is faster. if however youre on a very fast machine anyway, theres not a legitimate reason to use 200mbps all i.

[aldolega]

Ease of editing/processing is not the "only advantage" of All-I: its motion also looks more natural (and yes, "filmic"). Yes, long-GOP's image quality looks better (at these two compared bitrates) with static or near-static subjects, but not everyone shoots such subjects, or is able or willing to limit their camera moves to a snail's pace.

Personally, shooting an action sport (rollerblading), I am concerned with not just motion rendition, but also rolling shutter, so I expect I'll be using about 20% 4K, 40% 1080p IPB, and 40% 1080p All-I.

[Julian]

40% 1080p IPB, and 40% 1080p All-I.

 

Just curious: how are you going to decide when to shoot All-I or IPB? I can understand you'll shoot 4K when there is less motion involved because of the rolling shutter, but what would trigger you to change to 200mbps? Extreme movements?

[aldolega]

Yep, fast-moving shots with a fisheye or ultrawide, mostly. Usually following/next to the skater, or getting right next to an obstacle and doing a fast pan/tilt motion, following the skater as they perform the trick.

[hmcindie]

The same is true for the 5D3.

 

No matter how many times you say that in different forums...No. IPB is clearly worse than ALL-i on the 5d3. Even moderate movement completely smooths out every single detail with IPB. And the 5d3 needs all the detail it can get.

[jcs]

No matter how many times you say that in different forums...No. IPB is clearly worse than ALL-i on the 5d3. Even moderate movement completely smooths out every single detail with IPB. And the 5d3 needs all the detail it can get.

 

Starting at 3:45, see ALL-I compared to IPB, especially the "10" section on the lines (horizontal or vertical): 

There's more detail in IPB (from this thread ~2 years ago: http://philipbloom.net/forum/threads/intraframe-all-i-and-interframe-ipb.987/ ).

 

Here's an example I shot handheld (not a locked off shot- there's motion): http://www.dvxuser.com/V6/showthread.php?279229-Canon-5D-Mark-III-IPB-contains-more-detail-and-has-less-artifacts-than-ALL-I/ , as well as another discussion: http://www.dvxuser.com/V6/showthread.php?278746-ALL-I-vs-IPB-IPB-Wins

When policar posted a still frame grab of fast motion in 1080p in ALL-I vs IPB, the ALL-I still frame did look less smeared. However, during playback this kind of motion detail may not be visible. It's really important to test for yourself vs. simply following the advice of others. I did find that 720p60 always look better with ALL-I on the 5D3 which I added later to the thread.

 

ALL-I vs IPB is a popular topic. The common argument is "ALL-I must be better because the bitrate is higher and the compression is less". However, in real-world practice, most people can't see a difference. There are many tests out there showing this:

 

Again, for motions shots, if there aren't any objectionable artifacts for IPB (including excess high-frequency filtering / detail blur), then IPB can be a better choice to save disk space.

 

While I haven't done a test for medium or fast motion, for low motion, the 24Mbps FS700 AVCHD provides more detail than 5D3 RAW (in the same way the GH4 provides more detail than 5D3 RAW: '?do=embed' frameborder='0' data-embedContent>> .  I purchased a Nanoflash external recorder for the FS700, however I've never used it beyond testing as the 24Mbps internal codec's quality outweighs the extra weight and complexity of an external recorder's marginal image quality increase (again, except in motion shots and certain scenes such as water, etc.).

 

If you have an example or know of one showing ALL-I clearly doing a better job than IPB, it would be helpful to see under what conditions this can happen, e.g. what is the motion threshold where ALL-I starts looking better than IPB (and is it visible during playback vs. pixel peeping a still frame).

[Vesku]

I have shooted sports with 50P GH3 50mbs ipb mov. When shooted with high shutter speed lots of motion in frame I could hardly find any artefacts or smoothing.

 

Now GH4 has more efficient codec and double bitrate. I wonder if anyone could find artefacts in GH4 100mbs 1080P footage. I think that 200mbs frame compression quality loss is bigger than 100mbs motion artefacts.