GH3 GH3 vs. unhacked GH2

Not just a matter of detail - nor requiring magnification - the benefits were considerable;
The color improved greatly, the dynamic range widened (a bit at the top, a lot at the bottom), the motion smoothed, the resolved detail increased.
If one is shooting narrative, patching with Flow Motion 2 is highly recommended.
 
There is no doubt that hacking can produce slightly clearer footage of very busy images, viewed magnified... So can more expensive cameras. Just like I don't feel the need to change cameras just for that tiny extra, mostly not needed, detail, I don't feel the microscopic gain with hacking is worth the extra trouble, larger cards and files, instability and worries... And the upped detail level often brings more visible noise to tend to...
This kind of willful denial of hard evidence is truly a sight to see. When the unhacked GH2 is unable to handle the amount of detail in a scene, it doesn't just get a little muddy, it collapses into a severely degraded Fallback Mode. Panasonic uses this drastic mechanism to make sure the AVCHD encoder will not fail and stop recording when overloaded. What's alarming about it is that the GH2 can do this surreptitiously without giving you any indication that the camera has recorded a coarsely degraded video.

For typical consumer use, Fallback Mode may be a reasonable overload solution, but for professional purposes it makes the unhacked GH2 an unacceptable risk. Perhaps you're willing to examine all of your unhacked footage in Stream Parser to verify that it was recorded properly, but I am not. For the GH2 to be of any serious value, I must have confidence that every successfully completed video clip has been recorded at the intended level of quality I selected.

Fallback Mode is the dirty secret of the GH2, and reliably disabling it was my prime motivation for developing Flow Motion v2. Fallback Mode is the reason why I limited the AVCHD encoder to a peak bitrate of 100Mbps, and also why FM2's 720p video modes use a 6-frame rather than a 3-frame GOP. When running with Fallback Mode disabled, it's necessary to allocate enough bitrate to safely handle overload conditions, but not so much that it destabilizes the encoder. Successfully controlling all the variables involved is a delicate balancing act, even for Panasonic's engineers, who have direct access to the firmware code. With the GH3's All-Intra video mode, for example, they were able to stabilize the encoder to work reliably at 72Mbps, but that was still not enough bitrate to deliver consistently high quality video.
 
To me, personally, using an unhacked GH2 is like using a severely crippled camera. I don't understand why anyone, willfully, would do that. And please don't get me started on stability, etc. All of the major patches are perfectly safe and stable, most of them span. There are dozens of settings that suit all possible needs. Long GOP, short GOP, Intra, high bitrate, low bitrate, you name it, they got it.
To hack or not to hack - I don't quite understand why such dilemmas even exist. Pure nonsense.
 
You seem to forget that I have tried a load of hacks, including Flow Motion 2... And yes, I was able to see a slight difference (I'm not a blind, conservative idiot in denial ;o)... I haven't encountered the unhacked collapses, the muddyness you describe. Guess it depends on what you shoot, and how. I use a tripod almost all the time. No whip pans, expose very carefully... But even footage shot handheld from a speeding boat in the direction of the sun has been recorded without any trouble whatsoever, and with some post stabilizing looks magnificent, also when scrutinized on a high quality NEC monitor... I have shot a lot of running water, and haven't found any situation where I couldn't intercut the hacked and unhacked shots... this is NOT a severely crippled camera! IMHO of course!

I'm not on any crusade to persuade anyone to follow my example! So hack and develop to your heart's delight, and let me (and a lot of other users) choose not to! GH2 hacking is not a religion, although some of the language would suggest otherwise!
 
I think to a large degree, ya gotta let the footage speak for itself. The videos have been posted.
Interestingly, while the frame grabs show a very significant difference, when both shots are running they both have pretty nasty looking macroblock artifacts pop in and out of the scene. I'm not sure that the hacked one is significantly better in motion. That is to say, if it's better it's still not good enough to eliminate the blocks that pop up and catch your eye as being unnatural.
 
I haven't encountered the unhacked collapses, the muddyness you describe. Guess it depends on what you shoot, and how.
Yes, whether the GH2 collapses into Fallback Mode is dependent on what you shoot. The alarming thing about it is that no one really knows what may trigger it and for how long it may persist. And it may well have contaminated your footage and escaped your notice.
 
Well I shoot surfing footage for 2 years with 2 gh2. Also shoot many other junks. Never saw fallback mod. Always work perfect. I have another friend gh2 shooter and never have problem.I try hack one month and off it because looks same. It's palcibo imo.
 
Good to hear I'm not alone in liking the camera as is ;o) And judging quality on still frame grabs is a little strange really. It should always be judged by the final presentation while viewing a video, but of course on a quality display or projected to its full capacity, and not murdered by Vimeo or YouTube compression. So far my GH2's have done very well through 100ds of hours of shooting, after my initial hopeless learning phase of course... for the bad looking stuff I have made, I can only blame myself and not the camera...
 
If you can't tell the difference between hacked and unhacked then you certainly don't need it. Depending on your eye for quality the difference could be perceived as slight, or it could be perceived as the difference between usable for narrative film and not. For anyone on the fence, a quick search online will yield a ton of side by side footage from which you can make up your own mind.
 
This is what what convinced me to go to the moderate hacks of Sanity and Flowmotion (I use Flowmotion for things I really want detail in, and Sanity for every day shooting). Watch the two clips just after the 40s mark in 1080p, and tell me which was hacked and which wasn't. By my eyes it's not hard to tell at all. Throughout that video there is a mix of hacked and unhacked footage (as well as a little GH1 footage) and you can definitely tell the difference IMO.

http://youtu.be/uhSfstGzKGo?t=40s



Like I said above, in 95% of situations you won't notice much difference at all in the GH2 hack vs unhacked, even with super high bitrate. I do think it adds a tiny bit more detail to have those super high bitratees, but THAT part could be placebo. At a certain point the video is only as good as the sensor.

But in those other 5% of situations you'll notice a gargantuan upgrade. I shoot a fair amount of water stuff (like above, outdoorsy stuff) and I was really disappointed when a lot of them turned out a yucky mess. The average person may not notice it, but it's bad enough I think a client who didn't even know video would still pick it out. So I use moderate hacks. Sanity has been rock solid with my cards, and Flowmotion fails very, very rarely.
 
Last edited:
Just watched it - can't tell. A hack is not designed for watching mixed format footage on Vimeo or youtube. It is for high end viewing, broadcast and post work. Unfortunately, so much discussion goes into what makes up the high end original that very little thought goes into where the final edited footage is to be viewed - and that is total overkill and frankly just a waste of time for the average punter. Just like taking a high res photo and emailing a 640x480 image and saying how good it looks.
 
Last edited:
The codec breaks down most noticeably with lots of motion. You can look at comparisons all day long, but ultimately the codec lets you retain more detail (important in post-production, grading) and avoids PROBLEMS...

Photo below is from this thread... http://www.dvxuser.com/V6/showthrea...Motion-v2-Patch-Banding-amp-Blocking-Be-Gone!

attachment.php


These are obvious differences, if you know what you are looking for. The hack does NOT make the GH2 into an 'entirely new camera'. It just makes a good camera great! And useful alongside pros. Hence the Zacuto shootout results...

http://www.43rumors.com/zacuto-shoo...la-votes-for-the-gh2-and-beats-arri-and-sony/
 
There is more to the hack than bit rate. There is a good amount more detail, and you have the ability of going all I frames. This changes motion as well. Having done comparisons side by side of my GH2 unhacked and hacked, I can say with 100% certainty that it is FAR from a placebo effect.

Had to state that as I don't want members here getting the wrong information.
+1, agreed. There is a difference, if you don't notice it, then the hack is not for you. But for folks that know what to look for and see the improvement it's a huge bonus.
 
I'm not going to beat around the bush. If you can't tell hacked GH2 from stock GH2 then you either:

a) are an idiot
b) are in denial
c) require a visit to the opticians
d) are watching compressed comparisons on youtube (see option a)
 
I'm not going to beat around the bush. If you can't tell hacked GH2 from stock GH2 then you either:

a) are an idiot
b) are in denial
c) require a visit to the opticians
d) are watching compressed comparisons on youtube (see option a)

a) I'm an idiot
b) Not on a river at the moment
c) In the works
d) As often as not pass on YouTube links. Vimeo is much better. Am I less of an idiot?

Next time may I suggest you beat around the bush......:Drogar-Shock(DBG):
 
Back
Top