GH2 Is any GH2 hack - so far - worth it in the real world?

O

OldCorpse

Rockin the Boat
OK, I don't want anyone to take it the wrong way, I'm just asking a question. It's pretty much agreed that the GH13 is a demonstrable improvement over the GH1.

My question: is any GH2 hack - so far - as much of a jump over the plain GH2 as the GH13 is over the GH1?

Or is it too soon to ask and we're still waiting for the great GH2 hack to come out?

I ask, because before one goes to the trouble of hacking a brand new GH2, it would be nice to know what one is getting - and I mean this in the real world. I am not so much interested in "well, if you really pixel peep at a special angle while wearing blue glasses and standing on your head, you'll see a glimmer of a difference". I mean, what concretely is better so anybody is likely to see it, or for it to have a *practical* difference if someone is using it for real film work.

Let me add, that I have full respect for the work of Vitally and all the great hackers out there who bring us these marvelous possibilities. I just want to know if it's safe to jump in at this point, or is it still early days?
 
I'll second this - I hacked one of my GH2s and really can't see any difference. But it was with the early 42Mbps hack - so have just been waiting for someone to make an obvious breakthrough. Best examples I've seen so far are with the low-GOP hacks, but still don't see anything that justifies the enormous files created.

Anyone seeing that elusive breakthrough yet?
 
Another thing to consider is that if you have to send the GH2 to Panasonic for a warranty-related repair and, even if you manage to put the Pana firmware back on before sending it out; is Panasonic going to recognize the leaked firmware (the only available firmware) as legit and will it negate the terms of the warranty?
 
the higher the bitrate in 72060p for me is the main reason to hack for slo mo. i have it and all the high avchd settings at 32mb(this is supposedly the highest you can go with zero adverse effects) then i have the 24pL settings at 24mbps now. The possibilities that im checking for still are clean hdmi out and a higher than 1080p res for the etc mode. i had it hacked all the way to 65 mbps and couldnt really see a diff in the exact same shot at stock settings. im not much of a pixel peeper tho.
 
OldCorpse said:
OK, I don't want anyone to take it the wrong way, I'm just asking a question. It's pretty much agreed that the GH13 is a demonstrable improvement over the GH1.

My question: is any GH2 hack - so far - as much of a jump over the plain GH2 as the GH13 is over the GH1?

Or is it too soon to ask and we're still waiting for the great GH2 hack to come out?

I ask, because before one goes to the trouble of hacking a brand new GH2, it would be nice to know what one is getting - and I mean this in the real world. I am not so much interested in "well, if you really pixel peep at a special angle while wearing blue glasses and standing on your head, you'll see a glimmer of a difference". I mean, what concretely is better so anybody is likely to see it, or for it to have a *practical* difference if someone is using it for real film work.

Let me add, that I have full respect for the work of Vitally and all the great hackers out there who bring us these marvelous possibilities. I just want to know if it's safe to jump in at this point, or is it still early days?
Click to expand...



It's certainly a different game to this time last year, when the GH1 hack was exploding. The GH2 hack seems to be plodding a bit, and I have to admit that I can't realy see any difference between the hacked and unhacked samples (but that's probably just me)

My own opinion, fwiw, is that things are about to change dramtically with the Sony Nex 7 / Apha 77. I'd lay a modest amount of money that by this time next year 50% of the people currently using GH1/GH2 will have bought a sony, and will be posting in a new DVXUser Sony forum, along with former Canon & Nikon users :happy:

Seriously - they sound amazing, and Panasonic will lose shed loads of sales to them.
 
Well it's "simple": you can't do a comparaison between the GH1 and the GH2 hack process for a reason : the stock gh1, in many aspects was crippled, and its full potential was revealed by the hack of VK. Before/after = s**tty/decent footages.

On the other hand the stock GH2 is already pretty decent : in my view, near if not better (for the footage quality and because of special stock features : crop, hdmi monitoring, etc.) than the GH13. It need some (hard) research to VK and others to fully understand the specific AVCHD encoding process to maximize the quality in a stable and "definitive" hack.

For now, some see an improvment in their footages. Fine. If you're not seeing the same thing, well, you don't have to apply the hack, be happy with the decent little camera you have and wait for the improvement that will fit your need (clean hdmi out, customizable film mode, or whatever...).

And if the comparaison between the hacked/stock GH2 appears not convincing enough to some, it seems to me more solid than the rumorish/marketing bulls**t about the tremendous quality of camera that nobody have touch so far (same old, same old)... maybe Sony enginers have done a good job, and I'll be happy to havetheir new toy ; maybe not. For now we only have the proof of their marketing skillz.

PS : excuse my french!
 
I think it's mainly because the GH2 had proper 24p from the get go. IMO the hacks do make a very real and big difference to the 1080i and 720p encodes
 
I have two GH2's. I hacked one of them to 42 Mbits. In most cases you will not see a dramatic difference. You will see a difference in
water shots or anything else with really fine detail. I did a side by side test and could really see the difference in some shots
I did of a pool. I did not keep the footage. But the test satisfied me that the hack was worth it.
 
The GH1 was hacked because there were issues with the footage that people wanted to fix. The hack worked quite well to resolve those issues.

The GH2 was hacked just because it could be hacked. The stock 22 mb/sec footage out of the GH2 is indistinguishable from the $5000 AF100. With that being said it doesn't mean the hack is not useful.

You can use two simple settings to allow the camera to go up to 32 mb/sec which in all honesty is plenty especially with B-frames. There are no adverse affects and the camera will operate just like the day you bought it except at a higher bit rate.

All of that might change later today though. VK is supposed to release fixes that could make the 65 mb/sec GOP 3 settings have no adverse affects and it also might raise the minimum bit rate for less complex scenes. Those less complex scenes are where I think the Hack has the most possibility to show improvement.
 
I admit it's hard to tell immediately the differences in video quality. But for me, a frequent user of ETC mode, I can say that the grain is a lot smoother. Same case for higher ISOs in normal mode.
 
ed_lee83 said:
I admit it's hard to tell immediately the differences in video quality. But for me, a frequent user of ETC mode, I can say that the grain is a lot smoother. Same case for higher ISOs in normal mode.
Click to expand...

The bit rate that the camera encodes to varies directly with an increase in ISO value. The amount of noise actually affects the bit rate more than the amount of motion does.

Think of it this way. Noise is the constant changing of individual pixels throughout the picture. Even if you have very fast motion in a frame the moving subject does not always cover the entire frame. Noise is by far the hardest thing to compress without a quality loss. It stands to reason that using an extremely high bit rate would help accurately represent the noise. However, it is up to the viewer to determine whether that is a good thing or a bad thing.
 
Modestly hacking the bitrate upwards is absolutely worth it if you use the 720P mode! Shoot someone dancing, or whatever, in front of a foliage background and watch the background - not the dancer - in playback. On an unhacked GH2 you will see a strange kind of pulsing effect, which goes away with a 32Mbps bitrate hack applied.
 
High ISO and noise reduction

High ISO and noise reduction

The comment about bitrate and high ISO is right on the mark. I know many people suggest setting noise reduction at -2, I think in some cases this is ill advised. If there is a lot of noise the codec will be consumed with encoding it - to the detriment of doing a better job with more useful image elements. Over doing noise reduction obviously degrades detail, but up to the point where you reach the signal to noise ratio it should be helpful. The higher the ISO, the lower the S/N ratio, so the higher you can set noise reduction without removing any real detail. This is a different issue than sharpening - which never adds any real detail, rather it forces the codec to consume itself with the fake detail sharpening produces.

Chris
 
cbrandin said:
The comment about bitrate and high ISO is right on the mark. I know many people suggest setting noise reduction at -2, I think in some cases this is ill advised. If there is a lot of noise the codec will be consumed with encoding it - to the detriment of doing a better job with more useful image elements. Over doing noise reduction obviously degrades detail, but up to the point where you reach the signal to noise ratio it should be helpful. The higher the ISO, the lower the S/N ratio, so the higher you can set noise reduction without removing any real detail. This is a different issue than sharpening - which never adds any real detail, rather it forces the codec to consume itself with the fake detail sharpening produces.

Chris
Click to expand...

Some people actually want to retain that noise detail. In some ways having well defined noise is better than having noise that is smeared(even a little) from noise reduction or compression.

Anyway, the hacking game is about to change thanks to Chris and VK. The next Ptool will offer settings that everyone can live with. Thanks for your efforts. I think it is safe to say that the GH2 is the most highly desirable video camera thanks to what you guys have done.
 
Good point. In many cases, however... Can't you add well defined noise (or any kind of noise you want, for that matter) in post? I wonder if you could see the difference. It seems like there might be a balance here - the more noise that remains because of no NR, the more that noise will end up being corrupted by compression because the encoder runs out of bandwidth.

Chris
 
While the AVCHD is still plodding a bit, the hacked GH2 gives a superb 1080p MJPEG @ 100 mbps. I'm keeping my GH13's for the AVCHD/25p...but the GH2 for when I need MJPEG @ 30p. Best of both worlds...
 
Skeptikal said:
While the AVCHD is still plodding a bit, the hacked GH2 gives a superb 1080p MJPEG @ 100 mbps. I'm keeping my GH13's for the AVCHD/25p...but the GH2 for when I need MJPEG @ 30p. Best of both worlds...
Click to expand...

1080p @ 24 FPS and 42 mb/sec is vastly superior to MJPEG even at 180 mb/sec. VK's latest Ptool build has completely stable and optimized 42+ mb/sec 1080p,720p, and 1080i. If you haven't tried Chris's new optimizations then you have experienced really high quality video.
 
GrahamH said:
Modestly hacking the bitrate upwards is absolutely worth it if you use the 720P mode! Shoot someone dancing, or whatever, in front of a foliage background and watch the background - not the dancer - in playback. On an unhacked GH2 you will see a strange kind of pulsing effect, which goes away with a 32Mbps bitrate hack applied.
Click to expand...

GrahamH is right. The hack is good and cool for 24p, but for MJPEG and 720 modes, it's a godsend.
 
The devil is in the detail here. If you do grade your shot, the hacked gh2 (GH23) is much better. You don't see the typical macro block that usually pop up when the image breaks down. In shadows you will see marked improvement. Where it was a mush before now you can see details.
 
You must log in or register to reply here.
Back
Top