GH2 GH2vk Unleashed

I did. I ran another test with more detail, and a higher average bitrate (32Mbit), and my camera didn't freeze.

The file was slightly smaller than the last, so I still need to test it further.

Here is file analysis. 35Mbit for 24H, with GOP of 6 and 30 minute time checked.

View attachment 37464

greeches, the gop is set to twelve in that shot. Is that the right one?
 
I did. I ran another test with more detail, and a higher average bitrate (32Mbit), and my camera didn't freeze.

The file was slightly smaller than the last, so I still need to test it further.

Here is file analysis. 35Mbit for 24H, with GOP of 6 and 30 minute time checked.

View attachment 37464

Do you get the low bit rate frames at the beginning with this setting? Your picture is too small to tell. Looks like we are both in the same area(NC). What lenses do you have?
 
Yes, I got the low bitrate for the first second or so. You could see the pixelation in the LCD on playback.
 
So can the GH2 finally shoot 1080p30 with the hack?

How about defeatable AGC?

Those 2 things are all I care about.
 
So can the GH2 finally shoot 1080p30 with the hack?

How about defeatable AGC?

Those 2 things are all I care about.

It could always shoot 1080p @ 30 FPS slowed down to 80% speed. However, I don't think anything has changed at all with respect to frame rates. They are too busy working on getting the absolute highest bit rate without regard to whether it actually produces a visual benefit or not. I wish they would just leave well enough alone and start working on the frame rates instead.
 
We are not working on "on getting the absolute highest bit rate without regard to whether it actually produces a visual benefit or not ".

We are working on understandin encoder workings, improving 720p, improving stability, etc. Many things. And it actually produces visual difference.

About frame rates. I spend big time on this with no significant progress. It is just too hard for now.
 
@Vitaliy I have a few questions that could use your light.
1. What is the best 1080p24 setting you were able to come across that is stable enough to work with a class 10 card? (Compression and GOP)
2. What do the values mean for the AGC settings?
 
The only thing we have to do here is thank Vitaliy for his magnificent work. That's all. And wait patiently for new developments. BTW out of curiosity, any posibilities on Mjpeg 24p? Thanks again Vitaliy.
 
We are not working on "on getting the absolute highest bit rate without regard to whether it actually produces a visual benefit or not ".

We are working on understandin encoder workings, improving 720p, improving stability, etc. Many things. And it actually produces visual difference.

About frame rates. I spend big time on this with no significant progress. It is just too hard for now.

Vitaliy, I didn't say you. I was referring to everyone else. The work you and Chris are doing on Quantizers is where I would expect the most benefit. If you can raise the minimum bit rate along with the average bit rate that could be very beneficial. I just wish we could test the specific versions of ptool you guys are using right now as well. However, I understand why you haven't released that version yet.

I understand that altering frame rates may not be possible. However, I just can't write it off yet.
 
The only thing we have to do here is thank Vitaliy for his magnificent work. That's all. And wait patiently for new developments. BTW out of curiosity, any posibilities on Mjpeg 24p? Thanks again Vitaliy.

That and donate. Every user who has applied and benefited from the hack should consider donating. $20, or even $100, is peanuts considering the gain in image quality improvement. You would have to spend thousands more to buy a camera as good as the GH2vk.
 
I've read through most the information posted on these forums, & Vitaly's, still it's hard to understand which settings are truly stable. I previously hacked the GH1, so understand the basics.

While higher bit rate is desirable, stability is most important to me, & the only adjustments made were raising FSH/SH, and 24H, to 32000000. I've tested these settings with Pappas' CODEC Test Chart V.2, & the 14-140mm lens. Each clip was 10 minutes long.

With Transcend 32GH Class 10 SDHC both 1080/24, & 720/60, appear stable, even with zooming. So did EX TELE CONV. After rewrapping the files Quicktime showed an average bit rate a little higher than 32Mbit, & I'm looking forward to trying these in the field.

Are there any glaring flaws in these settings, or methodology?

Thanks,
Bern
 
Last edited:
Back
Top