@Vitaly It is ok for interlaced footage to have less details. Just make sure to use proper deinterlacer.

Can't imagine that alone this is the problem. Would understand it if there is something moving. I've been trying several players with the same result (Splash Pro EX, Media Player Classic, VLC)

@alcomposer
Please move deinterlacing and details talk out of this topic.

Dreamer, the pictures in 24p and 50i have slightly different either picture or pixel aspect ratio. That's clearly noticeable in your pics if you switch between them - it's like one of them is squeezed vertically. It might indicate that these are just two completely different sensor modes with different downscale algos. If so, codec tweaking is of no help here - the 50i will be always behind in terms of apparent resolution. It's just made so.

Have checked ExTele, as pal/ntsc and time limitation. The push on fn2 (my shortcut for Ex Tele) on the body result on the ability to choose YES or NO for Ex Tele even when connected by HDMI to an LCD monitor. BUT the EX Tele ( the zooming ) still does not function when connected to HDMI :(

1_ loaded modified firmware with 3 cases checked: a) time limitation; b) pal/ntsc; c) ex tele check disabled
2_ connected the GH2 to a lcd monitor by HDMI
3_That I see: I can now choose between yes or no when I ask to use Ex Tele but my choice between any of the two options has no effect on the picture : Ex Tele is inactive in the external LCD monitor as in the internal lcd.

Before the patch: I couldn't activate the choice between yes or no when I asked to use Ex Tele when connected by HDMI to an external monitor. Nothing happened when I asked Ex Tele connected to and external monitor by HDMI, even the display of the choice.

@sebalto "BUT the EX Tele ( the zooming ) still does not function when connected to HDMI :("

did you try to start recording? and check the memory card in a computer?

I not 100% sure but I think in the GF2 you don't see instant feedback of the Ex-tele Fucntion. but pretty sure the video recording is with the extender active.

@Vitaliy_Kiselev If PTool v3.60d has been updated, surely it is a new version and should have a new version number?

@Ptchaw

From Vitaliy
"PTool 3.60d updated
Like all minor updates previously it can be spotted by looking at the release date in the PTool title." The latest one has 060711 in the title block.

@andycorleone I have tried to record with "yes" choosen and eventually the zooming occured ! But it led to a crash: no way to stop the rec neither to switch off the body. I had to remove the battery. I repeated the process two times to check.
@dkitsov The magnification doesn't fonction even with patch above loaded when connected by HDMI to a monitor.

1080 50i was already ok before hack, as was 720 50p. 1080 25p isn't fonctionnal yet.

I know this isn't what everyone was hoping for. However, I have successfully got the GH2 to record 1080p @ 24 FPS at 17 mb/sec as an average instead of the original 22 mb/sec. The point of this was to determine what the camera did when we lowered the max bit rate. It appears to not only lower the max bit rate but it also lowers the average bit rate.

Whenever we try to raise the max bit rate it keeps the average bit rate the same. This causes the codec to over encode, under encode, over encode.... and so on. I expected the codec to do the opposite when I lowered the max bit rate under the default value. It did not do this. Instead it simply lowered the average bit rate along with the max bit rate.

In the tests I did the max bit rate was about 18 mb/sec and the average was about 17 mb/sec. When I use the default firmware the average bit rate is 22 mb/sec and the max is 24 mb/sec. Sorry I can't post screen shots and stream parser data at the moment.

There simply must be another setting that controls max average bit rate. It may be in the Encoder Setting variables. I haven't tested those yet. However, it could be an entirely different variable as well. Is there anyway to search the firmware for every setting that has a value of 22,000,000 or 18,000,000?

@mpgxsvcd I have the very same observation.

I've created HACKchart, a 1080p / 16:9 test chart (attached). You can print it, and then shoot it to debug frame corruption (especially fps / bitrate testers).

Having a standard chart will make it easier for V. and other testers to analyze wrong aspect ratios, corrupted colours and where parts of the sensor image end up in a scrambled frame.

If you get problems like that, shoot the chart so it completely fills the frame (or close-as) and post the a screen grab.