@bwhitz - If the XF300 is so similar in resolved detail, that means the 7D can resolve 1000 TVL? fascinating.

FACT: some fancy fringing and over-sharpened images.

cowpunk: it's is far less different than you think. It is the same sensor, albeit downscaled. Downscaling is done intelligently - not by skipping lines. Hence anyone can make a calculation of what the video mode SHOULD be capable of, with the right settings. (factory gh2 video compresses the video stream a lot, for instance)

The main difference to a full scan of the sensor is that pixel-level information is averaged - which in turn reduces noise and creates a lot of fine (but false) detail that exceeds the nyquist limit without causing a lot of obvious moiré or aliasing.

The EX3, and many other cameras - apply sharpening and decrease it in low light to avoid additional noise. Therefore, any such difference has only to do with perceived sharpness, and not to actual resolving power of the sensor.

@RRRR - actually, cameras actively apply noise reduction at high gain which smooths over detail. They don't merely stop "sharpening." Anyway, as I stated a while ago, when I submitted my comparison results, the XF300 was set to -4 sharpness and the GH2 was at 0. If you can't see the difference in actual detail, then it's because you want the GH2 to be better at resolving than it really is.

We can talk all day long about how the GH2 may downscale it's sensor to reach an HD image, but you're just guessing. Panasonic hasn't released that information. You don't know how intelligently it's actually working, so any calculation about how it SHOULD work is completely moot. It's nothing more than a wish.

Take a look at the chart right above. How does that documented performance of the GH2 in video mode factor into your calculation of how it "should" be working?

@cowpunk52 "If the XF300 is so similar in resolved detail, that means the 7D can resolve 1000 TVL? fascinating."

Well, the 7D was measured at about 700-800 in that last Zacuto test, so no, I don't think it can hit 1000lines... but the XF doesn't look like it can either. And I'm actually editing some XF stuff right now... it's pretty soft, and the detail really doesn't look that great... it looks electronic. Probably because the sensor is native, and doesn't benefit from the organic look of oversampling. A GH2 still image extraction looks photographic... and really could pass as a photo in most viewing situations. Where as a still extraction from the XF screams "video freeze frame". They just don't stack up.

I've compared some down sampled to 1080p from full RAW images of the GH2 of the same subject to 1080p AVCHD and MJPEG.

Luminance was pretty good in AVCHD, but is not a match for being downsampled properly from the full size image (meaning there is loss in Panasonic's down sampling and compression, though I think you'll find the same with uncompressed HDMI out), MJPEG was a little worse in luminance resolution, but significantly better than AVCHD in U and V resolution, but MJPEG still looks much worse than what 4:2:0 is.. so yeah.

I'm not sure what I'd like improved in the next generation. I honestly do not like cameras in the XF300's class. That 'handicam' form factor is the worst imho. People point to dSLR video's form factor as a weakness and then point to handicam style form factor as being a 'proper' video camera, when they area poor low end design, extremely bad ergonomics.

@bwhitz is probably comparing real luma resolution (an actual scene), ie: low contrast resolution, not high contrast resolution (test chart/silhouettes), high spatial frequencies in the real world are mostly low contrast, that is to say most detail and resolution people are interested in making films and taking photographs occur across low contrast changes.

As has already been pointed out, resolution numbers (for whatever they're worth) aren't absolute, but Alan Roberts (of the BBC), who uses a uniform methodology, puts the xf300 at about 900 lines of useable (clean) vertical resolution. The HM151 somebody referenced earlier (as being inferior to the GH2) comes in about 680 lines in Roberts' testing. The EX1/3 produces at or near the full 1080 lines in his testing.

Draw your own conclusions.....

What I don't understand is that when compared directly with other cameras in the admittedly flawed tests that Phillip Bloom did, it was visually obvious that the GH2 was more than up to the task of being usable for making films for the theater. I saw nothing that wold suggest that the image would fall apart or be a distraction for movie goers. It's not to say it would be the best thing you'd ever see, but it could pass in a low budget film just fine.

I don't film charts! Who cares about how good a chart looks in the end? If you have a detailed scene of actual moving life, that's the best test IMO. Does it look like a film or not?

What we actually learn from this is: the only thing that counts is a chart. Sony f3, canon c300, Scarlet and Sony fs100 obviously resolve less in a 1080 frame than xf300 and any real world experience is obsolete.

My next feature film will have a chart as it's main actor. That way I will know whenever I cock up the shot.

@RRRR - that chart was shot at 1080p on the GH2. Vitaly, I'm sorry about the link to another forum, but it's just satisfy naysayers who want to try to hang on to dogma despite all evidence. RRRR, you can see the chart, as well as a 1080i chart, MJPEG chart and 720p chart and more here: http://forums.dpreview.com/forums/readflat.asp?forum=1041&message=38912742

Not enough? Check out how these tests have the exact same results here:

And also at the 8:45 mark here:

But I'm sure you'll find a way to justify a reason why none of that is accurate. I'm sure some here could be presented with the a number of meticulous tests and charts and would still believe the GH2 can resolve exactly 1080 lines of resolution, even if you saw a test done right in front of your own eyes or if, god forbid, you did one yourself. The dogma runs deep.

Look, I'm the first one here to say that charts are irrelevant, except as a curiosity. What I can't stand, though, is untrue information being presented as fact just to further this idea that the GH2 is the greatest digital cine camera ever made (it's not). I feel sorry for people coming here to try to find honest information about their camera, and instead they just get a healthy dose of 'fanboyism.' The camera is just a tool, it's got limitations - all of them do. It's ok to understand what those limitations are. No one will think your d**k is smaller.

Just to put in a good word for charts - the attached jpeg (sorry, a bit low rez) is from a camera which has lots of passionate defenders and has been used to shoot features, commercials, etc. And good for the shooters, who got results with it. But, based on the the chart, most folks probably wouldn't choose to use it for a nature documentary, shots of building facades or men's tweed suits.

Which is why it can be useful to have charts -- to get a better idea of what your limits are on the screen, big or small. That was all the original poster wanted to know, and asking it wouldn't seem to be all that unreasonable.

I'm sorry cowpunk but you have taken the "evidence" you present out of context. Some high iso shots comparing high iso resolution differences between 3 cameras. Second you present a test that is only good for the shooter himself.. (it is lacking in terms of information). It is no better than other test footage that has been referred to or presented as evidence of what the gh2 is capable of. Let's stop wasting time with this nonsensical debate and get to work. Each to his own agenda.

Personally, I have no qualms over using the gh2 for shooting nature scenery whatsoever.. which was what the original question was referring to.

^ are those actual video stills? That's quite a dramatic difference. I didn't realize it was that much.