cowpunk: the chart above shows only that the user is possibly using factory firmware, shooting 50i because it is nowhere near the normal output.

@anthril - possibly, but even in low-contrast environs the xf300 outresolves fine detail when compared to the GH2's video output. Case in point are the 1:1 center crops i posted in the other thread on this subject. My guess would be simple user error, but without seeing footage for a framegrab it's impossible to know.

@bwhitz - the opinion that the XF300 image looks very video-like vs the GH2 is very valid. I'm sure that a lot of people might share that opinion, and it's not a bad thing either way. And, while I can't comment on why the XF footage you're working on right now looks soft, the fact of what the XF300/305 series is capable of in regards to resolution isn't really up to debate. Just look up any review or white-paper test of the camera.

I think @Anthril is on the right path here. Look, these are all just numbers and measurements and techno-babble whatever. It's ok if the what's on paper comes up short. What matters is picture - and the feel it is able to impart. And the GH2 is able to satisfy that need in spades. So if you want to know the numbers, that's fine. Just don't be surprised if the numbers don't live up to the emotional impact that the GH2's picture can give - because ultimately, that's all that matters.

GH2 doesn't match an ideal image of what you can get out of 1080p luminance wise, it's got a fair way to go when it comes to that, but leagues better than other 'dSLR' video cameras, and it's basically trash when it comes to colour resolution, worse than what 4:2:0 should be from examination.

But it's more satisfying than any other camera I've shot motion picture with so far (I haven't shot higher end digital equipment, just things like HVX varicam and others), apart from Aaton XTRprod (Super16) @ 2K scan which wiped the floor with anything I've seen digital.

That said there is nothing wrong with a single chip sensor, in fact, they do excellent, and demosaicing works (not total decimation like chroma sub sampling) even @ 100% off a still with a sharp lens it is excellent.

If it was sampled down perferctly, that huge amount of oversampling would provide tremendous image quality, and it would provide the maximum possible resolution/information you could store in a 1920x1080 size image if output in full RGB or 4:4:4. Single sensor wouldn't come into the equation since the huge amount of oversampling done, the effective down-sampled pixel size would be bigger than a cluster of RGB it came from, so essentially it would be "3 chip".

The downscaling used isn't ideal though (though much better than line skipping by a long way), so there is resolution loss, then comes compression.

Interestingly enough, just found this on dpreview - GH2 1080p AVCHD frame grab of a resolution chart. You can clearly see aliasing begin between 600-700 lines of resolution:

@RRRR - I think we're discussing video resolution and not stills resolution from the GH2. The two modes give very different results. I've yet to see a resolution chart for the GH2's 1080p video output, so definitive statements on lines of resolution are just guesses until someone does an accurate test. I did get a chance to do my own fine-detail comparison between my GH2 and XF300. Results were posted here in the thread you linked to above, and show the XF300 clearly resolves more detail, but I don't make any claims on the actual lines of resolution that the GH2 is capable of. My guess is 800-850, but that's just a guess. The EX3, like many cameras, softens as gain is applied. I don't understand what you're trying to say with "ever heard of sharpening?" and "true resolution indeed."

Cowpunk, if you doubt my word, you are free to look up any gh2 stills resolution test (there are plenty to go around) and do a calculation on an 1080 frame from there. That leaves bad settings and bad handling out of the picture. It's not rocket science. And there's plenty of footage which time after time shows the amount of detail the gh2 can resolve, so that is not exactly dodgy, nor misinforming either. Ex3 softens with low light? Hang on.. ever heard of sharpening? (and what it does to noise levels).. True resolution indeed.

Mind staggeringly bogging is an opinion.

The amount of misinformation and conjecture being spread around this forum as established fact, with nothing to back it up, is absolutely mind-boggingly staggering. So someone just jumps up and says the GH2 resolves 850-900 lines of resolution. Offers no test charts, no proof, nothing - and everyone just agrees on this "fact." Someone says that the the GH2 resolves more than an EX3 and an XF300 while pointing to dodgy samples taken at night at more than likely high gain (at which point the EX3 softens considerably to reduce noise) and it's just just accepted as "fact." Nevermind that I posted high-detail comparisons in that exact thread that showed just the opposite.

Look, guys - if you want to know an exact measurement, you've got to do an accurate test.

The GH2 is the first camera that I've used where I've been officially satisfied with the resolution. I loved the colors of the HMC150, but the softness, especially on wide shots, always bugged me.

There you go, there's the answer. 850-900.

Your debate around the importance of cheap resolution and theatrical release success is an oxymoron. Most succesful theatrical releases are due to overall budget and production value.

Fact is, if someone has secured talent with a $1m fee then they're not going to be shooting on a $700 GH2, unless there's a physical need for it, like shooting in ultra confined quarters and the DP knows it'll cut well.

The GH2 for the larger extent is a low budget camera. And low budget productions don't traditionally do financially well. That includes 5Dii, FS100, 7D...etc.... There's nothing stopping low budget being successful in DVD, TV and foreign markets.

Personally I'd take that, if it leads to larger productions with a budget to shoot on something better than a $700 camera. Until then, I'm so so pleased with the quality that is achievable within my means.

@JeffGibbsTC - "Does the latest firmware upgrade and the amazing new PTools and assorted hacks actually raise the actual lines of resolution resolved by the GH2?"

No. The resolution characteristics of GH2 video are determined by the image sensor, along with its Optical Low-Pass Filter (OLPF), and the proprietary demoisaic algorithms used to extract 1080p and 720p frames from the 16Mpixel sensor. I don't see any evidence that those algorithms were altered in Panasonic's v1.1 firmware upgrade.

"Lines of resolution" is a technical judgment that is difficult to pin down on digital cameras. The problem is that the OLPF doesn't block all of the spatial detail transmitted by the lens that lies just beyond the Nyquist cutoff limit of the sensor's photocell array. As a result, these hyper-fine details are mathematically folded down into the high-resolution region of the sensor and create synthetic aliasing artifacts that artificially enhance the sharpness of genuine edge details. Panasonic distortion-corrected lenses exacerbate this phenomenon with their built-in firmware sharpening filter - you can actually see the moire patterns shimmer on the LCD as the lens locks into focus.

This contamination of genuine image details with synthetic aliasing artifacts cannot be undone in post-processing. That makes it impossible to pin down at exactly what resolution point the genuine details fade into gray and the aliased details become dominant. Since lens resolution and diffusion filters also have an effect on the intensity of aliasing artifacts, the most relevant way to judge resolution is to shoot calibrated test patterns with your complete rig and evaluate the results to your own satisfaction.