I found this interesting video here. flower-4k on VG20E on Vimeo
The author mentions that the camera actually shoots in 4k (rather than using smaller portion of the sensor in the first place) and then the .MTS format makes it 1080p. He also decodes the format and reverses it back to 4K by the looks of it though the image quality isn't of 4K perhaps because even by reverse engineering the extra image data may of been lost.

Sounds like BS to me. He quotes the photo resolution as being the actual video resolution and his idea of reversing it back to 4K is just up-rezing the 1080 footage to 4K.

Hold on while I go shoot some 3264 × 2448 video on my iPhone.

Maybe so, but then again you never know with Sony. Just because the features aren't mentioned im sure the hardware could probally output a lot more than we think. Its just they love to sell overpriced cameras which lack basic features, and so the way to justify stupid price tags is to add the missing ones in new models.

The VG10 could of had power zoom from the start, it seems their ideas lack planning or just are based on profit margins.

I too call BS... clearly there is a lack of understanding of the way things like this work.

Yes, the "native" sensor resolution is higher than the "output" resolution - in theory with sufficient computing on board, you might be able to process "all" the pixels and produce a higher resolution output. But as a practical matter, part of the "math" involves DISCARDING at least some of the data to keep it manageable, hopefully with as little quality loss as possible..

The apparent theory is that by somehow "reversing" the math you can reproduce the original data... and I guess the idea was to produce a "better" video? I'd sooner suggest proper stabilization and focus... but that's just me.

And while I've seen algorithms that supposedly claim to "recreate" data points from highly compressed files, I've yet to see it in a commercial application (other than existing standard compression/decompression methods) - it works great on crime dramas and in theory, but creating something from "nothing" is still a speculative endeavour... one which I have to question when anyone claims to have done it...

Time to go back and make more gold, but my lead supply is running low...<wink>

This may not be as crazy as it sounds, im doing some test today and will post the results for review

test # 1 4k test3 - YouTube

Test #2 http://youtu.be/BuIgld-Q2MI

Now let me first say my method is completely different from the first video in this post ,but i think after buying my vg20 i have the right to get the best out of it

so let me explain how i came up with test 1&2 after looking at the specs of the camera it is possible for the vg20 to take a still at 4912 x 2760 dots /16:9 in photo mode ,i simply put the in continuous shooting from the drive mode menu and shoot continuous stills of my daughters stuffed toys, now this creats a whole bunch of stills, i then imported them into premiere on the timeline in a 4k project and then exported them as continuous video at 24 fps to h.264 file at 4k 4096 x 2304, which is like the standard for Red , now i know the video is not as smooth as one would like it to be but after all its just the first round of testing

Now while exploring the camera some more i discovered something interesting and you can try it for yourself ,
plug in an hdmi monitor or tv to the vg20 now in video mode 1080i/1080p looks just fine

Now press the mode button to still mode set the image quality to 16:9 13.6M
not setup a nice shot of whatever subject matter you choose making any adjustments you need
Now toggle back and forth using the mode button and watch you monitor , the live 4k view is far better , now if i can record the hdmi output to another device say like the Ninja we can get a better quality out of the camera, yes i know you cant get anything more than 1080P out over hdmi , but hey thats why im testing

Im not the only one who tested this theory

IBC 2011: Testing the SONY VG-20 - recorded via hdmi out - YouTube

I just tried the photo mode you suggested out to my HDMI TV and note that when I was pointing camera at grating on an air conditioner, using any of the photo settings, the image suffered much more from moire and aliasing. A lot worse than in video mode. I do have a Black Magic Shuttle which captures through HDMI cable. The file attached here is from a file I shot on the shuttle. Starts in photo mode, switches to video mode.

There is obviously something wrong with your video i took a screen shot of your settings and i think you need to make some adjustments and maybe choose a better subject matter that is properly lit and redo the test

Suggestion ? What setting do you recommend ?

Point of this test was that I shoot this same test in this same place with my all my cameras, and this is worst aliasing and moire I have seen on any of them. (Canon HV20, Sony FX1, Canon 5D Mark II, and the T2i.

I think its pretty clear that the HDMI output in photo mode does nothing to address aliasing/moire issues. I suppose I could set up a well lit scene, but that won't change the aliasing moire problems. AS you can see from the graph, exposure is flat and right in middle of where it needs to be and only at 9 db