HF-S10: All the Pixels Useful?

[Yellowdog321]

Hello. I am currently debating whether to purchase the HF11 or spend an extra $500 on the HF-S10. I really like the features of the HF-S10, but am really wondering about something. The HF11 has a sensor just over 2 megapixels effective, while the HF-S10 has just over 6 MP for video. My understanding is that full HD (1920X1080) is coveed by the 2 MP sensor of the HF11. So what is the use of having a 6 MP sensor? It would seem to me that all this does is require that the video be down-sampled to achieve 1920X1080.

Is this just marketing gimmickry, or is there any real practical advantage to the larger number of pixels in the HF-S10?
Thanks.

[Eugenia Loli-Queru]

For me, it's not even about the pixels and quality. It's about the zebra support that the HF-S has and the HF don't. Do you think it would matter if the HF had a sharper image but all your shots were over-exposed? It would still look bad. In my opinion, don't buy a camera that doesn't have zebra support. Regardless of any other feature that may have.

[Ian-T]

It would seem to me that all this does is require that the video be down-sampled to achieve 1920X1080.

Is this just marketing gimmickry, or is there any real practical advantage to the larger number of pixels in the HF-S10?
Thanks.The image would (should) be sharper as a result. Think of it as oversampling. For example...an HD image downsampled to SD would (should) render a much better (more detailed) looking image than even the best SD camera out there.

[thequads]

Another plus of the pixels could be the hf-s 1.7 tele-converter function.. which acts like like a telephoto lens adapter. Supposedly it's not like a typical digital zoom where you loose quality, but a zoom that takes advantage of those unused pixels to maintain image quality.

edit:

okay here's the stat from the manual:


sensor =8,590,000 pixels

effective pixels:

movies = 6,010,000 pixels

using tele-converter = 2,070,000 pixels
can anyone tell me if going from 6,010,00 to 2,070,000 pixels will show a loss in qualitiy?

or wait a minute.. is that 2,070,000 pixels added to the 6,010,000 pixels already used?

thanks

[JCM_GDL]

Another plus of the pixels could be the hf-s 1.7 tele-converter function.. which acts like like a telephoto lens adapter. Supposedly it's not like a typical digital zoom where you loose quality, but a zoom that takes advantage of those unused pixels to maintain image quality.

edit:

okay here's the stat from the manual:

can anyone tell me if going from 6,010,00 to 2,070,000 pixels will show a loss in qualitiy?

or wait a minute.. is that 2,070,000 pixels added to the 6,010,000 pixels already used?

thanks

2,070,000 pixels is enough for HD, and those 2,070,000 pixels resides the same sensor, using the pixels in a virtual 1920X1080 frame inside the sensor in the center of it.

I think that even, due to the HD resolution, one can go beyond the 1.7X with some digital zoom in order to render a little increased picture of some detail in a shot, without noticeable pixelation or degradation of the image.

[Sagefox]

Ian is correct... pixels do not make for added detail... they are only a grid on which contrast can be registered. Therefore, it is contrast that leads to detail, and so to capture detail that will be visible at the per-pixel level, you need many more pixels (for capture, not display... 2k/1080p is enough for most display situations including theaters).

An excellent industry based research paper is linked to here:
http://www.rent-cam.com/blog/2008/03/28/arri-4k-systems-theory-basics-for-motion-picture-imaging/
That is the very best info on resolution there is.

Oh, another cool detail is that this amount of pixels when added together minimize apparent gain noise.... this is what very powerful hollywood cams have been doing with their sensors even before solutions were created to store 4k raw motion.