hf s11 vs camera HD
 

Hiya
I have a HF s11. I am not saying that i am not happy with but i am not especially impressed neither.
And i admit that in photo mode it can get quiet limited.

I was wondering if someone tried to compare the video coming out of still camera (7D or other good camera of course...) shooting HD against it.
some video i have seen was quiet impressive and had more shallow dof, probably due to some lenses.
And thinking about it this access to a big range of diffrent lenses available for camera seems to be a very interesting point.

Any comparisons ? any though ?
cheers
Olivier

That shallow depth of field you're referring to is mainly due to
the very large image sensors in those still photo cameras...
they are substantially bigger than than the sensor in the
HF S series, which is roughly 7mm measured diagonally.

Digital SLR cameras that have an HD video mode are equipped
with an image sensor that measures anywhere from appx. 28mm
(APS-C) to 43mm (Full Frame), so there's your reason for the
shallow depth of field they produce.

You can tell by browsing our various D-SLR boards that they
are indeed very popular cameras for shooting HD video, but
your HF S11 has a couple of advantages over them.

First, you can record very long clips, filling up your entire
card if you want to, while D-SLR cameras are limited to
twelve minutes per shot, more or less.

Second, your camcorder has a motorized zoom. D-SLR
cameras do not. Plus, you have autofocus, which they
pretty much have to do without in video mode.

Also, your camcorder is more portable and less obtrusive.
But there's no real right-or-wrong way to go here. They each
have their advantages and disadvantages... it's all a matter
of which one is right for you specifically.

so in theory, the SLR gets a better quality.

5 min shoot is more than enough
and the focus changes is so much easy to get with a manual ring anyway...

:(

thanks

Yes and no. More control over depth of field, generally higher contrast, a wider variety of lens choices, and better low light performance: yes. But also a lot less resolution and way more artifacts. For my audience and style of shooting, my 5D2 comes out of the bag a lot, but I still use the XH-A1 frequently for the times when I can't afford the image to be riddled with artifacts or I need high definition.

If by 5D2 you mean Canon EOS 5D Mark II, than I can't imagine how much better image quality (and resolution) you expect to get. 5D Mark II captures full 1920x1080 (progressive) frames, unlike HDV devices (such as your XH-A1), which only capture 1440x1080.

DSLRs have been gaining legitimacy recently, as more and more professional productions embrace them as their principal acquisition device. Most recent well known case was the season finale of the American TV show Dr. House, M.D.; the episode was shot entirely on 5D Mark II.

It's important to understand that the XH A1 and other XH and XL H series camcorders are also progressive scan when they're shooting in Frame mode. And there's no substantial difference in resolution between 1920 square pixels and 1440 anamorphic pixels. They both produce the same size image. There is no easily discernible difference between them when comparing moving video. It's a non-issue.

It has worse aliasing artifacts than any other video camera in the history of mankind. It wouldn't even make the top 1000 if cameras were ranked by aliasing. Some people like aliasing, but I find it very displeasing.

XDCAM-USER.com alias

It's important to make the distinction between capture and storage. If I connect a 4096x2160 storage device to a standard definition DV camera, the format will have the potential to store 25 times more pixels; however, the actual level of detail will be no better than if I recorded to DV

In the same way, the 5D2 has the potential to store a full 1920x1080 in its file format, but in reality the level of detail is very poor. On vertical details it's little better than 720p, and for horizontals it's little better than standard definition.

That's nothing. Several big budget movies have been shot entirely on standard definition DV. All that means is that for some people, resolution doesn't matter.

----------------------------------------------------------------------
resolution doesn't matter
----------------------------------------------------------------------

Agreed!

To the thread starter: If the DOF is your concern, try backing up and zooming in. The HV/HF series are capable of getting some pretty shallow DOF using this method.

That is not true. When you back up and compensate by zooming in, the DOF remains exactly the same. (In fact, the DOF becomes deeper if the f-number gets slower.) The background, however, gets more blur. It's important to know the difference between background blur and DOF. Blur from backup-zoom-in sacrifices perspective, whereas blur from DOF does not.

I've sean lab tests of those two, and according to the numbers, they don't differ by much, if at all. Vertical is for both 650 lpph, while horizontal is 700 for 5D, and 800 for XH-A1. Neither is actually greater than the 1280x720 image size for 720p, let alone 1080p.

As for big budget movies, that is certainly true. The point of mentioning the 'Dr. House' TV show was, the show is shot and broadcast in HD. It would be extremely difficult to convince producers of the show to approve significant difference in image quality of their usual output (mostly 35mm film), if 5DmkII couldn't produce the image quality and resolution that is close enough to what was before. In all fairness, though, the show broadcasts on Fox at 720p (i.e. not the full 1080i that many others use). 5DmkII should comfortably output enough pixels to populate full 720 raster.

For the last two years I've shot the XH-A1 and 5D2 together on many occasions. When I bring them into post the 5D2 always has much less fine detail.

As for the lab tests, I think you may have made a mistake somewhere. Did you mix up lpph with lph? The theoretical maximum veritcal lpph for 1080p is 540: that would be exactly one line per pixel. In practice, any camera with that high of a value is aliasing badly, so well-designed cameras tend to have resolution around 70-80% of the theoretical maximum, from 380-540 lpph. 650 lpph is not possible, even in a computer-generated image. The 800 lpph is correct for the XH-A1, but the 5D2 is definitely not 700 lpph. Here is a lab test of the XH-A1:

BBC R&D White Paper WHP034 - Alan Roberts

It shows a measurement of 470 lpph, 87% of the theoretical max. This is more than the optimum, so that indicates that it will be aliasing, and in indeed it is. Here are the zone plates next to eachother:

http://www.dvinfo.net/forum/canon-eo...n-roberts.html

The difference in resolution is striking, as well as the difference in aliasing. The XH-A1 has more aliasing than I would prefer, and less resolution than several other 1080p cameras, but the 5D2 is so bad that the XH-A1 blows it away. Even the little old HV10 has far more fine detail:

http://www.dvinfo.net/forum/canon-eo...ml#post1132912

Well, "close enough" in this case is probably based on using extremely thin DOF and careful camera motion to hide the aliasing and skew artifacts and just accepting the very low resolution.

But again, resolution and aliasing don't matter to everyone.