DV Info Net - HD-DSLRs and EVILs used as camcorders --part 1

Now that my NEX eBook is done, I've been thinking about what I learned that would be generally helpful.

1) IMHO DSLRs don't make good camcorders because when shooting video the mirror is up and you only have the LCD. Clearly if you are making a film and have an HDMI monitor this can be overcome, but for most of us who shoot by themselves, this is an issue. Especially if one shoots under bright sunlight. If that poses no problem for you, the new Canon 600D is very nice. Depending on the price, Sonys HDMI monitor on a NEX would be an alternative. If you need a VF, the ones offered on the Panasonic G-series and the VG10 are very good.

2) Mirrorless designs work best for video shooting. But, you need to look carefully at the details. The Sony a55 looks great until you learn that unless you turn OFF the wonderful AF system, you have no control over video shooting. Panasonics G series offer total control as does the VG10. But, as I'll show in part 2, total control in any of these cameras is not the same as total control in a camcorder.

3). Looking through specs on a number of cameras, in general even when they offer a huge range of ISOs, when shooting video, you only get 3 stops (18dB) of gain. Which is about the same range offered by most camcorders. So if video is shot with AUTO GAIN, the maximum gain will be 800 or 1600 depending on whether the minimum gain is 100 or 200. Thinking in terms of film ISO, that is about one stop too much. Yet, looking at video shot with a NEX at 18dB gain, one doesn't see much noise. And, this is where numbers don't tell the story. Sonys chips use CDS noise reduction and have over 4000 A/D converters. Other companies may or may not use the same technology. Therefore, the visual noise at a given gain level may not look the same.

Keeping chroma noise low is far more important than keeping luma noise low. DSP based noise reduction can be used. 2D DNR under motion doesn't work as well as 3D DNR.

This is where the debate about chip size and pixel density comes into play. Some companies can pack 14MP into a slightly larger than 1/2 inch chip and others can't. Or, can't in 2010. Thus one really can't debate generalities. One really needs to look at video from specific cameras.

4) one thing that we can say, is that the larger the chip the smaller the aperture that can be used! I'll talk more about this later. This makes a huge difference when shooting outside.

5) the other thing we can say about chip size, the smaller it is, the smaller the optical system can be. Which means a 24X zoom lens can offer a really wide -- wide angle with a really large aperture (f/1.8 to f/2.8). This makes a big difference when shooting indoors. More later.

6) it goes without saying that even when a camera provides "full manual control" there are huge differences in HOW EZ it is to use the controls under pressure. And this is true of camcorders as well. Thus full control may not be practical to use even when available.

Likewise, you can't a fully trust photo reviews. The Panasonic G2 can be said to offer only aperture control like a NEX. But, it turns out that there IS a menu item called flicker control. The intention is when you are shooting in a country that uses a different poweline frequency, you can enable it. So when you take your 60Hz camera to Europe you can switch this on and the shutter speed will go to 1/100th. But, once enabled, you can also turn a control dial and select 1/50 or 1/60! Photo reviewers call this "limited" control. Videographers would call this perfect control!

7) unless you buy a RED or one of JVCs new cameras, you know that no matter the number of pixels on a chip, that number is not going to be used when shooting video.

http://www.dvinfo.net/forum/digital-...ml#post1620053

Typically only half the rows and columns are actually read from a chip when shooting video. This is necessary to keep heat under control. (Which is also why some chips are run at only 30Hz and not 60Hz.) The bottom line is that only about 2-3MP are processed by the DSP. Skipping rows/columns can produce nasty artifacts.

Panasonic uses this fact to provide a neat feature. Most short zooms provide a maximum focal length of 85-100mm which IMHO is just too short. So Panasonic enables a readout of the central 3MP which naturally captures a field of view that is over 3X less. Presto, you now have a 45mm to 300mm zoom lens. None of the pixels in this window will be discarded. So, in theory, there should be less artifacting. (But, since the anti-aliasing filter has been set for the full chip, the camera is majorly subsampling and aliasing may be horrible--unless electrical filtering is used. Once again, you'll have to check for yourself.)

PS: issues like these are present with the $6000 Panasonic AF100 and Sonys future NXCAM. Putting a "big" sensor in a camcorder looking body does not get you a low cost RED. In fact, one report on the Panasonic reported that is clips highlights just like other low cost Panasonics do. Which means that after 2-3MP have been read from the big chip, the same processing circuits may be used as are used on their other low cost camcorders.