In order to get the best possible results, we have to change our shooting techniques. One of the problems with using a HDSLR is that the camera itself is not ideally suited for capturing video. In order to see what you are capturing you have to hold the camera away from your body to view the LCD screen, this can introduce camera shake, which in turn will make the reference LGOP key frames less than ideal for sampling. A good tripod will cure the problem, but then this also takes away some of the spontaneity that is enjoyed when shooting on a HDSLR. Camera tripods are not ideal either, if you are serious about video capture then invest in a sturdy video head with a decent pan & tilt.
Panning too fast across a static scene can also make the image look blurry, a slow pan using a tripod will give the best results. The CMOS sensor, as used by Nikon and Canon, builds up the picture on the sensor from top to bottom. This can cause distortion when panning as the bottom section of the sensor will receive light after the top part, so if you are panning then you may notice some skew in upright lines, this is often referred to as a Jello effect. If you are using the camera in video mode during a wedding or other function and another flash is fired, then you may end up with a frame that has the flash captured on the upper or half only. A CCD sensor does not have these problems.Slightly blurred video can be due to the slow shutter speed that a camera records at. Video is usually captured at a shutter speed that matches the frame rate, eg., for Pal systems a cameras shutter will expose each frame at 1/25 sec – 25 frames each second, NTSC systems use 1/30 etc. You can use a shutter speed that is twice the frame rate – 1/50 or 1/60 to reduce the blur, faster shutter speeds could result in an unnatural strobe effect.
One problem is that many cameras turn off their autofocus and auto exposure once a capture has started. An out of focus HD sequence is really not worth looking at, especially when viewed on a 50inch plasma screen. Generally, this shouldn’t be too much of a problem, as most professional videographers never use auto focus or auto exposure. Using a HDSLR you can set up the focus and exposure through the viewfinder first, then switch to Live View mode to start the capture, this is not ideal for capturing a fast moving action sequence. To help with this there are a number of third party viewfinders available that clip directly onto the rear LCD, thereby making it easier to follow the action, especially on a bright sunny day.
As mentioned earlier, High Definition video can be captured at either 25 fps (PAL) or 30 fps (NTSC) this is, in effect, a very fast motordrive. The real advantage is if you want to take out a single frame and create a still image from it, you now have 25 or 30 frames every second to choose from. To get the best quality you should shoot in Progressive mode (25p or 30p). This means that every frame is a complete picture. The more common method is Interlaced which is actually twice the number of frames (50i or 60i) but each frame only contains half the information. The two interlaced frames are combined on alternate lines to make up a complete frame. This can produce smoother looking action but can be more difficult to obtain a crisp still image. A grab frame can be De-Interlaced in Photoshop to remove some of the tearing effect where the subject may have moved between the two interlaced frames.