Time to record
HDSLR cameras use high speed memory cards to store captured video, the two popular choices are SDHC and CompactFlash memory cards. High speed cards are essential as video requires fast data transfer, this can be any rate between 15 to 35 M/bs (mega bits per second). The higher data rate produces better video quality, think along the lines of JPEG compression and you are on the right track. Many of the consumer camcorders offer longer recording times by reducing the data transfer rate, typically these may vary between 10 to 24mbps. If you want high quality broadcast safe video then you should be looking for a minimum of 50mbs or ideally 100mbs. The real benefit of the higher speed codec can be seen on fast action and panning sequences. These will be very smooth and more detail will be visible.Recording video
HDSLR cameras record their video directly to memory cards (SD for Nikon & Panasonic, CF for Canon), many professional camcorders are now also using solid state memory cards or hard drives. The advantage of shooting direct to memory cards is a faster workflow, and the cards can be used over and over. The disadvantage is that you need to download the data to a hard drive or other storage media and you should ensure that you have backup copies of all your files.
File sizesShooting times are limited with HDSLR cameras; Nikon has a limitation of just 5 minutes recording, Canon 12 minutes and Panasonic 29 minutes. This limitation is due to the tax laws and different tax rates for still and video cameras. However, although 5 minutes may seem like a handicap, in reality you would never need to shoot any sequence for longer than a couple of minutes or so. Look at any documentary on TV and you will see that most shots are only held for a few seconds. The only time you may possibly need a longer duration is in the case of shooting an entire concert or wedding ceremony.
We are all used to capturing stills using JPEG format, for video there is MPEG and AVCHD H.264 formats. HD movies can occupy several gigabytes for a short sequence, so in order to fit it all on a memory card the movie has to use compression. MPEG2 is a popular choice. This uses a LGOP algorithm, LGOP stands for Long Group of Pictures. Compressing a still image is relatively easy, just analyse the single picture and compress it. With movies it is slightly more complicated. Movies are made up from sequences of still pictures. In the UK we use 25 frames per second (fps), in North America they use 30 fps (29.97 to be technically accurate). These frames are divided up into Groups of Pictures, generally 12 for a LGOP. The first frame is sampled and then the next 11 frames are made up from the first reference frame, then the process repeats with a new key frame. For most practical purposes this works OK, until you start to do a Pan or try to capture some fast action, then you may end up with a smeared looking sequence. Ideally, a shorter group of picture would be the better solution, but this would result in far larger files.
A four minute sequence captured in Full HD (1920×1080) using MPEG2 compression will occupy 1048 megabytes of memory (1 gigabyte), the same four minute sequence without compression occupies 32,457 gigabytes of memory.