- Product Reviews
- Gear Guides
- Tip and Tutorials
- Adorama TV
A still photographer's guide to HDSLR technology, terms, and features
High definition movie recording features are found on nearly every new DSLR and interchangeable lens compact camera. But a wide variety of video formats, quality levels, and video-related features make it more confusing than ever to pick the right camera for your still and video needs.
The HD Video DSLR revolution started with this camera...
The convergence of still photography and video recording has been happening in compact digital cameras and camcorders since the late 1990’s, but it wasn’t until Nikon added HD video capture to its D90 DSLR in August 2008 that serious photographers and even dedicated videographers got excited. Why? Compared to similar-priced camcorders, the larger imaging sensors typically found in DSLRs enabled extreme depth of field control, expanded low light capabilities, and wide angle views that were once the domain of professional camcorders costing over $20,000.
DSLR makers have branched out, adding interchangeable-lens compact cameras, such as this Sony NEX-5, with DSLR-sized sensors and HD Video recording abilities.
As a result of those capabilities and the buzz around the D90, nearly every other major DSLR camera manufacturer has followed Nikon’s lead, with each new generation overcoming more of the shortcomings of early model dual-media DSLRs, such as lack of autofocus or exposure controls in video mode. Now, more than a dozen DSLRs and compact interchangeable-lens cameras that featuring video capture (the latter of which include Micro Four Thirds models from Panasonic and Olympus, and Sony NEX cameras, as well as the Samsung NX-100), but these vary greatly in terms of video quality, controls, storage, and ease of use. And if you don’t have a video background, the terminology used to describe video capabilities can be very confusing, even for professional or experienced amateur still photographers.
In less than two years, HD Video in one form or another has become a standard feature on new DSLRs at all price levels. The Canon EOS 7D, above, is one mid-range example. But do they all deliver equal quality and ease of use? There's the rub!
This primer is designed to help you navigate the maze of confusing video rhetoric and specifications out there, and enable you to make educated decisions about the video features and capabilities of the latest DSLR and compact cameras (as well as HD video camcorders).
Let’s start with...
Just like the different resolutions offered by today’s flat panel HDTV’s (all of which display up to 16:9 aspect ratio videos), there are multiple video resolutions and settings available in interchangeable-lens cameras:
Full HD refers to 16:9 aspect ratio video captured at either 1080p (progressive) or 1080i (interlaced). In 1080p video, which is usually recorded at either 24 frames per second or 30 frames per second, each frame contains 1920x1080 pixels. In 1080i video, which is usually recorded at 30 frames per sec, each frame is made up of two fields containing 1920x540 pixels. (It may also be listed as 60 fields per sec or 60i in the camera specs). 1080p video records each frame without the motion artifacts found in 1080i video, and provides higher quality still images from the video for use in small prints or on web sites. It’s also the resolution used for most Blu-Ray movies (although your player or HDTV may convert them to 1080i if that is your TV’s maximum resolution).
Near HD (or HD) video records in 16:9 aspect ratios at either 720p (1280x720 pixels per frame), or 720i (1280x360 pixels per field, two fields per frame). Typically, 720p resolution is what you’ll find in commercial DVD movies and many broadcast HDTV programs. Earlier model DSLRs, such as the D90, recorded video in 720p resolution, but many newer DSLR models feature up to 1080p resolutions with a choice to capture 720p as needed to save on storage space (more on that later.)
The Pentax K-5 offers a 3:2 default aspect ratio, setting it apart from other HDSLRs which typically deliver a cinema-standard 16:9 ratio.
Custom HD such as the 1536x1024 pixels per frame recorded by the Pentax K-5, are usually slight variations of the above determined by the camera’s aspect ratio. (Basically, the Pentax K-5 records similar detail to a 1080p camera, but the video is cropped to 3:2 aspect ratio instead of the wider 16:9 aspect ratio video).
iFrame format is used in many Apple video editing programs, such as iMovie, but it's now starting to appear as an option on Full HD and HD cameras. It records 960x540 pixels per frame at 30fps.
VGA (also called 480p) resolution is offered as a choice on most cameras and camcorders, and records each frame at 640x480 pixels. This is adequate resolution for viewing on older, non HD TVs with a 4:3 aspect ratio. A slight variation called WVGA records similar sharpness but at 16:9 aspect ratio, for 848x480 pixels per frame.
More on frame rates
In the United States, NTSC broadcast standards require video to be recorded and played back at 29.97 frames per sec. That rate is often rounded to 30fps or listed as 60 fields per sec. Some DSLRs can be set to either a true 30fps (perfect for playback on computers on digital TV) or 29.97fps (if video is destined for broadcast.) However, the more popular choice is between 30fps and 24fps (the frame rate closest to that used in cinema projections.) Some cameras can capture 1080p video at up to 24fps, and can play them back on HDTVs as 30fps movies. However, due to some frames being interpolated, you may notice that 24fps video played back at 30fps on an HDTV appears slightly jumpy compared to video captured at true 30fps.
(Note: The 60fps frame rates found in the specs for 720p cameras stem from doubling each frame captured at an actual 30fps. This is done to accommodate the AVCHD Lite storage standard, and doesn’t actually improve video quality over 720p movies stored at 30fps. )
Stills from Video
If you do the math, each frame of a 1080p video contains more than twice the number of pixels per video frame as 720p video, and all things considered equal, that equates to approximately 1/4 more resolution. (That’s the same difference you’d see between a still camera capturing 2MP stills and one capturing .9MP stills.) But in reality all factors are not equal, and image quality and sharpness can be affected by the compression codec used, as well as the camera’s exposure and focusing accuracy. The bottom line is that each “sharp” frame of a video captured at a true 1080p resolution can be used to make up to a 4x6 inch print at near photographic quality, and is more than adequate for use as a full screen image on a web site or in digital slideshows.
Setting your camera to capture video at a 16:9 aspect ratio (or close to it) instead of one of the lower resolutions at 4:3 aspect ratio is usually a better choice. True, video shot at the narrower 4:3 aspect ratio displays without blackout areas on older TVs and most video projectors, plus takes up less storage space. However, a wider 16:9 video format is more future-proof, and fits better on flatpanel HDTVs, mobile smartphone displays, and on HD video-hosting websites such as Vimeo. Plus the wider 16:9 aspect gives you greater leeway to crop the video or resize it to a narrower format later.
Video Compression Algorithms
If you thought that the variety of compression choices available for still shooters (including various JPEG levels, TIFF settings, and RAW format settings) was confusing, wait till you try to master video compression options! Basically, the compression/decompression algorithm (also called codec) used by your camera determines the amount of memory needed to store your video as well as affecting its compatibility with video editing programs and mobile displays. Just like the JPEG “quality” settings found on most cameras, selecting a video codec designed to maximize storage space may degrade image quality—especially in scenes with lots of action or moving objects. But at the same time, advanced algorithms used in some codecs can actually reduce storage requirements by 4-8X over older codecs, and still deliver higher quality video!
So far, no DSLR has offered a “RAW” video storage option (only a couple high-priced commercial video camcorders have the equivalent of a RAW video setting.) The reason is that uncompressed 1080p video with sound requires approximately 160MB per second of storage. At 30fps, that adds up to almost 10GB per minute!
So, each camera compresses its video using one or more of the following options:
M-JPEG (Motion JPEG): The Nikon D90, most early dual-media DSLRs, and several current models use M-JPEG compression to make the video stream more manageable. MJPEG compression works by treating each frame of video as a separate JPEG, resulting in “bit rates” of approximately 40 megabits per sec (Mbps) for typical 720p video. (Since each byte of stored data is made up of 8bits, M-JPEG 720p video should fill up a memory card at the rate of roughly 5 MB per sec, or 300MB per minute.)
Because the data found in each frame is compressed separately, M-JPEG video files are larger than video optimized formats which may duplicate some data between frames or use i-frames (more on those below), and no cameras offer M-JPEG compression for 1080p or 1080i resolution settings. However, the advantage to this compression choice is its nearly universal compatibility with video editing programs and ease of playback on most computers—even older ones.
The Nikon D3s uses M-JPEG compression. Newer cameras are moving towards H.264/AVC, which offers more compression efficiency.
H.264/AVC (advanced video codec) While even some higher-end cameras such as the Nikon D3s and Pentax K-5 still use M-JPEG compression, it’s rapidly being replaced by H.264/AVC. This codec can compress Full HD 1080p video up to four times smaller than M-JPEG, but requires a faster, more powerful camera processor to compress the data (resulting in slightly higher cost to the manufacturer.) In addition, most older computers (3 -5 years) have a hard time decompressing this codec (resulting in freezes or skipped frames) and fewer older or low-cost digital video editing programs are compatible.
Yet due to its compression efficiency, 1080p video recorded on a camera such as the Canon EOS 7D using H.264 codec requires nearly the same amount of storage space as lower resolution 720p video recorded by the Pentax K-x using the M-JPEG codec.
MPEG: This codec was once widely used in compact cameras, and may still be a choice in some interchangeable lens cameras or cheap pocket camcorders. However, its use for Full HD or HD video is questionable, as it uses an i frame structure that throws away data between full resolution i frames. For example, if detail in a background hardly changes between frames, that area might be heavily compressed (visually blurred) in all but a few frames of the video. The detail found in the i frames can fool the eye at normal playback in a low res video, but lower quality is obvious at slower playback rates or at HD resolution and larger screen sizes.
After compression, video and sound data (yes, the sound data is also compressed, with a wide variety of choices outside the scope of this article) gets stored in one of four file standard “containers”, which are identified by .AVI, .MOV, .MP4, or .MTS extensions, depending on the camera and resolution setting. Cameras using M-JPEG compression are more likely to use the .AVI container, which limits them to 2GB-sized clips; while those using the H.264/AVC codec tend to use .mp4, .MOV, or .MTS containers.
Again, there’s a catch: The software or hardware you use to play back or edit your video with must be compatible with both the file container and compression formats used. So, it’s possible that a 720p video compressed using the H.264 codec and wrapped in a .MOV container (originally developed for Apple QuickTime) might not open in a PC editing program, but another H.264-compressed video wrapped in an .mp4 container will.
AVCHD or AVCHD Lite
The Panasonic G series and Sony NEX cameras, as well as several HD camcorders, use a more elaborate file container known as AVCHD or AVCHD Lite to store H.264 compressed video. In this system, the camera first saves its H.264/AVC videos in an .MTS container, then places it in a STREAM folder three levels down from a master folder on the storage card labeled “Private” (see photo, below).
Several other folders and subfolders store data that can be used by AVCHD-compatible programs to create slide shows and other features. Unfortunately, this system makes it much more difficult to find and open your videos directly from your memory cards using your operating system finder. Also, few affordable programs (and next to no mobile devices) are compatible with AVCHD or AVCHD Lite, or as in the case of Apple iMovie, automatically convert them to another format that takes up more storage space.
Final word on image quality
A camera’s resolution and the compression codecs it offers can affect the image quality, storage required for video, and compatibility with video editing programs and computers. But that’s not all you should look for when deciding on a dual media DSLR or interchangeable-lens compact camera.
Other important features for improved image quality include the ability to autofocus and track action in the video mode (the recently announced Panasonic GH2 and Sony Alpha 55 appear to be the fastest and most precise in this department). An image-stabilized lens or a camera with built-in sensor stabilization is also a big plus, as is a lens that focuses or zooms quietly.
Shooting video requires more power than shooting stills, so a high capacity rechargeable battery is a plus, and cameras with a built in stereo microphone or a separate port for an external stereo mic can greatly improve sound quality.
Now that you know the basics, which camera is best for you? Learn more about video-enabled DSLRs by reading Adorama's HD Video DSLR Round-Up.